Hello, I have a problem with a cylinder head temperature guage that I bought from a company called Auber. The problem I'm having is I'm getting noise in my reading while the engine is running and the only cause I can find may be the sensor grounding out on the engine where it is installed underneath a spark plug. The gauge reads fine before the engine is started, then when it's started it goes crazy and is all over the place.
I'm wondering if there is a way to create an isolated ground to power the gauge with therefore eliminating the noise from the vehicles ground system? I'd like to do this without adding a second battery which brings up questions of how to charge it and still maintain isolation while the engine is running.
This is all being done on a 1970 beetle so I don't want to overly complicate the electrical system. The solution may be to buy a different gauge that can handle the sensor being grounded.
Ive looked at stabilizers, isolators for dual batteries, and dc/dc convertors but I'm having a hard time figuring out how these would fit into my solution.
The guage only uses 2W so my solution doesn't need to be heavy duty.

 

The guage only uses 2W so my solution doesn't need to be heavy duty.

Link to the guage?

 

Sure
https://www.auberins.com/index.php?main_page=product_info&products_id=504

 

Link to the sensor?

 

i would try determining where the noise comes from - power circuit or sensor circuit.
to test power, temporarily power it from separate 12VDC (battery or PSU) to see if problem goes away. if so then you can use small isolated DC/DC converter to power it up.

to test sensor circuit, check what type the sensor is and how it is wired.
some are isolated, some are not, which is why the manual tells that the gauge itself need to be powered from isolated source. chances are this is the issue.

using twisted pair shielded cable can dramatically reduce noise. if needed one can connect capacitor in parallel with the sensor. that slows down response but temperature is a slow process so not an issue...

 

beware thatmany isolated DCDC converters have limited input voltage range:
https://www.digikey.ca/en/products/detail/traco-power/TMH-1212S/9383360

in car 12VDC fluctuates, so you would want one with wide input range like R2M-1212S
https://www.mouser.ca/datasheet/2/468/R2M-3073919.pdf

 

I think that the TS told us the sensor clamps under a spark plug. That is about the noisiest possible place in a car. Does the sensor have two wires or only one? That is an air cooled engine and so you may be able to add a second wire and isolate the sensor from the engine block and attach it to a cooling fin near the spark plug.
Please letus know if the sensor has one wire or two.

 

It's a K type thermocouple. Under the spark plug is the common place for it and the best place to get the closest reading. The sensor is from the sensor connection and is a reputable company.
https://thesensorconnection.com/pro...er-head-temperature-sensor-ring-terminal-type
I haven't had a chance to test powering it with a second battery yet. Busy painting the house unfortunately. Hopefully, this weekend. The sensor is supposed to be potted to avoid grounding at the transition.

 

So you’re trying to measure a 4mV signal in the vicinity of a sparking plug with a couple of kilovolts on it. Are you surprised it doesn’t work? Thermocouples were available in the days of the VW Beetle, but guess what? Cars with temperature gauges used thermistors to measure the temperature, which had a huge change of resistance with temperature, and ran at a few tens of milliamps, driving a hot wire gauge with a time constant of about a minute. Which is the better design?

 

So you’re trying to measure a 4mV signal in the vicinity of a sparking plug with a couple of kilovolts on it. Are you surprised it doesn’t work? Thermocouples were available in the days of the VW Beetle, but guess what? Cars with temperature gauges used thermistors to measure the temperature, which had a huge change of resistance with temperature, and ran at a few tens of milliamps, driving a hot wire gauge with a time constant of about a minute. Which is the better design?

Wow. You obviously don't know much about old VWs. I'm just gonna assume you woke up on the wrong side of the bed today.
Most people with any kind of high performance work done on their type 1 are running CHTs under their #3 spark plug. Some have decided to epoxy them to the head near the plug, but only because it is an absolute PITA to change a plug with one of these on there. There is NO potential between the spark plug body and the head. When assembled they are the same.

 

It's a K type thermocouple. Under the spark plug is the common place for it and the best place to get the closest reading. The sensor is from the sensor connection and is a reputable company.
https://thesensorconnection.com/pro...er-head-temperature-sensor-ring-terminal-type
I haven't had a chance to test powering it with a second battery yet. Busy painting the house unfortunately. Hopefully, this weekend. The sensor is supposed to be potted to avoid grounding at the transition.

My brother had one on a two cycle motor cycle under the plug it worked fine. Perhaps yours is defective. I like Mr Bill's idea. Test it by clamping it between a fin and a plate of steel.

 

While the steel base of a spark plug screwed into an aluminum head is supposed to be well grounded, it does not take a lot of resistance to develop a bit of voltage. AND if it is some electronic circuit rather than a low impedance meter, then a millivolt of noise can lead to a problem. And if it is a thermocouple welded to a thin washer then it could easily be damaged.
A separate welded thermocouple coated with ceramic goo could be attached to the head near the plug and held in place with high temperature adhesive.

 

Wow. You obviously don't know much about old VWs. I'm just gonna assume you woke up on the wrong side of the bed today.

looks like you did... what Ian0 said was spot on. every wire is an antenna. if it is only 10mm long. and wireless transmission has long been proven.

the mV signal from sensor is amplified. and along with it any interference. and it does not take much to induce noise in a mV signal. the thw circuits do not need to be touching to get the interference. it is enough for wires to be close enough. and simply having both inside a vehicle is close enough. we all have seen clunkers whoes engine operation can be picked up using AM radio placed 10m away. so having sensor mere inches from spark plugs is a valid concern.

what you failed to do is:
a) tell if sensor mount is isolated from sensor wires. some thermocouples are not galvanically isolated. datasheet does not state it so did you check this? with sensor leads free (not connected to anything), is there a continuity between attachment ring and the wires?
b) if the noisy reading is due to ground loops - did you ever try running it with gauge powered from a separate (galvanically isolated) source? documentation for the gauge states that supply need to be isolated.
c) are the connections solid? are you sure this is not simply caused by vibrations from running engine?

 

In industry, your colleagues would never thank you for making a simple problem difficult.
It certainly is possible to find the 4mV signal from a thermocouple, remove enough interference so that it doesn't overload the amplifier input, filter the amplifier output to remove any remaining interference, then measure the cold junction temperature (often with a thermistor) then subtract that figure from the output of the amplifier to give the hot junction temperature; but when the alternative is a thermistor which would give a 6V output with no amplification and a capacitor across it which would remove the interference, I don't quite see any advantages for the thermocouple.
My classic car experience is British and water cooled - Morris minors, minis, Hillman Imps and Sunbeam Rapiers, so I don't quite know how hot a VW air-cooled engine fin runs at, but the oil will have to be below 125°C, and you can get thermistors with M3 screw threads for use at 150°C.

 

What about a ignition interference suppressor, would that work

 

What about a ignition interference suppressor, would that work

Never heard of that one. I was thinking the first thing i might try is a cap across the sensor output.
Do they make 1:1 dc transformers?

 

Never heard of that one. I was thinking the first thing i might try is a cap across the sensor output.
Do they make 1:1 dc transformers?

No such thing as a DC-DC transformer

 

No such thing as a DC-DC transformer

You are right. Superman and Batman are DC, but Transformers are Marvel.

 

I was thinking the first thing i might try is a cap across the sensor output.

You could try a large capacitor (e.g. 1000uf or more) across the leads to the display and see if that helps.

 

I am GUESSING that the TC is welded to the washer that clamps under the spark plug. THAT means that it is solidly connected to the engine block. IF the engine block is not solidly enough grounded to the lectrical system common there can be many volts of noise present on the TC connections. IF that is the case, moving the ground connection of the temperature display module to the engine block should reduce the noise. But so will adequately grounding the engine block to the vehicle power common negative.