Solar Pool Heater Sensor - Sunlover CSC-V3

[crunch]

It seems to be the month for me to try and repair electronic equipment. This is my second one I have asked about in as many weeks. (Where is the best place to post this question for the best help from people? )

I have a Sunlover Solar Solar Pool Heater controller; "Computer Solar Control Model CSC-V3" the unit was purchased in 1999 so its 21 years old.

When, hail came through our area the other week (Melbourne, Australia) the roof sensor thermister cable took a direct hit from a hail stone about 30mm from the sensor and was sheared right through. (Not surprising for a 20 year old UV damaged cable). The cable is too old and brittle and the break too close to the sensor for me to do soldering and repair the cable and have it be reliable, so I need to arrange a replacement sensor.

The company that originally made it, Sunlover Solar, appears to have been rebranded to Sunlover Heating. They have been contacted but have no interest in helping us out with information and want us to buy a new unit which can be upwards of $500. I suspect the only relation they have to the original Sunlover Solar is the name. They probably don't even have the IP (intellectual property) to the original CSC-V3.

I don't want to have to spend $500 simple For want of knowing the correct 5cent component we need to use. The unit itself is otherwise reliable and works efficiently and effectively for what we need it to do. (well, it did before the sensor cable was broken)

I purchased a new cable labeled "Sensor - Roof Analog Suit Ascon Controller TS11PP" made by Dontek Electronics that the pool shop I brought it from claimed to be the type of sensor people use on these Sunlover controllers when the need to replace the sensor. HOWEVER when I plugged it in I get a reading of 01 on the unit. I have checked the terminals and ensured the new cable is properly seated and tight and making contact with the unit (as best I can) When the roof sensor cable is removed the unit readout reads 99 for the roof sensor.

We have no circuit diagrams, no technical information, nothing except the old sensor the new sensor and a bunch of photos of the inside of the unit. (And the unit itself, of course)

I was able to test the original damaged sensor and appears that it reads 25.4Kohms at about 23degrees c and when I heat it up the resistance gets lower and cooling it down the resistance gets higher. This appears to make it an NTC thermister.

When I test the new Sensor it appears the cable is actually polarized.    In one direction I get 5.3Kohms at 25 degrees and in the other direction I get  141 K ohms.  at 25 degrees.  (I suspext it's not a standard NTC thermister cable like the one it replaces. .......  )

The internal Circuit board is labeled CSC-v4 although the external label is V3.

On the internal circuit board there is a micro POT labeled 'roof'. But I have no idea what it does.



Twisting it violently to both extreams with the new cable installed doesn't seem to make a difference (neither does swearing at it, calling it names or threatening it with a pair of side cutters). I'm assuming its meant to be used to adjust the calibration, but without a technical manual for the product (and somebody to interpret them to easier non technical English for me) I am just guessing. In the picture, the IC4, beside the pot is labeled MAL - LM358N 90F836, in case it's not clear in the photo.

I'm hoping there is somebody here who can provide more information, or at least enough information to get our unit back online without spending huge amounts more money then we already have.

David

[Ed.Kloonk]

I know you said the wires are short but can you repair the old sensor and use a bit of heat shrink?

[crunch]

I'm going to try if I have no other luck elsewhere.

I would love to know what those POTs do.   Amongst other things, now that I have foolishly played with one I have no idea what its original setting was or what i need to test to get it back to the original setting! (totally my fault).

[Ed.Kloonk]

I'm going to try if I have no other luck elsewhere.

I would love to know what those POTs do.   Amongst other things, now that I have foolishly played with one I have no idea what its original setting was or what i need to test to get it back to the original setting! (totally my fault).

From the pic, it should be easy enough for someone to draw a schematic.

[Ed.Kloonk]

Years ago I made solar water heater motor circuit using information I ripped of the net. The pots might be for a 'set' temp or perhaps duration. Or maybe the amount of hysteresis. Who knows, until you draw out the circuit.

Something like this shows how pots work around the lm358 op amp
https://www.homemade-circuits.com/solar-water-heater-with-battery-charger/

edit: I found the Silicone Chip schematic which I modified for my own purposes using the LM339 instead:



My knowledge of any specifics further is limited, sorry.

[peter-h]

In a previous life, many years ago, I was manufacturing differential temperature controllers for solar heating We sold many thousands...

We started using AD590 sensors - literally the Analog Devices ones. Accurate to 0.5C. We sold a few hundred and most of them failed. The reason turned out to be that the AD590 didn't anywhere near meet its operating temperature spec. We used the metal can TO-18 version, too... They drifted by several degrees, after a few months of being cycled to say +80C. The company sent its fattest salesman driving the biggest company car, to take me out for a £50 lunch (1979 - a lot of money!) but he failed. I sent back the whole stock. But our company reputation took a big knock.

Then we moved to thermistors from Fenwal. Epoxy dipped. Accurate to 0.2C. They were better but still drifted. Apparently the solder used to attach the wires (inside the epoxy bead) migrated into the thermistor, even at the low temps like +80C.

Finally we solved it with glass bead thermistors from Midwest. They looked like this

They were brazed at something like +500C and were really stable. Not so accurate, about 1C, but this was ok for the application. I still have a few in a bag, 30-40 years later.

Obviously we needed interchangeability of sensors, hence the need for absolute accuracy.

The circuit was just an op-amp switching a relay via a BC337. More expensive versions had a digital readout done with an Intersil 7107, which used a diode network to linearise the thermistor's output.

So, it was not trivial to get it right.

Nearly all the customers installing solar water heating were crooks, shysters and in some cases outright criminals, and the business became harder over time, dealing with these types.

One of the complications was to get a long term watertight seal past the cable entering the sensor, which was in a 8mm OD copper tube, epoxy filled. We used irradiated PTFE insulated cable. The irradiated material enabled the use of a primer which epoxy would form a hermetic seal to. I see this issue has still not been solved by most manufacturers of industrial sensors (e.g. pressure) who are relying on the sensors not being "too outdoors". Lots of them fail after a few years, due to moisture ingress.

[floobydust]

OP I think you're wrong about the sensor being a thermistor. It looks like a diode or diode-connected transistor, based on the circuit I see as well. The MCU can't do thermistor linearization (1KB code space 128 bytes RAM) so it is best to use a linear sensor (diode) with the CAL trimpots, which makes the firmware much easier.
Try test a sensor using a multimeter on diode-test instead of ohms. Your 114k ohm reading implies the sensor or cable has water inside, especially if the reading climbs and moves around. It might have gotten cracked in a few places. The cable might have to be shielded as well.

edit: The ROOF, POOL trimpots by op-amp IC4 LM358 are for calibrating the temperature sensors (offset I believe). A pic of the board's other side would reveal those secrets.
For some reason the power supply voltage regulator VREG LM317T is adjustable, the third trimpot is to adjust that. It would also scuttle the temp sensor calibration if you changed it. I imagine it is a 5.000V setting.