DIY thermal differential controller (solar hot water pump controller)
 

In anticipation of using a solar hot water system, I am working to design a DIY thermal differential controller. The main reason behind this is that it is a simple device that costs $130+ for a commercial controller. I know I can make one for less than half that price. Plus, if I put together good documentation others should be able to easily construct them as well.

So, what is a thermal differential controller anyway? A thermal differential controller is what is used to control the pump between the solar panel and hot water tank. It senses when the solar panel is hotter than the tank and starts up the pump so that the water can go get warmed up. When the panel cools off, the controller then turns the pump back off so it doesn't cool down the tank. Sounds pretty simple doesn't it? Well, it is and thats why I don't want to pay $130 for one!

The setup is pretty simple. You have a micro controller and two sensors. One sensor goes in the panel and ones goes in the tank. You also need a relay of some sort to switch your pump on and off.


For the micro controller, I am going to use an arduino. It is very popular micro controller and is more than capable of handling this small task. I'm also some what familiar with using it.

The sensors will be 10k thermistors. These are what are commonly used on the commercial controllers. This makes for quick and easy replacement should the user choose to go that way. The other option is making your own sensors which will save a fair bit of money. This is the route I will pursue.

I'm lazier than you. Wouldn't a solar powered pump be easier? If there is enough solar energy to run the pump then should also be enough to solar energy to heat your water. You would still have monitor the temperature of the tank to ensure it doesn't get too hot.

A PV powered pump will work okay. However, it will start up before the panel has warmed up and it will shut off after the panel has already cooled down. So, you will loose some efficiency.

With a normal thermal differential pump, you set it so the pump only comes on if the solar panel is 8 or more degrees hotter than the water tank. You also shut it off when the solar panel is only 4 degrees (sometimes more) hotter. These temperature differentials are set to compensate for thermal losses through the lines. The longer the line between the tank and collector the greater the temperature differential. It of course matters how much insulation is used around the lines as well. In the commercial controllers, these temperature differentials can be set to as much as 24 degrees.

I've been in contact with Gary over on BuildItSolar to get an idea of what features he would like to see and what is important in such a controller. Does anyone here have any input?

The easy way is two temperature sensors and a comparator with adjustable offset and hysteresis.

Quite right, NiHaoMike. However, that wouldn't leave much room for additional features. For instance, Gary commented that it would be nice to have an LCD display that shows a couple things. He also mentioned a max tank temperature is a nice feature or displaying the number of hours the pump has been in use. I'm not sure if I'll end up going that far with it, but I'd like to know that I have the option.

Then add all the fancy stuff like control outputs to start use of hot water on the spot (such as washing dishes), PWM pump control, thermostat for home heating that uses excess hot water production, maybe even Ethernet connectivity.

Differential Controller...
 

Daox,

Great idea on the controller.

There is a variant on the differential controller idea... it's a controller which is specifically for hydronically heated homes. I'm not so sure what the general name of the controller is, but the feature is that it monitors the temperature outside the house as well as inside the house. Hydronically heated houses have a very large thermal mass (AKA: concrete radiant floor). This thermostat can sense the temperature drop (or rise) before the inside sensor can and helps prevent undershoot and overshoot conditions.

I'm not so much of a circuit hacker, but I suspect that the 'adjustable offset and hysteresis' features could apply tho your project and with some tuning could apply to this type of thermostat as well.

Bst Regards,

-AC_Hacker

Mike, that is definitly doable. Perhaps, once the base model is done I/we can start developing add on modules for it for doing additional things. I was also thinking about integrating data logging (perhaps thats what you meant by ethernet though). I'm really no electronics expert though. Even adding a simple LCD screen will be a learning experience for me.

How would PWM control of the pump help though? As far as I can tell, the larger the temperature differential between the fluid flowing through the panel and the panel itself, the more efficient then panel will be. I'd imagine slowing down the flow would decrease efficiency.


That is a very interesing idea AC Hacker. I like the idea of a 'smart' hydronic controller. You'd definitely be able to get some extra energy savings that way. I'll have to think about using that as a possible add on.

Differential Controller Curcuit...
 

Daox,

Just to prime the thought-pump...

...have you Googled: app note op amp differential

...such as here:

app note op amp differential - Google Search

...or Googled: differential controller circuit

...such as here:

differential controller circuit - Google Search

Also, as far as the ethernet monitoring part of your project, Ward Cunningham, the guy who invented Wiki (the structure upon which Wikipedia is built) has offered to help me get the Homemade Heat Pump Manifesto project on line, with monitoring data publicly available in real time. Obviously, I will be sharing these development details with you as they come available to me.

Best Regards,

-AC_Hacker

Hi Tim,
Just learned today that the Goldline differential controller is no more. This was (to my knowledge) the cheapest controller out there at about $130. It did not have a lot of bells and whistles, but very solid. Surprised to see them decide to stop making them -- they had been around for many years.

On the direct PV power of the pump where the PV panel acts as the power and controller.

- Its difficult for some systems. For example, drain back systems where the collector is quite a lift above the tank need a pump with high starting head, and these tend to be AC pumps with fairly high power demand -- expensive to do with PV.

- Its nice to have a controller that knows what the tank temperature is. On days when the tank starts out cool, it can start the pump earlier in the day -- as soon as the collector can produce water hotter than the tank water. When the tank starts out fairly hot already, a controller that knows the tank temperature will wait until the collector is getting enough sun to actually heat the water in the tank. The PV panel has no knowledge of the tank temp.

- The efficiency of a PV panel and a thermal collector opposite dependencies on ambient temperature. So, on a very cold morning the the PV panel will make enough electricity to start the pump earlier than usual, but the water heating collector would ideally like to be started up later than usual, because it has higher losses and is less efficient.

All that said, a lot of people use the direct PV successfully, and it has a very appealing simplicity to it.

Gary

Oh YES!

I need a few of differential controllers around here . . . . .
Attic Fan . . . . ++ I think this would be a big "market" for these . . . .+++
Pool Solar heater . . .
Updated domestic water heater . . . . .

First thought Idino Know nothing about them but they are flashable, displayable, and output controllable for a decent price . . . . seems like overkill tough . . . .

Have also thought in the past about some sort of electric way to wire up temp sensors and relays to do the same thing; but I can't remember my idea now, and may have been dreaming . . . .

Why would you need a differential controller for an attic? Wouldn't a normal temperature switch do?

It has a normal temp switch; I'd like to upgrade.

Take this morning as an example; 70deg outdoor air temp, 95 deg attic temp (my roof is shaded in the afternoon, but not AM) I could have dropped attic temp down in to the mid 70's and taken some load off of the A/C. How ever if I had set the thermostat to 75 to take advantage of this situation the fan would turn on sometimes in march and shut off in early september.

Just another case of taking maximum advantage of available of available resources.

Sorry for crappy post; in a hurry . . . .

Ah, yeah that does make sense.

I had Solar Hot water for about 17 years, before the system started leaking a lot..
When we took it out, I kept the pumps (one is solid brass) and the controller.
I think it might have used 10k thermistors. It has pump control relays etc .
Anyways, if you are interested, I can dig up the controller and check it out.
Maybe we can do a deal.?. :)

Hey, just found it. It's a 1981 Novan C-30-1S, made by Independent Energy Inc.
The power LED still lights up, and I can hear the relay(s) working..

http://i46.photobucket.com/albums/f1...r/NCL/F014.jpg
http://i46.photobucket.com/albums/f1...r/NCL/F015.jpg

Got it open.. No insects! ;)
http://i46.photobucket.com/albums/f1...r/NCL/F018.jpg
Still looks okay inside. There is a temperature limit pot (blue).
The chip seems common.. RCA - CA324E

Just hook up the sensors, two motor leads into the sides and plug in the 120AC. :)
The relay contacts seem to be rated at 1/2 HP - 7.2 Amps at 120vac..

There is a slide switch on the side for OFF-AUTO-ON (on is for testing).
IIRC, I think you need two of the same type thermistors to make this work right.

This unit was mounted on the heat exchanger, next to the big tank
down in the basement. Never kept outdoors.


You never know, this thing might still work!!

I wonder what's the standby loss on that old controller.

For an extremely high energy efficiency, it's hard to beat some modern low power opamps or a MSP430 microprocessor. It takes a lot more effort to design with a MSP430, though.

The standby isn't going to be very much, when compared to those two cartridge pumps
running almost the whole time the sun is on the collectors.
The motor run-time might be lower, if you used a small storage tank, or kept
using hot water all day.
We had three full sized Novan collectors and used an 84 gallon tank.
In the winter, it would average about 120 deg F.
In the summer time, I think it was about 30 or 40 degrees warmer.

One could just use a Sun detector gizmo to power up the system every morning
when the sun started warming things up..

Edit:

Maybe even a timer? Have the AC switch off at night.

If I was going to use this again, I would use a Sun detector/Switch
It would control the AC to the hotwater system with a SS relay (Solid State)
that was being driven by a small PV panel.
At night, or when it was very over-cast, the system would go into shutdown.

Zero watts on standby! :)

Haha, nice old controller. I don't suppose you have any of the sensors? They do have to match the controller since every thermistor out there is a bit different.

I'm pretty sure they were just plain old 10k thermistors.
IIRC, the ones that came with the system were junk. After a few years,
I replaced them (Radio Shack) and epoxied potted the new thermistors
into some old brass .357 mag cases.
One was tucked into 82 gal tank and the other was inside the insulation
on the 'down' pipe (3/4" copper).

I may have one spare somewhere in my junk, but it would be easy
to test the controller using a couple of 10k pots..
Just turn the 'panel' pot up and down and watch the relay.

I have noticed that a local solar company uses bi-metallic thermal disk switches, on the order of this:


You couldn't get a differential function, but at 25 amps and 240VAC max, it could turn the whole system on & off, very reliably, without standby current, without the need for a relay.

Regards,

-AC_Hacker

The problem there is that all thermistors are slightly different. A 10k thermistor made by company X does not have the same resistance at 100 degrees as company Y's thermistor.

Perhaps they are close enough. I just haven't done that much research on it yet.

Just buy two of the same type and they would pretty close..
I think these RS models are 1%..
RadioShack Thermistor (271-0110)

I did a search for "Differential Controller Kits" and came up with this:


...in fact, he even gives you his circuit:


...and if that leaves too much to the imagination, he gives you the top side of a perfboard board prototype:


... and also the bottom:


... and here's another one with a nice box to enclose it:


Best Regards,

-AC_Hacker

Hey, I wonder if that LM324 chip is the same as the 324 chip in my antique 1981 controller?
Of course mine requires no soldering. ;) It's just plug-n-play!

Interesting, I haven't see the DIY kit. I have seen the ART-TEC unit. The cheaper one is fairly limited.

Looking at that diagram..

http://www.jc-solarhomes.com/DTT/DTtherm1.jpg

The thermistors form a simple (2.5V) voltage divider. So, you could use two
5k or two 10k etc. As long as they were nearly the same resistance
(when at the same temperature).
If they weren't the same, that Adjust Pot can trim things out. :thumbup:

I think my old controller could easily be tweaked in using a small pot
on one of the input lines. (If it was indeed out of balance).

I found this thread after pondering how to regulate the fan for my hot-air solar collector. Given that I'm trying to make it from my junk pile, I'm thinking of making a big bi-metallic strip by riveting brass and steel together, and using it to control a microswitch. The operating range can be set by distance along the strip, and the temperature by proximity to the strip, na? Any better ideas? Could I just use a refrigerator thermostat, and adjust the heck out of it? Or how about the thermostat from an electric heater with the on-off reversed?

How about using and old room thermostat? When we went to programmable
thermostats, we saved our old Honeywell units. I have one that does both AC & Heating
somewheres in my junk. That would be prefect for controlling a fan, if you have a relay
or and SS relay.

If you don't have one hanging around, I think Lowes has some in the $10 range.

How about a thermal disk switch?
 

How about a thermal disk switch?


They're fairly inexpensive, and highly reliable, and they can handle considerable current without additional electronics thingies.

Here's one that should be right in your range.

-AC_Hacker

remote thermistor?
 

I wonder if the thermistor could be removed from a modern Tstat
(like the RTH221B1000)

Honeywell at Lowe's: 1-Week Basic Programmable Thermostat

And re-connected using a twisted-pair?


If you could locate the thermistor inside the warm-air collector, the Tstat
could be used in the AC mode. Coming on whenever the temperature
inside the collector got above ~75-85 deg..
I think these things have about three degrees of Hysteresis..
So the fan was moving a lot of air, it would shut down pretty quickly,
as the collector cooled off a bit.

~~
In the early evening, you could try using the heating mode. If the outdoor temp
dropped below 70, air from the house would be sucked into the collector
and cooled down a little.. (Assuming the house was still warm).

Again, you are the MAN!
Anybody try this yet?
If nobody else does, I'll try one in Feb/March for the pool heater.

"They're fairly inexpensive, and highly reliable, and they can handle considerable current without additional electronics thingies.
Here's one that should be right in your range.
-AC_Hacker"

Thanks! "On at 40 C, off at 32 C." I thought I'd want to leave shutoff closer to my indoor temperature, but I see that their range only goes up from that model. I suppose that if the collector is active, it will usually kick on the fan, since it hits 70 C pretty easy in still air. Has anyone tried these limits, and watched a partly cloudy day get wasted?

doax mentioned an arduino in the first post, I like that idea, they can be had for as little as $15 for a clone. They have tons of support online so getting the code right should'nt be a problem, and once we do everyone can copy and paste it. Kinda like MPGuino project. They also sell cheap lcd displays for like $20 so we can use the serial print function to display temp and pump on/off status, ect. I bought one a few months back and have been real busy with things so I haven't got past the first few little tutorials(I can blink some leds like you wouldn't believe) but it is fun to play with and I see tons of uses for us.

Arduino Based Controllers
 

I was looking around for the same type thing, with a bit more... Brains then just an off/on differential controller. These two links are great resources. The Arduino board seems to be the way to go.... Tons of possibilities, very low cost.

Hearth.com Wood Stove and Pellet Stove Forums....and much more! | My New Control System

arduino (woodnotoil)

Another cpu controller that is very easy to program is the PicAxe, they have a mini-basic interpreter inside the chip. Plug them into a breadboard add a couple of resisters and a capacitor and you can connect to the serial port of your PC to program them. They have inbuilt basic commands to read the DS18b20 1-wire bus temp sensors and are absolutely ideal for making solar controllers. Here is a link to the UK site that manufactures them, I believe there is someone in the USA importing them.
PICAXE

Cheers
Mike

I built a differential controller that has 4 temp inputs and temperature logging that is pretty cheap ($49). I wasn't going to pay $150 for it either. Here's the link if you are interested (just take out the spaces).

http://sites.google.com/site/mydtcstore/

Looks interesting...

What is the current draw?

In what format is the data stored?

-AC_Hacker

Here is some from a screen capture. The line under the s command upto the relay states and relay on time is what is stores off each tick.

I just noticed the picture I posted is for a depopulated version that just does the temp logging. You can get one with one or two independent differential or HEAT/COOL thermostat functions (selectable).

Current draw is about 15-20ma. When the relays are active, add about 33ma more each.

>m
myDTC Deluxe v1.2
s)Status c)Config t)Set LogTick f)Freeze Mode
0)Set SI Units 1)Mode_R1 2)Mode_R2 3)SP_R1
4)SP_R2 5)Bump Minutes 6)Bump Hours
7)DeltaT 8)LowT 9)HighT m)Menu r)Rules
>s
DAY:HH:MM TICK T1 T2 T3 T4 R1 R2 R1TIME R2TIME
000:01:12 012 59 59 58 58 OFF OFF 000.0 000.0
>c
DTMP LTMP HTMP MODE1 MODE2 LOGTICK SP_R1 SP_R2 FRZMODE
04 F 05 F 60 F DTC_R1 DTC_R2 5 MIN 25 F 25 F ALLOFF