Hydronic Heat Pump NTC Thermistor - Incorrect readings

[aqhc]

Hi guys, I’m a complete newbie to this forum so please be patient/kind!
I run a hydronic heating firm in Frankston South, Victoria. I’m a technician/installer by trade.

I’ve got a client who’s got a Tivok Hydronic Heat pump and I’ve fully serviced and cleaned it, with the last issue being the faulty thermistor, which I believe is a generic type, as I’ve purchased a couple, all with the same false temperature readings.

Please see the photos attached. The Tivok heat pump (HP) heats up a 200L buffer tank (BT), and does so via an NTC Sensor in the dry well, about 1/2 the way up. The original one worked fine, up until last year. System is about 7-8 years old. There is a digital controller on the front of the HP and it reads, among other things: The Set and Actual temperatures.

Set temp. would be 45 Degrees C. The actual temp will be cold and so the heat pump SHOULD turn off once the 45 Deg. C is reached, based on the NTC Sensor detecting the correct temperature in the dry well.

I’ve cut and joined the wires from a brand new NTC sensor and crimped/joined them to the existing wiring connector that goes on to the PCB. There was a fault code coming up that said “P3” which meant that the PCB was not sensing. With a new NTC Sensor this fault code goes away but the reading on the controller reads -9 Degrees C.!!

What happens is that the heat pump thinks there’s a demand for heat so it fires up and runs up to 65 Degrees and then shuts off with fault “E4” which is an O/T Fault. It then continues to fire up when the temperature drops to, say 58-60 degrees.

I’ve checked all the connections, used 4 different makes of NTC Sensors from 2 x very reputable hydronic heating suppliers (Hunt Heating and Australian Hydronic Supplies) and they all show the same readings.

What does happen, however, is that the -9 Deg. C does, sometimes goes up, and the maximum temperature it read was 20 Degrees C. But, the true temperature inside the Buffer Tank was 65 at the time.

So, it’s not working and I’m running out of ideas! How could 4 x different 10 MegaOhm NTC Sensors be wrong?? The “P3” Code goes out but then it doesn’t sense correctly. I’m sure I’m missing something simple? It was working fine for years!

Could it be the PCB possibly? Or, do I need to join the NTC Sensor wiring in a different way?

Normally I just ring the manufacturers, speak to the Tech Team but Tivok no longer exist and I can’t seem to find anyone that can help, hence thinking that I could find some help/advice here.

I’ve attached photos and the manual also.

Any help appreciated!!

Thanks,

Nick Harris

[Craig_2048]

aqhc

Thank you for contacting DigiKey , From the manual I found on line the E4 code is High pressure protection/Pressure valve break down (system 1) as far as the temp not reading correct I am not sure why that would have changed when you put the new thermistor in .
I am not versed in heat pumps and have not worked on one myself , from the pictures all the components look ok on the board I do not see any caps bulged or any components burnt .
I will send an email out to the engineers to see if any of them have some ideas as to why this is not working correct .

Thanks Craig

[rick_1976]

NTC temperature sensors change resistance in the opposite direction of temperature, so a low temperature reading would correlate to an abnormally high resistance reading or open circuit.

10MΩ would seem a rather high resistance for such an application; if one replaced a 10KΩ thermistor with a 10MΩ thermistor, low readings would be expected. One might see something similar if a lead wire was broken, there was a bad connection on the board, in a connector somewhere, or some other condition existed which was causing the circuit to become open or high-resistance for some reason.

One might consider measuring continuity of the wiring to check for opens/breaks, and perhaps substituting a dummy resistance/potentiometer connected directly to the PCB to evaluate board function and/or determine what an appropriate thermistor value might be.

[aqhc]

Many thanks for your reply Craig. Yes, the “E4” code means O/T in plumbing/engineering terms - OverTemp - The unit has stopped short of overheating. So this is not good/normal, and the NTC Sensors I’ve connected up don’t seem, at this stage, to be sensing correctly, although they are eliminating the “P3” code which is detecting NO sensing at all. I’ve just seen Rick’s post below so will answer that. Thanks.

[aqhc]

Thanks Rick, that all makes a lot of sense. I’ll double check the thermistor that I’ve bought as the ones supplied from my 2 x very reputable suppliers were their “Generic” brands that have the correct KΩ or MΩ.

So, it could be a connection issue or something on the board? I’ve attached photos of the board connections and the original connector.

When I replaced the sensor I actually cut and spliced the 4 wires - 2 each as a pair: I didn’t crimp them. But they come as 2 bare wires so, for testing purposes, I just spliced them - Could that be the issue?

I removed the original connector in the photo but left the one connecting to the board, with about 300mm worth of wire, as I didn’t have a connector on me to go direct to the board - But I can source one and crimp it onto the sensor if that would help? Or, have I got the wrong Ohm rated sensor in the first place?

Thanks again!

NH

[aqhc]

Here’s the manual also…FYI.

Thanks

thp60ab.pdf (5.2 MB)

[rick_1976]

@aqhc

So, it could be a connection issue or something on the board?

That’s a possibility. Again, a low temperature reading from an NTC thermistor loop points to an abnormally high resistance, which could occur anywhere in the circuit–in the cabling, in a connection, at a solder joint, etc.

Looking a bit more closely at things, it would appear that most of the control magic happens in the unit with the display. The sensor signal obviously has to get there in order to be displayed, which suggests some additional places to go looking for bad connections etc.

My suggestion at this point might be to look at that display/control unit to see if a person can’t figure out where the sensor feed comes in. If one can find that and the sensor measures that same looking back through all the funny business in between as it does directly at the sensor, that’d point to an issue with the controller itself. If not, well then the issue would seem to be somewhere in between.

[aqhc]

Many thanks for your reply Rick. I’ll look into this and feed back to you, probably next week now that we have a long weekend in Australia due to ANZAC Day. Cheers