How can I use resisters to replace a Thermister?

[RonRock]

I have a pool heater that I have to replace the electronic control board often. I have now replaced the board 5 times. Each board shows the same issue when it needs replaced. I have tested all of the control switches all test good. The thermister is one item that is not basically an open or closed switch. As I'm sure everyone here knows the resistance value of the thermister changes with temperature change.

The one I need help with has three wires, ST-COM-SR. The manual shows to test Resistance based on the water temp. So Lets say that the water is at 60*F. The resistance between ST -COM is 15.31(K), resistance between SR-COM is also 15.31(K) within 10% of each other is good.

My thermister tests good with my VOM. But I still think that it may be bad. It is a fairly expensive part to just guess needs replacement. So I would like to make a "test" tool with two resisters. Seems simple enough to do use one side of the resisters together for COM then the other side for ST and SR.

But I am not savey enough with electronics to know if this will work or not to"trick the board into "thinking" that it is a working thermister. I also don't know how to choose a resister to give me the correct value of 15.31.

I will replace either the board or thermister after my test. But at this point it is hard to belive that the boards keep going bad with the same symptoms.

Any advice? TIA
Ron

 

[Nigel Goodwin]

If replacing the control board cures it, then obviously the control board is faulty.

What does the thermistor read between ST-SR, that's the rather crucial part you've left out - presumably (as the other two both read the same) they 'might' be connected together, so should read zero ohms (or very low). If it reads 30.62, then there's possibly two thermistors?.

But I don't see how replacing the thermistor with a resistor will help, as the entire system then has no control - either it won't heat at all, or it will heat on full power permanently, it's the thermistor that makes it work.

How long do the boards last?, and can you post a picture of one (both sides, top and bottom of the PCB.

 

[Lightium]

Thermistors are chemically different from resistors, so no.

 

[alec_t]

From your measurements I would guess that the 'thermistor' is perfectly ok and the fault lies elsewhere.
General purpose thermistors are cheap. Get a couple of NTC ones rated at 15kOhms and connect them in series, the join being COM.

 

[RonRock]

Thanks guys for the replies. I have taken photos of the board. I will post as soon as I can figure out how to get them from my phone to my computer. Always a PITA for me. Seems like it should be simple.

I will also take a reading between ST-SR. But unsure what it will read, the water temp is not likely 60*F as that temp was just an example I gave from the diagnostic chart in the service manual.

 

[Reloadron]

Looking at your numbers a 10K standard thermistor will have a resistance of 15.310 K Ohms at 60 F. Leads me to believe you have a dual 10K thermistor which is doing just fine. I would try as Alec suggested in post #4.

I see where you only posted an example. I would still run with the suggestion by Alec.

Ron

 

[RonRock]

This is one of the boards that I have replaced. Part of what makes me question the thermister is that, if I unplug the connector while the unit is powered on I can hear the relay click and the gas valve click open and close quickly. In the past I have been able to get the unit working if I tried this eventually it would work and the heater would work as normal. Now I am unable to get that "trick" to work. It still acts the same, but will not make the system work. Almost seems as a bad connection at the thermister plug. But that is where I have taken my readings and they are always within spec. I can provide any info needed on the board. There are some numbers that I assume are part numbers, but don't help me.

 

[RonRock]

Sorry about the backside photo. I should have orented it as the frontside is. The thermister connection in the second photo is the three solder joints at the bottom of the photo.

Is this correct thermister as suggested in post 4?

NTCLE100E3153JB0 Vishay Beyschlag/Draloric/BC Components | Sensors, Transducers | DigiKey

Order today, ships today. NTCLE100E3153JB0 – NTC Thermistor 15k Bead from Vishay Beyschlag/Draloric/BC Components. Pricing and Availability on millions of electronic components from Digi-Key Electronics.

www.digikey.com

 

[rjenkinsgb]

Part of what makes me question the thermister is that, if I unplug the connector while the unit is powered on I can hear the relay click and the gas valve click open and close quickly. In the past I have been able to get the unit working if I tried this eventually it would work and the heater would work as normal. Now I am unable to get that "trick" to work.

Could it be the relay contacts that are burning out over time? If the loads they switch are not properly suppressed, the relays may not last long.

They would still click, but not always turn on the connected load.

 

[alec_t]

Is this correct thermister as suggested in post 4?

That might work (for testing at least), but from Reloadron's post #6 it looks as though 10k would be a better choice.
I'm wondering why two thermistors are in the sensor? It occurs to me that they might be used in a bridge circuit; but in that case I would expect one to be an NTC type and the other a PTC type, for enhanced sensitivity. Can you check this possibility by warming the sensor up and seeing if the resistances from COM to SR and COM to ST are still approx the same?
Is the sensor connector polarised or reversible?

 

[tepalia02]

Thermistor's resistance is changed in accordance with the temperature. It's not like a fixed resistor. Replacing the first one with the second will most likely cause the whole board to dysfunction.

 

[rjenkinsgb]

I'd try a 10K resistor in series with a 10K potentiometer (wired as a variable resistor; wiper joined to anticlockwise end).

That would give you a range of 10K to 20K so you could see what points the heater switches on and off, compared to the commanded temperature.

You would need two of those setups to simulate both thermistors.

Are the actual ones both in the same housing, or are they separate and monitoring different parts of the heater unit or system??

 

[RonRock]

The existing thermister is a single unit.

I may not have explained what I am trying to accomplish with the test thermister idea. The board has a potentiameter that adjusts the temp setpoint. I have checked and the resistance does vary with the turning of the knob on the pot. I am pretty sure that the thermister is simply what measures the water temp and when the board sees the resistance from the thermister it determins if the temp is at setpoint of the pot. What I want to accomplish is to have the board "think" that the temp is below setpoint and turn on the heater. And since the resistance won't change the heater will stay on.

The board actually has two control sides. One Pot is Pool, the second is Spa. I get the same responce out of the board on either side Pool or Spa. When working properly it works on either side.

Sorry about the spelling errors. For some reason my spell Check isn't working.

 

[Reloadron]

I'm wondering why two thermistors are in the sensor?

You and I both.

Thermistor's resistance is changed in accordance with the temperature. It's not like a fixed resistor. Replacing the first one with the second will most likely cause the whole board to dysfunction.

Yes and when calibrating thermistor circuits The measurement and display are normally checked against a standard resistance.

<EDIT> I wrote a lengthy post and all of a sudden the server was down so I lost it. Well that sucks! </EDIT>

The one I need help with has three wires, ST-COM-SR. The manual shows to test Resistance based on the water temp. So Lets say that the water is at 60*F. The resistance between ST -COM is 15.31(K), resistance between SR-COM is also 15.31(K) within 10% of each other is good.

I have never seen a three wire configured thermistor. Not to say such a design is not out there. Three wire RTD configuration yes. The example you mention above the numbers would match up with a 10K thermistor. I am not sure what the designations ST and SR are about?

Ron

 

[RonRock]

Thanks guys for the help.

The pool temp is 68*F as measured by the pool thermometer.

My readings are SR-COM=12.73K ohms
ST-COM= 11.87K ohms

Seems within the 10% differential. But I am not sure how to test the rest of the board. Still seems odd that I have had several boards go bad in the same way.

The unit was unpowered for these photos. The Pool-Spa switch is normally left in "Pool" mode. But as previously mentioned don't matter to the working of the heater.

 

[Reloadron]

Well for 68 F, a 10K thermistor should have a resistance of 12,491 ohms or 12.491 K ohms. Your numbers are pretty much inline with what I would expect. Keep in mind thermistors are not quite linear animals. Figure 67 F would be 12,809 ohms and 69 F would be 12,182 ohms. For applications like you have thermistors offer an inexpensive sensor solution for process control.

When several boards all fail in the same manner it sort of points to a design flaw. They do mention a 5% differential maximum. Very important is make sure the thermistor connections to the board are clean and secure. There is no reason to have to replace control boards over and over again.

Ron

 

[alec_t]

The board actually has two control sides. One Pot is Pool, the second is Spa.

That probably explains the 2-thermistor package. But if one is for Pool and the other for Spa, how can they be in the same physical location?
I note the package has two red wires. That suggests the two thermistors are matched and interchangeable in theory.
I wondered if the designations 'ST' and 'SR' had any significance in some foreign language, but Google Translate doesn't throw up anything obvious.

 

[rjenkinsgb]

I'd guess it is a dual sensor system for safety, to prevent the heating running away as it could if a single sensor failed.

If the two sensor readings do not match pretty closely, it will shut down rather than heat to an unpredictable level.

>google<
Like this?

WO2012103356A2 - Dual thermistor redundant temperature sensor - Google Patents

A redundant temperature measurement probe is described that may include any number of features. In one embodiment, the probe includes a pair of resistive sensors, a pair of output wires, and a shared ground wire for a total of three wires. The first and second resistive sensors of the probe can...

patents.google.com

(It does not really seem to be a patentable concept, as two or more sensors for redundancy & error checking is a very common concept - look at eg. most Diesel car throttle pedal units.)

Two precision thermistors with one connection common to both should work just the same, or two equal resistors for a system setpoint test.

 

[alec_t]

That makes more sense. Our previous car had a dual-track pot on the throttle pedal for safety.

Each board shows the same issue when it needs replaced.

Does it display an error message pointing to the sensor as the culprit?

 

[RonRock]

Yes the Sensor Fail Led that can be seen in the photo. It is mentioned in the diagnostic flow chart. That can also be seen in the photo. Of course the suggested "fix" is replace the board.

I have 5 of them. I'm tempted to start removing parts and testing them. Sadly I don't know how to test several of the parts. Most seem pretty straight forward to test with a little Google help. But I don't know about the SIP's. No idea what they are.

Any suggestions how I would do that in a logical manner?