index

Putting it all together...

I have always wanted Radiant Floor Heating. It is the greatest invention ever. The greatest heat efficiency is at a low temperature difference, and you have it with Radiant Floor Heating. Heating is low in temperature, distributed over a large area, and is placed low in a building. All three are optimum conditions for efficiency.
Also, required is a lot of cement for Thermal Mass. Thermal Mass is the number one deal breaker for an RV; You can not carry around tons of cement. The best that can be done is to use water. The number of PEX lines would have to be increased, and they would be placed up against each other with no spacing between.
The water in the lines would be the Thermal Mass, as well as metal brackets to hold the lines up tightly against the floor.

I got under the RV and observed two problems. Three tanks were in the way, a deal breaker. And also, the light construction used in an RV would not allow drilling holes in the floor joists. An RV is flimsy enough without a lot of holes drilled in critical support structures. That is a no-go for pex lines under the floor.

But lines can be laid on top of the floor in two areas: the sides. Both sides of the RV can hide warming lines under sinks and cabinetry. And, in addition, these areas have genuine warming needs of there own. There are water lines all over the place that need freeze protection. There are Lithium Batteries. There are water filters and pumps. Good to have these areas never freeze. I already have electrical warming that automatically come on at 35 degrees. But more is better. And I have internal BMSs, and smart code, for the Batteries. But again, redundancy and more is better.

This area is not a loss. Besides freeze protection, there is Comfort Heating. In the nonwalking space on the sides, there is not much floor heating. But the extra heat here will still make its way into the living area. Not as efficiently, but still a win.

Another plus, with planning, is that I will have, for the first time, heated water at the sinks. RVs suffer from this problem in general; It takes time for the hot water - from a 6 gal tank located far away - to finally arrive at the faucet tap. Tankless Water Heaters are not the solution for Calorie Conscious Solar users, although they can be placed near faucets. An RVer can not just sit there waiting and watching his valuable water run down a hole in the center of the sink.

I have solved this problem years ago for the shower. But it works differently. I have a brass valve that you can rest your hand on. Turn on the valve and the hot water is diverted back to the FreshWaterTank. The water is not used up, rather just put back in the tank from which it came. And the process does not require an additional dedicated pump; it is under pressure. After about 30 seconds the water turns warm, as your hand can feel the brass heat up. Turn off the valve and open the shower head above. Instant hot water.

But we have never had hot water at the sinks. Ever! And sinks are different from the shower. At the shower, 30 seconds is a minor wait. It takes me 30 seconds to get my clothes off and to pull the shower curtain. Sinks are different; When I come in to wash my hands, I have no time to wait. My hands are dirty, and I need water NOW! My little heated plastic filter does not work good enough. The response time from the little reservor is great; It is just not hot enough. It is fast. I have it withen a couple of seconds, it is just luke warm. And at night, forget it! You have to throw a switch to feed it from the batteries, which takes 15 more minutes to warm it up.

The PLAN...
For many years now, I have heated my 60 gal FreshWaterTank with a dedicated Solar Hot Water Panel up on the roof. It uses a lot of copper tubing. For hot water, a solar thermal panel is more efficient than a PV electrical panel. But its use is restricted to only for hot water. Hot water is so important that it deserves its own dedicated space on the RV roof. In my system, it works great as a booster for the Hot Water Tank. But the 60 gal tank never gets to 100 degrees on its own. It never gets hot enough for the faucets or, heaven forbid, radiant floor heating. But I have an idea about floor heating: The 6 gal Hot Water Tank does get hot enough!

I will build a Circulating path for the water. Under microprocessor control the water will actively circulate, starting first from the HotWaterTank, through normal PEX lines, to just under the Hot Water faucets. Then to 50 feet of PEX or vinyl, snaked across the flooring, and around key chosen areas of water lines, filters, pumps, and batteries. The line returns to any "Cold-Side" that returns to the HotWaterTank through normal lines. Upon demand, the water circulates hot water through both hot water lines, and returns in cold lines of the normal Hot Water System. The down side is that the "Cold Water" Tap will be only 10 to 15 degrees colder than the Hot Tap. And that is not really "cold". And this may be a problem for filling ice trays. I can not imagine any use for cold water - unless ice trays.


I went out to the RV...
Outside temperature is 52.1F degrees. Inside temperature of the living area is 56.5F degrees.
I do not want to disturb the inside temperature until I run these tests.
I could easily turn on the furnace floor heat and heat up the inside to a nice comfortable 76 degrees. But no.
I could easily turn on my modified oven, converted to a heater/oven and vented to the outside. And heat things up. But no.
The Circulator could THEORETICALLY raise the inside temperature! If so, I need to measure this.



It looks like a real mess under the bathroom sink.
My only defence in the appearance is that with tubing, you can always run more lines in three dimensions than two. If you do not buy it; "Confused brain" is possible too.

The bulky blue device on the left is a homebuild Tankless Water Heater, a plastic WaterFilter, heated by light bulbs. It also serves as an immediate heat reserve for the faucet, a few inches above. I am not happy with it. It only produces barely warm water. Just noticeably warmer. And during the day it sits there sucking on my solar, and at night, after an hour of cooling down, it has nothing left and does anything at all.

Things are about to change...
The normal hot water line comes into this WaterFilter, and leaves with a few inches to the HotWaterTap of the faucet. But now, between the WaterFilter output and the Tap is a Tee. The Tee feeds both the normal tap faucet, and also 50 ft of vinyl. The vinyl returns to a Flow Display (Which was leaking and I had to remove it.) and to a vane pump. The pump pumps the hot water, which now has cooled, to the cabinet shown below.


The cooled water returns to the Cold-Side PEX.
For this project, I would have prefered to work with PEX because it conducts heat out better than vinyl, but I can not bend the darn stuff.
Thermal properties are better with PEX, for this project. But vinyl works too; so easy to bend and work with. Barb connectors are bulletproof.

This lower cabinet holds paper towels and toilet paper which do not flex the water lines. You can not ever have heavy things in a bouncing RV moving against PEX lines. Vinyl can tolerate a lot of flexing, but not PEX.


I turned on the HotWaterTank at 9:45. The water is at 55F degrees.
I brought the 6 gal HotWaterTank up to 99.3 degrees, and turned it off. It continued up to 102 degrees.



The time is 9:57.
That is 6 gal of water (22.7kg) in 12 minutes for a rise of 47F degrees, Delta 47F = Delta 26.1C.
At a Specific Heat of water of 4.2kJ/kg/K, that is 4.2kJ/kg/K * 22.7kg * 26.1C = 2488 kJ = 2400BTU
Suburban-SW6DE HotWaterTank is a 12,000 BTU/Hr. 12 minutes is 2400BTU. (Perfect agreement with above!)
I do not know if that was just luck, or if there is some validity to science.

I turned on the Circulating Pump to start to bring the tank temperature back down, and transport the heat further inside of the RV.
I felt of the Hot Line under the sink, and in my hand it turned warm in about 30 seconds.
I do not need a flow meter to tell the warm water has hit my hand. I then turned on the HotWaterTap, and it turned warm in 1 second. Almost instantly.
Wow! For just this one feature, instant hot water, this is worth it. This is going to work great.



The HotSide is 86.8F degrees. It is going to the reserve filter canister, and then to the HotTap and Tee.


The Return colder side is 77.8F degrees.
Looks like, after the water current stabilizes, the 50 foot of vinyl will dump almost 10 degrees worth of heat.
After running the Circulator Pump for 23 minutes, the 6 gallon tank has lost aprox 10 degrees. Hopefully all of the heat has been transferred to the inside areas, where the heat retention and efficiency is much better.

And the inside room temperature has gained 2 degrees to 58.1F degrees.

Looks like this thing is going to work good. I am writing the code for the control electronics. I have two events:

I have a motion detector that turns onn the BlackTank PowerVent. This has been working great, creating a negative pressure in the BlackTank. Negative pressure is not related to what I am now doing. But the motion detector is.

I need to also connect the Circulator Pump to this motion detector. Actually, I do not need to write code for this function, just solder some wire. Hopefully, by the time I or my wife get off the pot, standup and get to the faucet it will already have some hot water. Washing of hands is optional, and now for the first time, warm water is an option too.

I need to write code to turn on the Circulator if the temperature on the floor or in critical areas reach, say, 38F degrees.
With, of course, the condition that the HotWaterTank is above 38 degrees. Otherwise I would be cooling down things, instead of heating up.





When you combine functions, such as a Manual-Onn, MotionDetector, and Microprocessor Control, then you can use diodes. Over on the left are three black diodes that combine functions. The diodes activate the small relay that activates the Circulating Pump for HotWater. The relay is an Optocoupled-Relay, and it only needs about 10mA; And it does not need big honking 10 amp Schottky PV Diodes, such as what I have used. However, this is what I grabbed in my parts box. Still works good. I constantly need instant gratifications these days, and I may have to grab the closest handy things to get my fix. But I got it; Works good!

The jumper on the relay board is set to the Right=H High 4.5v to 12v.

One diode comes from manual switch.

One diode from RV Microprocessor Module #3 output.
The Microprocessor may have several objectives. One is freeze prevention. Another might be to heat the floor and store energy by a wider and cooler distribution.

One diode from Bathroom BlackTank Vacuum Motion Detector Relay.
This is to prepare availability of hot water directly at the faucet.



There is a master switch on the RV Panel that disables or enables the Circ Pump regardless of diode inputs. It does this by controlling power to the RelayModule.


The new flow indicator arrived. Amazon makes it ridiculously easy to return. The first had a water leak in the center of the side, at the wheel bearing location. This red one is a little too sensitive, and turns too fast. It is a blur. Also a minus, it is smaller than my other indicator.
The other flow indicator had a temperature gauge, way cheaper, bigger. But the leaking part sorta cancels a million good things.


One disadvantage...
The MotionDetector also turns on the Powered BlackTank Vent, as indicated by the toilet paper being illuminated. And the reason that the paper is illuminated is to indicate that the BlackTank is under negative pressure, and that the suction fan is working. It is so quiet that you can not hear it. It is not even in the RV; It is up on the roof.

The acoustics in the bathroom have changed, and I can definitely hear something now. The HotWater Circulator now comes on with the BlackTank Vent. As you sit on the pot, you can hear a soft "whirl" somewhere in the bathroom. I guess the noise is low and quiet enough. It is OK, because I have to appreciate that the Circulator is preparing HotWater for me as a convenience. And if you leave the bathroom without washing your hands, no alarm will go off, so you really do not have to use the hot water if you do not want to.


All my devices talk to each other.

Speech

There are three Solar Charge Controllers.
This one is "Sol-1". It controls the charging of a Lithium 100 Watt Battery.


This one is "Sol-2". It controls the charging of a Lead Acid 100 Watt Battery.


This one is "Sol-3". It controls the charging of a Lithium 100 Watt Battery.



Circulator Pump turned off


The Solar Audio Monitor is playing in the background, and sounds like a lot of wind noise; and some of it is. But most is Solar Corona Discharge.

Here is an example, in this video, of the Speech module not knowing how to say a word.
If the module does not find a word in its library, it says the word "ahh" in its place.
You hear the attention getting sound, a Sonar Ping, characterist of the "RV System" Module which controls several water systems.
Next, it says the word "RV", another identifying sound of the source, which it gained indirectly from packet information.
Next is the "Clearing of the throat", indicating speech is about to begin. I got this from my Dad.
Next is a silent 400mS delay, which is automatically inserted.
Now for the main message...
The next word is "ahh", meaning there is a missing word. The missing word is "Circulator", which it can not find.
The remaining words are "water pump is turned off".
And the last sound is a "spit" sound indicating a period, or the end of the sentence.
I learned this last part from the great Victor Borge before I became a programmer.




There are a couple of functions of the RV Module that do not have anything to do with water:

Refrigerator Onn "The Refrigerator is onn."
The RV Module also keeps track of when, and how long, the refrigerator is onn.
And it does not matter if the Refrigerator is on AC power or Gas power.
But this information can only be seen on a computer or laptop.




Motion Detector, outside "I think someone is here."
This was from a motion detector.





Air Heaters Off and Circulating Water offAir Heaters Off and Circulating Water off
Here are two events that happened exactly at the same time.
And the speech buffer can memorize several events that occur at exactly the same time, to later be parsed out and spoken.
The inside of the trailer had been increasing in temperature, and the temperature was above 37 degrees.
The "Air Heaters", which are incandescent light bulbs, scattered throughout the trailer, are just now turning back off.
Also the "Circulating Water" has turned off. The Circulating Water is floor and cabinetry heat, coming from the HotWater Tank.
I am changing the 37 degrees to 35 degrees off. 37 is too warm.
Also the Speech Module knows all about the word "Circulator"; Its location in Chip2, and its duration.


At the present time, my emphasis of my work is on the Circulator. It can not heat up the whole RV, only the water lines and some cabinetry. And that is OK.

Also, I am changing some code. Now, instead of using hysteresis between the Turn-onn Temperature, and the Turn-off Temperature, I am using only ONE temperature. Normally, this would produce a chaotic situation of a device turning on and off too much. In fact, this can produce highly destructive "relay-chatter". After decades of using a Two-Temp Hysteresis code, a lower temp for onn, and another higher temp for off, I now see a merit to a One-Temp logic:
Freeze protection is so important that once a minute a vocalization is performed until the condition is resolved. I want the RV to jabber its head off about freeze protection, and the use of precious heat resources. With Two-Temp, the iteration occurs at the exact top of the Minute; Any important ongoing endeavors are reiterated at this time. Each and every minute, no matter if the temperature is in "no man's land" between the two temperatures.

With One-Temp, I do not have to do this. With heating resistors in roof water panels, with heating incandescent bulbs below, and with warm water flow in the roof water panels, and with warm water flow of the Circulating system, I like pulsed usage of 20 seconds, 30 seconds, or in some cases 45 seconds. And, by mandate, the status can not change during the time period. I have found that a pulsed operation can clear air bubbles producing a temporary AirLock in vane type pumps. And pulsed operation can help keep wires from getting too hot too, mitigating bad engineering.

After the mandatary Onn-period, the temperatures are remeasured and reevaluated.
And at times, there may be a 50% probability of change, of either an Off or an Onn. If the evaluation is unchanged, the condition continues, uninterrupted, for the same time period. The One-Temp is extremely responsive, hovering near ONE temperature. It is precise! For example, I could place the temperature precisely at 33F degrees, save resources, and have the maximum efficiency. However in a practical sense, this would be foolish. Temperatures are more spread out over a larger surface: Some my be 29F degrees, and some 35F degrees. And two temperatures would only exacerbate the problem. I have thus changed it to a One-Temp, and I will choose 35F degrees to be on the safe side.

I have changed it to a One-Temp, and have a vocal iteration any time it changes.


RVSystem: The HotWaterPump be off 70 degrees"RV: The HotWaterPump be off 70 degrees"

The RVSystem Module introduces itself with a Sonar Ping.
The 70 degrees is at the center of the Hot Water Panel on the roof. The Hot Water Panel heats the 60 gallon Fresh Water Tank. About 30 gal half full. I think the Fresh Water Tank was about 59 degrees. If the sun comes out, the Hot Water Pump normally comes onn and off all day, heating water. The pump comes onn when the Solar Water Panel is 10 degrees hotter than the Fresh Water Tank. It used to have a 20 degree hotter come-onn, and had less frequent come-onns. But I changed to 10 degrees hotter, and traded a hotter water for a cooler water with more volume of water flow with greater frequency. I am hoping the tank will get hotter faster.


RVSystem Freezing"RV: AirHeaters be onn freezing in here 35 degrees"

The RVSystem Module introduces itself with a Sonar Ping.
Air Heaters are many incandescent lamps scattered all over the RV, inside cabinets, and around waterpumps, waterlines, waterfilters, batteries etc.
They pull 6.2 amps. They seem to run several minutes before raising the temperature a degree.
The trailer is unattended.


Clock Official"Clock: Official DayTime now is 07:23."

The MasterClock introduces itself with a clock-chime.
The MasterClock keeps track of all Time Events. One of those events, is to know all the Sunup and Sundown times for the year for the Chico, Ca area.


Solar3I hear the sun may be ready for shining.
The RVSystem Module introduces itself with a Sonar-Ping. The Solar Modules introduce themselves with Geese-Calling. Here the RVSystem Module and one of the Solar Modules are hearing the MasterClock traffic, and commenting that they are ready for the announced sun to come up. Which is kinda ridiculous because they know when the actual solar panels are producing, and it is by their own measurements. Solar Controllers measure Solar voltage-output themselves. They do not need to be told. There is no "anticipation". I could make it so, but there is no need.

In fact, when we are over on the coast in the RV at Fort Bragg in the morning, fog may obscure the SunUp. A solar controller, even if they are told when the sun is coming up, can not see it themselves, nor do they have a concern. Actually, they can see it if the morning skylight exceeds about 1% of a Sol even in the fog. Still, they have no way to identify it as a "SunUp". The SunUp Event, as an "event", has no meaning for my Solar Controllers. It is stupid. Even so; kinda cute too at the same time, to hear non associated modules talk about it. And I am sure to wake up with more Modules gibber jabbing about it.


The "Event" is a left over from my broadcast days when a broadcaster was required to know the time of official sunup and official sundown. However, at my broadcast stations, Modules such as the Tower Monitoring Module, needed that information from the MasterClock to know how much current from Side Lights and how much current from Beacons was appropriate for the time. Tower Lights and subsequent currents were controled by tower photo sensors, but were checked by my systems. By the middle of the 1990s KHSL-TV and KNVN-TV knew exactly when so much as a light bulb was out. My system could even tell if the Blink Rate was off, which is another requirement of the FCC. My system was very advanced for its time. So now in an RV, modules do not know what to do with the information. But I know what to do with that information; I wake up and go to the window and look at a brand new sunrise, or look at a beautiful sunset. Maybe step outside and take a picture. I taught my modules to talk, and now they are teaching me.

Maximum Power Point MPP

The Maximum Power Point occurs when the amp-volt product, or power, is at a maximum. The problem is with Chemistry; the amp-volt product of LiFe batteries can not be just any combination of amps and volts. The Amps portion can be anything, generally the higher the better.

But the Volts can only be one value, or very narrow range: the value of the Lithium ion transfer.
Generally, this value is 3.40, 3.41, 3.42 volts/cell, depending on Charging Current.
Higher voltages of 3.45 to 3.50 can only be used for short periods of time without degradation and damage.
I listed three charge voltages, in a narrow band, for the Chemistry of LiFePo.:


3.40v Termination v: the bottom of the band. 100% Capacity is always possible with a large enough Absorption time. For example 5 amps, 1/20C for 100AH, and a large, many hours, Absorption time. This marks the Chemistry Threshold voltage. The time period is determined by when the cell is full, when the cell has reached a full AH capacity charge. The longer the time period, the more efficient the charging.

3.41v Termination v: 100% Capacity with shorter Absorption times because of the increased voltage giving increased Current, for example 10 amps charging current.

3.42v Termination v: I do not know for sure, but this begins the point where some heat is produced in the battery due to "Forced" Bulk and compromised Absorption. For example: 20 amps gives a 0.02 volt internal resistance to the Passive Absorption point. Or 0.4 watts/cell, small but measurable, if I am right.

3.44v Termination v: 30 or 40 amps will give this voltage. Absorption times have went ridiculously low. For example 100% Charge with zero Absorption time. And the cell is getting warm. And I do not know, but coulomb counting may show more going in than being stored. Personally, I am staying away from this cell voltage, 13.8 volts for my batteries. The only time I could see this justified, is early in the morning with dead batteries, and a forecast for heavy clouds later in the afternoon. Then you should take it while you have it. But what microprocessor can forecast the weather?

And lower voltages can not be used at all. For example: 3.30 volts/cell will not charge effectively.

The delima: The proportion of Amps and Volts in the IR product, is not ideal at the battery charge-point with solar panels. A 12VoltPanel has a maximum voltage of 21 volts and under full sun a MPP of about 17 or 18 volts. The same Power can be expressed by many combinations of Amps and Volts having the same product. Ideal is maximum Amps and a minimum voltage at the charge-point, which has the same IV product. For example, the MPP may have a power of 17 volts times 13.6 amps. But what is desired is 13.6 volts times 17 amps at the battery.

This curve is very important because of the difficulty of charging Lithium while the batteries are under load. My microprocessors use this (discharge) curve for charging.

You must get over 3.40v/cell before the charging voltage is effective, or worthwhile. And below, or at, 3.3v - forget it! You are not charging the battery. What is not shown is the "Residual Absorbtion" time. 3.40 volts will charge a cell to 100%, just as completely as 3.50v or 3.55v. It just takes more Absorbtion Time to reach 100% capacity. In my case, after I reach 3.40v/cell, a Full Charge will be reached in about 2 or 3 hours additional time. And the cell will still show 3.40 volts. I choose 3.40 v/cell as a "Constant Charge Voltage" which is 13.60 volts in a four-cell LiFe battery. The nomenclature is the same as "Float" in LeadAcid, but has no physical corollary in LiFe. But "Absorption" does! And adds a connotation of "Time" to the Constant Charge Voltage. As long as an Absorption time of a Constant Charge Voltage is less than a solar day, I am fully Topped Off. And I did it simply. CCV is not only simple, but also, I can easily see the state of charge. At full absorption, the current quickly drops off to zero. No need to "hit hard", for hopefully a short time, with a toxic voltage.

From 3.4v to 3.3v is a very narrow window. And if the batteries are "attached" or integrated into the system at the same time, charging methods become even more in dought.

Years ago when I first started with LeadAcid, I would charge to just below gassing, and then stop charging for 1 minute, then reevaluate. The problem is that you introduce, what I call, "Microcycling". And it can reduce the Capacity Life of any battery. MicroCycling becomes a factor especially with Parasitic Loads, or when batteries are permanently attached in the system, and there is a draw when not charging. At night the batteries naturally come down a little, but the depth of discharge is minor. Although the cycling is minor, it is unavoidable.

MicroCycling is not the only problem: BMSs discharge the high cells with only 100mA equalization discharge. For example, if the cells are operating at an amp, either charging or discharging, then 100mA will have negligible influence. BMSs will not work because there will be no Top or Bottom Balancing. The BMS can only work while the battery is quiet, and not while the battery is working with high currents, in or out.

Another thing I might mention: I have, what I call my "PostRegulators", and they have been running lately at around 12.1volts for the entire coach. They are responsible for the overall Coach voltage, regardless of any particular battery voltage. This makes it possible for every battery to have its own charge regiment and charge voltage, and to have its own life, independent of the requirements of the RV. Every battery has its own dedicated Post-regulator.


Here is the Curve above developed into four or five bands, or areas, in my code. Lithium batteries have their own BMS management systems, and my Code is in addition to their own internal systems.

My motto as a control engineer: Redundancy and backups.


And another motto:
Lets throw in the woofers, and test at the ocean!




MPP DumpLoads

MPP DumpLoads

Other people use a buck-boost to trade high or low voltage for current, in maintaining the MPP. And efficiencies above 90% are only possible with very low currents. Less than an amp. Otherwise, 80% is the norm. And with a voltage span from 17 volts down to 13.6v, the conversion is not worth it.

And when in the fog and rain, or in moonlight, a boost has no power to work with - to start with. It is not worth it.

I have a better way...
I will divert waste power, maintain and use the raw MPP, and use consumption to lower the voltage.


I will start with this...
For example:
I want to heat the 6 gal HotWaterTank from PV Solar. I want 99% efficiency of Solar Power, and I want to be able to use more of the Solar Power.

I want to use 100% of all Solar Power in all - repeat ALL - light conditions. Normally, this can not be done. At least, to my knowledge, it has not been done yet. At times the Solar Input may be at or below Battery Charge Voltage. Normally this takes a high power buckboost. Good luck in getting any efficiency out of this.
I have a totally different way. Hope it works...

There are two types of DumpLoads.
And both give no regard innately to MPP Maximum Power Point.
PostDump: After all higher priority users have consumed some power, a PostDump will assign the power scraps.
PreDump: A Source voltage ladder parses Dumps according to the Active Source Voltage available. The higher the voltage, the more dumps available. The lower the voltage, as in fog or clouds, the highest priority loads are exclusively allowed. The decisions are made upstream of all Charge Controllers.

In the picture, I show some of my Charge Controllers. But they can be modified to be my new PreDumps. The boards contain a PWM Regulator with an adjustable voltage pot. I have already manufactured the printed circuit boards. An adjustable voltage setting allows these boards to be PreDumps.

A ChargeController uses the Remote Battery Voltage as its controlling reference. But the important point is that it is an Output.
In contrast, a PreDump must use the Input. A PreDump must monitor the Source Voltage, which is its Input. A PreDump must know its assignment on the Ladder.

The ChargeController in the picture is a HighSide regulator, and the Logic dictates that a drop in the battery voltage causes a greater conduction in the regulator, resulting in a greater correction voltage at the battery. The correction is in the opposite direction.
In contrast, the PreDump uses a different Correction Logic: A lower voltage sensed on the Input must decrease conduction, resulting in a dropout of the Load. The correction is in the same direction.
Fortunately, this TL495 chip can do both jobs at the same time. It has two comparators. I have configured one for Upstream and one for downstream control.

I am just thinking about this. I could change to a 24 volt system...
There must be more HeadRoom. There must be at least 12 volts between the MPP Maximum Power Power point and the Battery Charge Voltage. In a 12 volt system, the MPP with strong Solar, is 17 or 18 volts. The Lithium Iron Phosphate voltage is 13.6. That is only 4 or 5 volts difference! A 12 volt system gives no room for a ladder. And it is worse than that: It is not the Beginning Charge Voltage of the battery, 13.6v; It is the Finial Charge Voltage of 13.9v or 14.0v. That is only 3 volts! And it is worse than that: It is 0v or less in dim light.

It is easy to change the panels...
I am going up on the roof and take two 100W 22v OC panels, and connect in series. This should give an equivalent panel of 42 volts Open and 37v MPP.

And I will set PreDump for the HotWater Tank to 24v, and the Cabinet Fans to 26 volts, and the cosmetic accent lights to 29 volts. The RV accent lights would be the first to drop out. Cabinet Circulation Fans second. HotWater Tank third. Batteries have the highest priority and never drop out by choice, unless temperature, overcharge etc. But Batteries share loads with PostDumps which is another discussion.

Here is the discussion:
I am working with resistive loads at high voltage, represented at the band of voltages from VOC to Battery Charge-point. Here, there is no need to trade voltage for current. Here, the MPP is not relevant. Here, power is power in any form.

Batteries are a totally different matter. Batteries need coulombs or electrons - not voltage - coming in for charging. Batteries need quantity, as in coulombs, they need electrons, for battery chemistry. Extra volts are only heat and waisted power. The MPP power should be available for charging; It is desirable and it is unfair not to have it at the charge-point of the battery. However, it usually is not the case. The proportion of Amps and Volts in the IR product, is not ideal at the battery charge-point. The same Power can be expressed by many combinations of Amps and Volts having the same product. Ideal is maximum Amps and a minimum voltage at the charge-point, which has the same IV product. For example, the MPP may have a power of 17 volts times 13.6 amps. But what is desired is 13.6 volts times 17 amps at the battery.

But if Charging is not the only goal, then resistive Dump loads open the door to nearly 100% efficiency. The dump loads bring down the voltage to the battery Charge-point with absolutely no waisted heat. At least, no waisted heat that was not planned for. If the battery was the only kid in town then yes, I suppose it was robbed of potential extra current. Then again, maybe not. Because my Buck-Boosts will be more effective at higher voltages into the same inductors. My experiment just might work.

There are hazards to my scheme:
In the Combiner, some schottkys may be 20TQ035s. Breakdown is 35 volts. They could blow when 42 volt panels backfeed into a 21 volt panel. This will occur if the 21 volt panel is 7 volts or less. The 35v schottkys would normally be OK unless a 21 volt panel became shadowed; That be a huge problem.
No way around it, they need to be STPS20M60D, 20Amps at 60volts. Gotta check it.

In Charge controllers, the power MOSFETS may be 4PO3L04, 80POs 80 Amps at 30 volts. If the batteries are 12 volts, they may have a two volt margin. They will blow for sure if nature is involved. But perhaps they are IRF4905s, 74 Amps at 55 volts. In that case, they would be OK. Gotta check it.

Charge Controllers must be modified to accept 40 volt input representations on their ADC inputs instead of 20 volts. The PIC 18F2523s have protection diodes on the inputs, but who wants to chance it. If this works, the Code has to be changed anyway, as well as resistor dividers.

I could pile a huge pile of zeners together as a safety protection. Never to be engaged. Just protection in case all of the Dump loads failed for some reason. But that would be a huge amount of zener watts: Each panel is about 3 ohms. Two in series would be 6 ohms. I would like to operate at 30 volts. That figures to be an 18 ohm zener load. A 30 volt zener pile at 18 ohms is v^2/18 is 50 watts. That is a big bunch of hot. I would need a big stack of 10 parallel 5 watt zeners. If each stack is a 24 volt with a 6 volt equalling 30 volts, then that is 20 5 watt zeners. Kinda comforting to know they would never be used. But still, kinda stupid just to look at.


Here is what I am working with on a sunny winter day. Solar flux is about 53%. All batteries are charged. Some lights are onn as well as a scope and soldering station.
At 53% SolarFlux and a cold 55F degrees on the panels, MPP point is about 17 volts. I have 19 volts, which is in between the MPP and the VOC of 21 volts. I will build a PreDumper with Tracking, to load the panels to 17 or 18 volts. I will use first the heating of the 6 gal hot water tank. This is a good choice for second or third priority. And it serves a useful purpose which I would want anyway.
Here I am placing 4 100 watt resistors against the tank. They are first mounted to aluminum bars, which are forced under the insulation with pressure. Contact surface is not the best: two contact bands per bar. They are two ohms each. 8 ohms in series, about 40 watts. About 2 amps extra on the Ampmeter. FET heatsink is ice-cold in all conditions: shutoff, regulation, full-onn duty cycle. Great, I do not have to worry about that. And I can place the MPP point any where I want from 19 volts down to 15 or 16 volts. The manual adjustment works great. The problem is that the 40 watts only brings down the Combiner voltage 0.2 volts. Hardly any effect at all. From 18.9 volts to 18.7 volts. I need another row of resistors. 80 Watts, twice 40 Watts, might put a better dent in the voltage. I can place an arbitrary MPP point anywhere, and have it feebly attempt control. But it is not enough to select an MPP point; The Load Devices must have the power to actually put the MPP there with enough assertion and current capability.



Another problem is that two of the resistors are too hot. The resistors are not heat-sinked to the rough steel of the tank well enough. R1 is 172.4 degrees, R2 is 143.6 degrees, R3 is 152.0 degrees, R4 is 184.1 degrees. Actually, they are all too hot. They will definitely burn my fingers. Rough steel on a curved surface. A challenge.
Here is the next day, despite the fact that I am not done. The 60 gal fresh water tank is at 55.3F degrees on the left in red. The 6 gal tank is in blue on the right at 66.7F degrees. All day today there has been no sun at all, not even the hint of shadows. Solar flux is 25%. Yesterday the 6 gal tank was less than the 60 gal tank. That is a switch. And the 6 gal tank was at 48.7 degrees. So the resistors are working. That is an 18F degree rise. But it will take more effort on the next 18 degrees: There will be more and more heat loss. In fact, eventually there will be no longer a heat rise.

I will take a look and see if this is even reasonable:
I have 6 gallons of water and/in a steel case.
Heat Capacity of water is 4182 Joules per kg per degree C.
The steel is 445 Joules per kg per degree C.
20 lbs of steel is 9kg, or 4005 Joules per degree C.
6 gal of water is 22.7kg, or 94.9 kJ per degree C.
Together is 94.9 + 4, or 98.9 kJ per degree C.
The temperature came up 18F degrees.
Centigrade is 5/9F, or 10C degrees.
So, 98.9 kJ/C x 10C = 989 kJ.
That is how much extra energy is in the tank.


How much did I put in?
I am putting in 40 watts minus a huge amount for thermal losses.
Thermal losses are burnt fingers and warm air above the tank.
Call it 30 watts...
30 watts is 30 Joules/sec
That is a rate of 108,000 Joules/hour.


Time:
989kJ / 108k/Hr = 9 Hours.
Yes! It is reasonable.


What would be nice as a final objective is to be able to shower at the end of a day. At the end of a 7 hour sunny day, to increase water temp from 50F degrees to 100F degrees.

If I can have a tight bond on the resistors, then I could have 40 watts. That would bring down the 9 hours to 7 hours.
If I double the resistors, then the time is down to 3.5 hours for 18 degrees.
I need 50F degrees. That is 9.7 hours. 10 hours of sun is hard to get in one day. But the attempt is feasible.

Suburbans have a glass liner. Will concentrated heat on the skin crack the glass? Water exposure, slipping through cracks in the ceramic, and forming a rust spot on a steel tank may be fatal.

The thing is not even built yet, and I have things to overcome. Another, is what to do if the water temperature goes too high. Too high, for me, is anything over 110F degrees. For example the system could turn onn the Circulator. That would be useful. Hot water would be moved out of the tank and into the flooring and other places.

Here is a 10 minute run the next day with the circulator onn.
The HotWaterTank starts out at 75.3 degrees
The input lines to the floor system come up to this temperature. It reads 76.1 degrees, a degree higher than the tank sensor.
The HotWaterTank drops temperature due to cold water returning from floor tubing.
In 10 minutes it reads 68.7 degrees.
The technique is effective if the floor is cold. But after an hour, the cooling effect on the HotWaterTank would be minimal. The floor lines and the tank would reach a steady state condition of everything closer to the same temperature.

I could cool the tank down by having the Circulator come on at 105 degrees. And a total electrical shut off of the resistors at 110 degrees. But that is a drastic measure and should never happen. Showering would be the same as what I am used to. At present, I never need to mix in any cold water. I only turn on the Hot, and adjust the flow for a comfortable temperature-feel on the skin. Low flow rates compensate for too hot of water on the skin, especially light sprays mixing with the air. In an RV, used for boondocking, you do not have the luxury of adjusting temperature with hot and cold knobs. Water is at a premium. And time waisted screwing around with knobs is not only stupid, it is not necessary. Only a tankless water system can do what my system does. And, in addition, you would have to have a tankless water heater at every faucet.


The above placement of resistors was parallel with with the circumference. I guess it could be described as "longitudinal". The resistors were mounted on a bent peace of 1 inch aluminum of a few inches long. This gave two points of contact, 1 inch each, for a total of 2 inches.


The new design places 3 resistors on an 11 inch bar placed laterally. This gives one long continuous line of contact down the middle of the bar. This gives 3.7 inches of contact per resistor. And should be an improvement.


End-on view.
The Hot Water Tank is about 11.5 inches in diameter, and about 11 inches long, front to back, from seam to seam.
The Bars are 1 inch wide and 1/8 inches thick aluminum.
Because of the roundness of the tank, the edges of the bars do not touch the tank.



The bar will not sit flat on a round tank. No dought there is an equation to find the distance from the outer edges of the bar from the tank. But it is more convenient to just draw it in CAD. I thickened the lines to make them more visible, and it is cosmetic, and does not effect the accuracy.
The gap at each side edge is 0.023 inches.




Actually, it should not be that hard to at least estimate the Separation Gap mathematically.
I will use part of the CAD drawing as a "scatch pad", and draw a second blue Tank. The smaller blue tank in the center is my doodling. And it is not to scale as the big blue CAD tank which is exactly to scale. But I will trade drawing-to-scale, for an algebraic solution. I sketched out an equation that duplicates what I have done above in the CAD program.
Ya, it is easy. Just a little Trig.
The hight of the bar above the tank, at its edges, will be the radius minus the cosine side of the triangle.
Everything is right: The Gap h will be a hair more than 0.0218 inches.
Therefore, verified. The Gap is indeed 0.023 inches from CAD.
(Not that I doubted it; Stupid aspergers I guess.)


I am wondering if I could place a wire of solder on the outer edge. This could potentially tripple the area of contact.
Hopefully I could have the two edges touch, and the centerline touch. Lead is kinda soft. I have all kinds of solder around here.
One roll of solder reads 1 mil thick, and I have no idea what that means. Another roll is 0.38mm. Another roll is 18 gauge. Another three rolls are 0.05 inch.

11 Now the whole world had one language and a common speech. 5 But the Lord came down to see the city and the tower the people were building. 6 The Lord said, "If as one people speaking the same language they have begun to do this, then nothing they plan to do will be impossible for them. 7 Come, let us go down and confuse their language so they will not understand each other."" 11 Genesis

WTF


If I can compress the solder to halve its thickness, then the 0.05 inch solder should work, giving the desired 0.023 inches.

On the other hand, a more ideal method would be to have an 11.5 inch grinding wheel and take out 0.023 inches in the center of the bar for the length of the bar. But I have never seen such a ridiculous grinding wheel. And besides, it would fly apart at any speed. The requirement for precision, and everywhere on the bar, makes this technique unreasonable.

Silver Paste has a conductivity of 3.17 W/mK. But lead has 34.7 W/mK. Way better, but less surface area.
But solder on the edge may be utilized to form a dam for the paste inside.
But how do I hold a thin delicate solder spacer trim in place?




2 ohm Resistors mounted. Two parallel sets of 4 resistors in series. Two 8 ohms in parallel. 4 ohms total.



Shaved off mounting screws, and mounted two strips of smashed solder on each bar at the edges.
Filled area on the bars between the solder strips with Silver Heat Sink Compound (not shown).



Epoxied two wooden 3/4 inch square dowels to the side of the 6 gallon HotWaterTank.
Used blue masking tape to keep the runny epoxy from invading the working area of the clean steel.



Laid the populated aluminum bars over the clean steel.
Applied pressure to all resistors from three vertical wooden dowels.
Pressure is carefully applied to all resistors with adjusting screws.
The screw points have not been dulled, so a penny was placed at each contact point.


Need soft springy washers to guarantee a steady consistent pressure, as the hard adjustment of the screws may be too critical.
There may be no pressure, or excessive pressure, depending on heating expansion and contraction.


Nothing is exact here. My holes are not precise and it looks bad. Using pitiful things like makeshift pennies. But the epoxy bond is super strong. I can not bust the wooden supports off even with a lot of force. I have to close my eyes to the look, and instead, tell myself it is a terrific Proof of Concept.





I have just turned on the MPP diverter that heats the HotWaterTank. The voltage of the Combiner is 20.3 volts, close to the Panels Open Circuit. The MPP pre-dump is running 80 watts, full out; And still can not bring down the voltage. At least not on it own. The current was 2.2 amps, now 6.2 amps. That is 4 amps. That is 80 watts. Oh well, that is OK. I will build several more.


Wow, this wonderful! I can just barely feel the resistors getting warm. Actually, they feel cool. With the other design they were running 180F degrees. And they would burn your fingers before you could jump back. The temperature measures only 70 degrees after a long time. The resistors are heat sinked to the steel and water very well. This is wonderful. Everything is working great. It took a lot of engineering, but worth it.


It is going to take about 11 amps to move the voltage down to the MPP. Here I have turned onn the 7.5 amp Furnace.


Individual meters only go up to 5 amps. And I know because with my design, each 100 watt panel has its own meter. A lot of people do not realize that 5 or 6 amps is all that you can get from a 100 watt panel. And that is at optimal conditions, and also at the MPP point. I never fuse my panels. It is stupid. That 5 amp current is all you are going to get, and you want every drop. In fact, I routinely short each panel out to ground to test their performance. They can take it. In fact, they like it. They run cooler while shorted out. I use the simplist of a little push button switch to short each out when testing. Never have I had a problem even with 5 amps hurting my switches. With good design and blocking diodes, you should never fuse your panels.


I have to figure a way to insulate the tank. I will set here cutting out different sizes of styrene. Even though the resistors do not get hot, I am keeping the styrene from touching.





I need to note the size of the filter: 12x12x1. It is not dirty yet.

The RV did not come with a filter. I had to put this one in.

It gets worse...
The RV did not come with an opening for air to inter the furnace. The Louvers on the left are, in effect, cosmetic. I was in shock to see what Springdale had done. What is behind the Louvers is a sheet of metal. It is the back side of the furnace; up close and personal. There is no room for air to get through. After I cut a proper air opening, and installed a filter, the furnace blew dust up out of the floor vents. Damn! All this time! The factory had installed the Louvers on the wrong side. They installed on the left instead of the right.


There are two switches under the Dinette. The top turns onn the 120 volts to heat the HotWaterTank.
The bottom switch turns onn this dedicated Solar MPP diverter. It also heats the HotWaterTank.

The Top switch is a three position switch, and used to also allow 12 volts into the 120 volt Heating Element. I had it that way for a couple of years, one position for 120 volt operation, one off, and one position for 12 volt heating of the element. I used the switch all the time: When Boondocking 12 volt operation, or Off for Forced Propane. And when in a paid park 120 volt operation. For short stays 120 volt heating is free. After two years of flipping this switch up and down, it finally arced closed in the 120 volt position. Normally, I already had this switch in the 120 volt position, before I rolled into a park. I never switched a heavy load while hooked up because the contacts could arc.

I immediatly disconnected the 12 volt system from the switch, and never ever will I hook it back. What can happen in an arced switch is the 120 volt system could short over to the 12 volt system, and destroy everything in the RV. I knew this at the beginning of the two years. I knew very well before I hooked it up. I have seen plenty of switches arc-closed. Some pure silver contacts. Stupid! Just plain stupid.


All day heating. I knew it would not make a hundred; only 92.3F degrees. I knew I could not get 10 sun-hours in a day. The calculations were right. Not quite enough heat or time. I can take a shower at 92, but it is cold. No thank you.

The next morning, the tank cooled from 92.3 to 75 degrees before starting to rise again. So the insulation needs improving too. If I start out higher, then I will get higher. Better insulation during the night equates directly to a higher temperature at the end of the day.

At present, I have two rows of four 2 Ohm resistors; or 4 ohms. 18v/4=4.5 Amps, which reads closer to 4 Amps on the meter. About 72 Watts.

I could make it two rows of three 2 ohm resistors; or 3 ohms. 18v/3=6 amps. 108 Watts.
I think that I might take this option, as I only have to move two wires.
I could make it three rows of three 2 ohm resistors; or 2 ohms. 18v/2=9 amps. 162 Watts.
A little more work: I have to drill holes and mount one more resistor.
That is plenty of watts for this amount of surface area.


Perhaps it is time to recheck my wire...
I have 15 feet of 14 gauge which is 2.53 ohms/1000, or 0.04 ohms. That is OK, but all my fuzes are getting a little warm. So, I am loosing more on fuzes than wire runs. But I guess that is how fuzes work: they get hot.


You can see the exposed resistors on the HotWaterTank.
I am going to look at the control module in greater detail on the scope.


Here I have adjusted the module to conduct almost maximum, or a full onn duty cycle.
The solar panels are reading 56% Sol, And their output from the Combiner reads 17.1 volts.
The release is straight and fast, but the power onn or pull up is slow and sloped, and not at all ideal. The pull-up should produce a little heat in the Fet due to not being a clean switch. After about 10 minutes, I am not sure that I can feel heat with my fingers on the heat sink. I think I can, but it is minor. And feels about the same as the heating of the 14 gauge wire. I have to use my imagination too much to feel it; It is fine. And the resistors themselves are as cold as ice. The current meter shows about 6 amps. But I think I can get more power by actually backing off on the onn-dutycycle adjustment. This shows less dutycycle on the scope, and I think, less current, but the voltage on the resistors goes up from 15 or 16 volts to 17 volts with a lesser onn-dutycycle adjustment. It is intuitive to see the voltage come up, but to see the power come up is subtle. The 17 volts at the resistors is actually 18 volts at the combiner and panels, and 18 volts is closer to the MPP than 17 volts. But it is hard to say because the MPP keeps changing. The Sol meter, subject to the changing sky conditions, keeps changing from a high of 56% down to 35%. The total current keeps changing dramatically from 9 or 10 amps down to 7 or 8, but with a lesser effect on the voltage. The voltage is one thing that is not bouncing around. It fluctuates very little: 17.0 volts to about 17.2 volts. Unless I bring it up with the module adjustment; Then there is a big change. With no conduction, voltage is about 18.5 volts, sometimes 19. Variation is also caused by other loads. The Combiner feeds three other Charge Controllers that have changing demands. Despite all the clouds and uncertainty, I know what I am expecting, and the facts are not disagreing with it.


The scope shows well below ground ringing. The negative excursion shows minus 5 volts. But it is not real. In fact, it is impossible. It would indicate the Flyback Schottky is not working. It is fine!
Wow! It is a problem with scope lead grounding.

The frequency of PWM Module is 5.77kHz. The target was 8kHz but the lower the freq the more efficient on switching.


Just by grounding the scope lead, the section of ringing that is below zero is eliminated. I think some of the ringing is actually real; and evidently there is some inductance of the wires getting to the resistors. Who knows, the resistors may be wire wound, and not purely resistive. Probably are. On a Release, the impedance is high, and the current momentum is high. Positive ringing is diffinity possible. The Schottky Diode prevents negative swings, and also gives a little current "boost" at the same time. Looks like it is working. If the ringing is real, it is not dangerous, and does not produce any high voltage swings at all. I will have to check and see if any ringing or switching noise is getting back upstream into the Combiner.

This can be very important for another reason: RF noise. Anytime I invent something new, I should check outside the RV with a Spectrum Analyser. FCC inspectors are nice guys but I do not want them knocking on my door. When I was a broadcast engineer, I was checking, almost everyday, for spurious signals.




The temperature of the HotWater Tank rises to 100F degrees. The the Circulator kicks on, as directed by the Microprocessor in the RVModule. Then the HotWater Tank temperature begins to drop. At about 98F degrees the Circulator is directed to stop. The temperature of the tank continues to drop to about 97F degrees. Then begins to climb again. Here is climbing again at 97.5F degrees, on its way to 100 degrees again, and repeat the whole cycle.


The circulating water is heating the floor and a lot of miscellaneous things. Here it is at 28.4c, or 84F degrees, as measured by the under-the-sink local preheater. The water at the bathroom sink faucet should be nearly 84 degrees if turned onn. The circulating water now only heats up the faucet water and the floor, but also cools down the the HotWater Tank. By circulating cooler water into the tank, the tank is protected from overheating.






The TL494 and the TL495 Chips can be configured for TWO simultaneous regulatory functions: The chip contains two comparators; Pins 1,2 and 15, 16. The chip is unique in that it can simultaneously regulate its input as well as its output.

On some PreDumps the output can not easily be regulated. For example on the 6 gal hot water tank, there is no filtering, no choke or cap, on the output. The output is 17 volt PWM spikes. Spikes are hard to regulate. In addition, the chip does not work well if the input frequency is the same as the output PWM frequency. I do not think I will use output regulation for this particular PreDump.

Spikes are OK in this application. It is pure resistive heating. A noisy square wave makes heat just as easily as a highly filtered DC. Just so there is no acoustic noise against the tank, and there is not. I do not hear anything.

Coach Regulator

But this is not the case with my next PreDump...
It is a Coach Regulator. The output is regulated at 12.10 volts, and helps serve the intire RV 12 volt system. Here the output voltage is critical and important.

Before this PreDump, the RV Coach voltage was basically from the batteries. And of coarse the batteries are religiously maintained constant by the SolarChargeControllers. Two independent Battery Coach Regulators keep the RV at 12.00v to 12.10v. The problem is that under heavy loads this voltage would sag. Sometimes by as much as a volt, despite the batteries being held constant; Constant, for example at 13.60 volts. This is due to the wiring losses from the batteries to the Coach Regulators. Another reason is the fact that the Regulators at times have only one volt to operate with: From a supply source of aprox 13 volts, down to the regulated 12.05 volts. That is not enough headroom. For most of the time the regulators are not even in regulation.

24: No one can serve two masters. Either you will hate the one and love the other, or you will be devoted to the one and despise the other. You cannot serve both God and Mammon. Matthew 6:24

To solve this...
Originally, years ago, I had a switch to either supply a Remote Sense from the Distribution Center and Fuses, or from the Batteries. A SolarChargeController could decide to either serve the batteries or the Coach. But it is a bad idea for Solar Charge Controllers to give more voltage to a battery to compensate for saging Coach voltage. This was the case with a RemoteSense wire connected to the coach fuse panel.

In fact, it is a real bad idea:
A lithium battery could go above safe limits despite a BMS.
The batteries would have a fluctuating charge voltage, giving rise to MicroCycling.

I decided then and there, many years ago, to not only have every battery have its own SolarChargeController, but for each Controller to be exclusively dedicated to the feeding and care of its own battery. That is all they care about; They do not care whats going on elsewhere in the RV.

In addition I invented my Liquidator Board, which contained a PostDumper. My Liquidator Boards still work today. At the same time that they give dedicated voltage to the batteries, they also divert power to other choices. With batteries as the top priority.

I need to think of a name for this next PreDump. It functions as a Coach Regulator, similar to the Coach Regulators that come from the batteries, except it is a tenth of a volt higher on output. Its input will be one of the lowest of all PreDumps, about 17.5 volts. And of coarse will have direct access to the Maximum Power Point at the Combiner, which is directly from the Solar Panels. PreDumps can design the MPP; PostDumps can not. They can only divert waisted power. I can not decide what to call it.


Here is my first prototype of a Coach Regulator functioning as a PreDumper. It has both Input Regulation and Output Regulation. Again, I used one of my Liquidator boards, and all I used was the PWM section with the FET and the Output Regulator. There are a lot of cosmetic things: Like where do I put the Coil? And instead of threading the heat sink screws into the aluminum, a nut would be better. Not because of strength, but because there can be a bigger hole to give more tolerance of placement.

I brought the current up through this regulator by turning on the furnace and a soldering gun. At over 15 amps the heat sink gets warm. Which is probably OK. What bothers me is that I may be out of regulation because the Coil normally gets warmer than the HeatSink. But that is not indicated by the output voltage: I set the Output to 12.20v and at heavy current the Output dropped to 12.09v. Which is OK, but that is not the whole story... One of the Battery Supply Coach Regulators came on, and was flickering. Meaning some of this current was supplied by a battery. No matter; I am close. The thing is working.

I am starting to think about the operation more seriously. I have a bigger issue with the concept. I have been classifying this module as a "Pre" Dumper, consistent with the batteries having the highest priority. But do they? Ever sense I got the RV, I have considered charging-the-batteries as the most important thing. Available current should go to the batteries first. Now, I have my doubts. When I first set up the module I set the Input Regulator section to 17 volts, and the Output Regulation to 12.20. And, even under moderate current, the input would drop out, or begin cutting back. This of course hampered the Output, and the Coach voltage fell. And this is great. This is how a PreDumper should work. But do I want this? After all, the Coach Voltage is pretty damn important. Perhaps there should be three catagories: The PreDumpers and PostDumpers, but also "Shared Loads". Which would be a type of opportunistic and chaotic sharing.

Or, better yet, multiple catagories, with Coach Power, the lowest input-voltage-catagory, the highest priority, unrestricted on input, no MPP considerations. This is logical because the Coach Voltage is way less than Battery Charge Voltage. 12 volts vs 12, 13, 14 volts. Under clouds or morning the MPP is not high enough for battery consideration. But it is for House Coach. As the light increases, the MPP progression will track the logical placement of priorities.







Yes, this should work:
As the light energy increases out of the Combiner, and after the voltage exceeds 12.00 volts, the Coach Regulator is engaged. The MPP is not considered. The Coach Reg has no input restrictions, and the Output is determined by Output Settings. I have more modules to build, but this will be the first module to come onn. The Coach Regulator is the first module to come on, and according to my new way of thinking, has the highest priority.

As light increases, and MPP increases, Battery charging begins as determined by Output Settings. As with Coach; There are no input considerations or constraints. Batteries are second in priority, and second only due to a natural higher input voltage; a natural consequence, by default.

As light increases further, up to about 16 volts, the PostDumpers begin to operate. PostDumpers begin to operate from about 14v to 16v, and above. Coach, Batteries, and Posts are fulfilled. Coach is running at rated voltage; Batteries are charged or charging; PostDumps are beginning to supply current.

Above 16 volts, the PreDumpers begin to operate, and each operate in voltage sequence. MPP becomes more important as sun reaches maximum. Indirectly, the sequence of PreDumpers lends itself to MPP Adjustment, and directly to MPP with its associated voltage sequence. With all PreDumps the input voltage is insisted upon and regulated. This is the MPP. Although modules engage in stages, - Individually - each PreDump operates in an Analog fashion, and the result collectively is an Analog operation. For example, at full sun, all devices feed from a regulated, and approximate, 17.5 volt MPP. Are my devices opportunistic or self-depriving? Stable or Unstable? I know they are Self-Depriving. And I have seen a little bit of flutter. But everything seems stable.

My system is superior to all other systems, at least in concept, in that the MPP is created with no wasted power. But there is an implementation hardship. It has to be implemented with catering to the individual RV. If PreDump loads are heavy enough, and the voltage is above 16 volts, the MPP is determined by an Analog System, and is very smooth. Hierarchial devices can not only operate at the MPP, but can also create it. Unique! Will this be Self Regulating, or will this be Unstable? Still worried about it...

In my system, all of this can be established, and tested, by the changing of one pot. The pot is located on the Coach Module. By turning Input Pot all the way CCW, there will be no Input Regulation. And everything should work as described. So, in the schematic above, the pot on the left will be turned all the way CCW. And the the operating point for input voltage should be 2.5 volts plus 15 volts, or 17.5 volts. That may be a little high. I may solder a 12 volt zener across the 15. That would make it operate at 14.5 volts. In any case, I have now made this function as the Highest Priority.


OK, to condense...
I have an hierarchial energy distribution system for an RV.
The rules of hierarchy are overlaid on the budget of available sunlight.


At the moment, the hierarchy is preset by static adjustments.
However, I can migrate that to dynamic intelligent control with my microprocessors;
Incorporating the "feel" of the RV, including history and changing needs, with presently available and anticipated energy. Here is the control input to the FET.

It runs between 4 volts to 17 volts. Sides straight and vertical.

Frequency is only 5.7kHz, instead of the 8kHz target, which is fine.
The lower the freq the less heat.


...So long as the lower-type freq does not show up in the output, and it does not. The lower the freq the harder to filter.


Here is the output into the choke.
The choke is 2.4mH, and 0.101 Ohms.
-2 volts to 16 volts.

Side straight and vertical on the Power Conduction Stroke.
At first there is some opposition posed by the Choke, that decays down from 17 volts to 16 volts.
Which is perfect.

Cutoff looks fine too.
Choke keeps the Release going a little, which is fine.
The reactance is robust because the slant is more linear.



Here is a quick look at the output of 12.17 volts.
There is no detail, but looks good:
No PWM noise.
I will later revisit all traces in detail under a higher current of 15 amps.

Trip to BlackButte

2023.04.21

At BlackButte Lake, I faced the back of the RV, with the big window, towards the lake. Taffy has never seen BlackButte. Obviously, Taffy wants to back in a little closer to the lake.


Taffy is on the left, and is a brand new puppy of 5 months and already weighs 8 lbs, the same as Callie.





This is the first year that I have noticed StorkBills being a problem in the fur of the Woofers.


Have this thing now parked at BlackButte Lake to test it out.
So far, looks good inside. You can see several of the solar panel current meters shining through the window.


When the Circulator comes on, it takes hot water out of the HotWaterTank and distributes it throughout the RV, and returns cooler water to the tank. This reduces the temperature of the tank. It also makes hot water immediatly available at the faucets. Which is a good thing. And, most importantly, do not have to reduce the temperature directly. I am still using, and taking full advantage of the full effects of the sun. I am still operating in the spirit of energy management where usage is maximum. Seems counter intuitive to call this "Conservation", but it is. Maximum energy usage, in this context, is Conservation and efficient Management of available resources.






The temperature control on the left is for when Heating is from Propane.

There is an identical temperature control thermostat when Heating is from AC Electric.

The third control, Solar, has no thermostat, and is controled by a microprocessor, and it turns on the Circulator. This third control is normally always on. There is a switch to turn it off, but I only use it for testing.

Here, I have turned off the Circulator, which would normally limit how high the temperature can go. Presently set at 100 degrees, but I will now change it to 110 degrees. After being at BlackButte for three hours with the Circulator disabled, and being shaded by an oak tree, the temperature has gone to 117F degrees. All on its own; Free energy! Wow! I am impressed. Can it really be this hot? I need to take a shower to see. I turned on the Circulator to bring the hot water to the faucets and shower in advance. Next, I took off my clothes, and jumped in the shower, much to the dismay of Linda. She could not figure what was happening to me. I gave her a quick explanation as I turned on the Water Diverter. The Diverter is another thing, and returns the hot water to the fresh water tank.
Normally, with the Diverter, I have to run the water for about 20 seconds before it gets hot. Not anymore! Wow, the water is already hot after only about 5 or 10 seconds. I know this beforehand by holding on to the brass fittings at the shower hot water tap.

I turned on only the hot water, and kept the water spray low, so as not to be too hot. I took a long shower, about 10 minutes. I used less than a gallon. Wow, it works. Better than I hoped for.





Some people go through life not living a "real" life. They kid themselves in a make believe world, distorted by stupid human biases. If you are one of these people, do not look at the next pictures...

Great! You read the disclaimer, and are still here.
Here is the BlackTank being dumped. The lever on the left is pulled out.
Waste is always colored brown. Contrary to intuition, and contrary to what people think, it is not lumpy. It looks like pure, and fine, muddy water. Solids are quite well broken down on their own; even with the RV just sitting. And a few miles of sloshing further guarantees it. It is nearly impossible for solids to stick to the bottom or sides of a tank because they do not exist. However, with enough ignorance and abuse, the impossible can happen. Leading to pyramids, buildup at the low point exit, and sticky walls.

Don't use much Chlorine in your Fresh Water Tank! It will get into your Blank Tank. Every time you flush the toilet, you will be using water from this source. As a result, your BlackTank will have particulates and have an odor. I have a water filter for the intire coach, that according to litmus paper, stops my small amount of chlorine from getting into the Black Tank. It is so small I can not even measure it. For the last three years I have not used any bleach or fresh water treatments at all. But, for me, there may be no need. I am the only RVer that I am aware of, that heats the fresh water tank in a continuous loop, cycling the water through a good water filter and exposing the water to hot temperatures in solar heating panel. Also, the hot copper tubing has some antibacterial properties. I probably, however, should go back to using some, little amount, of bleach because nothing is foolproof. I have been getting lax and lazy. I do not want to get sick.

Don't treat odor with chemicals. If chemicals invade the surface area of the inside of the tank and get into the plastic, the tank may be permanently damaged, and the tank may have to be replaced.

Use a biological enzyme. I use one, not so much for digestion, but to keep surfaces clean. I examine with a strong flashlight every few years, and it is always very clean.


We have no odor at all because our tank is under constant vacuum when we are in the bathroom. I am the only RVer, that I am aware of, that has this one too. If you do not have my invention, do not turn on your exhaust fan when you flush; You will suck odors into the bathroom. This action is exactly opposite to what most people plan to accomplish. Instead of exhausting bad odors, they are bringing them in. These people would rather pull an outhouse down the road, than to give their RV the love and care of an engineered habitat. My RV is a spaceship.


After you dump your Black Tank, open the Grey. The Grey will flush out the hoses. Massage the hose while the clear is running by raising low sections of hose to get out any pockets of sewage. Be careful, as the stream nears the end of its draining. Here, you may find things that float. Some floating debris and toilet paper. Don't close the gatevalve on any of it. If you have a rinse system, you can use it now.

13 You shall not murder. 14 You shall not commit adultery. 15 You shall not have a garbage disposal in an RV. 16 You shall not steal. 17 You shall not give false testimony against your neighbor. Exodus 20 New International Version

Most RVs do not have garbage disposals. And that is because RVs are not supposed to have garbage disposals.

Pour hot cooking oils into the garbage can, or well lined waste basket. Not in the sink!
Put cooking scraps in the trash can too.
I clean off my plates and silverware before I wash them.


If you violate these laws, you can try this: You can continually flush out the stale water in the P-trap. And continually add a small amount of water to the trap after trips. But really, there is no hope for you.

Don't use chemicals in the GreyTank, or BlackTank . They can atrophy, harden and destroy the seals in the GateValve.






Here is the temperature in the center of the copper coils in the SolarHotWaterCollector. The temperature is not 1.73, but rather 173F degrees. On some of my meters, voltage readings are literal. For example 5 volts reads "5.00". This reading is coming from an LM34 temperature sensor chip, where 1 volt translates to 100F degrees, and a literal "1.73" volts is 175 Degrees. No real conversion is needed beyond mentally moving the decimal spot.

The reason the temperature is so high, is due to the water flow being turned off. I turned off the water pump about 30 minutes ago because the tank was about empty. With the tank near empty, and water sloshing back and forth on the open road, the pump could suck air. Pumps running with no water can be damaged. It is better to disable, and wait for a water fill up spot. Or better yet; Don't let the tank get this low in the first place.

Several years ago, the system was disabled for 3 hours, and steam was produced. After the steam was produced, the temperature read 240F degrees. Super hot! Melted and burnt vinyl houses was the result. Things can get exciting in a homebuilt RV. My mother told me; If you play with hot things, you are going to get burned. Strange how a mother knows. When I was all grown up, sort of, I was in the hospital over a week with second degree burns all over my face. That was from racing cars with a overheated radiator. Another time, I burned off my eyebrowes and my eyelashes, as well as some hair. That was from washing carborator parts in the sink near a flame. Can't say I am a fast learner. Actually, kinda stupid.

Miss Mom. Miss her a lot. Wish I could take her to BlackButte lake.


Email to C.A.Pennock