Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Amplified thermocouple issues

Not open for further replies.


New Member
Hi, I'm a newbie to this board and hope someone here can help me out. I've trying to troubleshoot an erratic reading on a cylinder head temp gauge for automotive use. It uses a K type thermocouple (grounded), a 'black box' amplifier and an analog gauge. The operating range is 0F-600F, and the amp ranges 0-5 [vdc] output. Here's the problem, with everything wired, the amp wired directly to the car battery and the TC screwed to the head I get a fairly accurate reading,,, provided the ignition is OFF. As in the TC setup is hardwired to the battery. As soon as I turn the key to the run position (engine not running but car powered up) the gauge shows ready about 30% high. Start the engine and it just bounces all over the place.
There's a few more scenarios I haven't had time to try yet like floating TC ground though it's not possible in this situation, but you get the idea. Just trying to ID the problem and go from there.
I've done a little research on this topic, but I'm still a rookie so if I have something really wrong let me know. My first thought was potential difference at grounding points (ground loop?) so I tied the grounds together and still no luck. Then thought Alternator noise, tried a genereic Radioshack filter kit for Auto use, still no change. The only way I can seem to get this to work is with a separate power source, but that's not really practical unless someone knows of an easy way to achieve this. Maybe a line conditioner on the power lead to the amp? Anyone have any ideas?
Thanks much, --Ryan
Not much I can say, the whole setup is sold by Autometer, but so far they've been no help. TC leads connect directly to it, then there's 12VDC in, ground, and three leads out to the gauge. I would just crack the box open, but then I void warranty on $130 setup should something go wrong. Assuming of course it goes right first. Also should mention that my regulator puts out 14.4VDC, possible problem? Though I would think that would be accounted for considering it's normal to have more than 12VDC in cars.
Is there a way to mimic a separate power and ground source for the amp w/o actually going through the trouble and expense of a true separate power source? Or is there something more I'm missing. Thanks again, --Ryan
Earth loop ?

My guess is an earth loop, the thermocouple (t/c) signal is only a few microvolts and easily swamped by 'noise' signals.
If the 'black box' doesn't have a true earth-free input and use balanced line circuitry ...

First isolate the t/c from ground, try that (the norm in industry).
Is the t/c cable screened? do it (earthed only at the amplifier!).
Does this fix the problem ? ...

For your idea of an independant power supply, yes, if the above doesn't work.
Small (size of a matchbox) switched-mode units are available, using a small capacitor to 'hop' power without a common connection, I've definitely seen 5v to 5v versions so I guess 12v to 12v will exist.
A little more info. Did a quick check with my multimeter tonight, not only did I find a ground loop to be the cause, but the bouncing needle corresponds to the 'bouncing' potential I measured. Actually I found that the chassis ground at the TC measures about 7mV relative to Batt (-). With the problem side taken care of now, any ideas for solutions. As I mentioned before, due to the circumstances the TC must be grounded. Is an isolated power supply the only solution, or are there other less involved ways. If I have to go that route, anybody know of a specific model they can recommend? Thanks, --Ryan
Hi rcb78,

i read your posts carefully.
The only way I can seem to get this to work is with a
separate power source

Bit ambiguous this.
I take it to mean that you can get it to work properly
if you use a separate power source.
I assume you mean a power supply that is not connected
to the engine in any way, other than the connections to
the thermocouple ?

If that is the case, and that is what you mean, then my
opinion is that the thermocouple is touching to case in
its fitting somewhere.

Which would not affect its operation on a separate supply,
but which would interfere when the engines system is in
use for this.

The answer is to remove the thermocouple assembly or
housing, and inspect it carefully, and ensure that it is
properly isolated from ground, and wont vibrate itself
into making contact.

I suspect that the thermocouple has one side connected
to ground, at the fitting or housing, meaning only one
wire comes out to the loom from it.
This might make it difficult to check for leakages as
thermocouples are quite a low impedance, i suggest you
disconnect the grounded side from the fitting or casing,
so as the thermocouple is 'floating' while you check the
insulation, and wiggle the insulation, and if necessary
improve the insulation until you are satisfied with it,
then re-connect the grounded side.

If there is a slight drop in the battery line which is
interfering with the thermocouple signal, you could
possibly run the ground line of the thermocouple out
separately, so you have two wires to the thermocouple,
and ground it direct to the battery,
i suspect that may be impractical.

Or you could run the battery main earth strap, the
major conductor, over to the engine block instead, and
use that point for a strap to bodywork.
All such connections, bat neg, body, engine, sometimes
extra strap to starter, are likely to give small volt
The thing is to arrange them so that any volt drop(s)
wont interfere with the thermocouple ground being the
same as the amplifier ground, and maybe the instrument

Have a look at the layout of the ground wires, try to
see them as poor connections, and think how best to
arrange them.

Best of luck with it,
John :)
Ok, let's see. The TC is itself grounded. It's a washer style with the TC welded to the washer, this is the norm for CHT devices on automotive engines. What isn't the norm is the amplifier, most simply drive the gauge with voltage from the TC. However I had reasons to go with this particular gauge and just make it work.
Now then, like I said, TC is grounded, both wires, no way around it. When you attach the TC to the head, it's grounded there, no options. If I measure potential from the head to the batt (-) I find that relative to the batt (-), the head is (+) .007vdc, but only with igintion on. Likely culprits are stereo amp and ignition controller. With the engine running, the reading goes nuts. It bounces from .005-.009 vdc from head to batt (-). Oh, and the gauge itself recieves it's ground from the amplifier, nothing to bother with there. If I ground the amplifier next to the TC, it still reads incorrectly, but differently, sifnificantly low. Lower than ambient temps with engine running, however the gauges readings are stable (not bouncing) with the engine running.
Tomorrow I'm gonna try powering the amp with a separted car batt and see what the results are. Before I meant I got correct readings off the TC with a MM on it's own power supply, not the amp on it's own power supply. But the amp does work correctly wired directly to the batt with everything else turned off.
I'm a little concerned about the isolated power supply idea. I understand that the (+) side will be isolated, but will it just pass ground through, or will that be isolated as well? Thanks, sorry if I seem a little incohesive when writing this. That's just how I get when I'm brain storming, too many ideas to get out in short order. --Ryan
Thermocouple woes

You say that the thermocouple (t/c) must be earthed but I don't see why...
Can't you use a couple of mica or PTFE washers to isolate it ? maybe the insulator kit for a TO220 transistor ?
I appreciate that this will slow the thermocouple's response but given the size of the lump of metal you are monitoring ...

To give an earth-free power supply for your system (your other option) we need a bit more info - you said the amp gives a 0 to 5v output but we need the current to drive this amp.
I suppose it will require less than one Watt (but I could easily be wrong - check)
Take a look at THIS. The RI1212 (highlighted red) will give a completely floating, regulated power supply for your system. It will supply upto 2 Watts but...
it is rated for a MAX supply voltage of 13.2v - you should have a simple regulator before this device - a 7812 would be ideal.
Well ive tried to follow what you rote,
But you make it hard work.
Both sides of the thermocouple cannot be grounded,
or it would not work.

I suspect that you mean one side is connected to the
casing of the thermocouple in such a way that when you
bolt it in, the body of the thermocouple housing will
be fixed to the head metal-to-metal, and that it is
made in some way that prevents you from disconnecting
it without damage.

I also suspect that when you say the other side is
grounded also, you simply mean that it is a very very
low resistance. This is normal, a thermocouple is a
very low resistance.
But the wire coming away from the thermocouple should
not be grounded. You should read a very low voltage
from it, to ground.

Now i am not clear where you are putting the probes
of your meter when you say:
"the potential from the head to the batt(-)"

You are dealing with very small voltages and you have
to measure in the right places.

I have attached a small pic, you may be aware that
these places are potential problems, but you may not
be aware that wherever dis-similar metals are touching
there will be a potential difference.

This normally does not matter, as such voltages are
quite small.

However where thermocouples are involved these voltages
can be significant.

They are not just caused by poor connections, that may
require cleaning, these voltages are because of dis-
similar metals coming into contact.

Also, you mention there are spurious voltages getting
into the signal area with the engine running. These i
guess are from currents flowing along 'shared earths'

You say you have tried with the amplifier earthed at
the thermocouple, but of course the battery ground
would go through several other junctions to reach the
case of the thermocouple.

I think your best answer would be to modify the thermo-
couple into a two wire unit with no ground.

I think your problems now are with earth/ground currents
interfering with the signal, such currents when the unit
is running could be from charging, or just the ignition,
or maybe injectors, most units share the common ground
connection, and i don't see how you could easily isolate
the thermocouple signal from other currents sharing the
Batt(-) terminal.

It might be possible to run a (slightly) lower voltage
supply from the 12v rail, with its own earth at the
thermocouple casing, if you were willing to make an
additional connection at the thermocouple.

I will follow up with a possible diagram.
Best of luck with it, John :)


  • Earths_or_grounds.jpg
    12.6 KB · Views: 953
Possibly the problem ...


  • Th-Cple.jpg
    26.2 KB · Views: 990
The potential difference I measured isn't necassarily absolute, it's a comparison. i.e. measured with key off vs key on. I tend to agree that it's ground currents I'm fighting here.
The TC is a washer style I picked up from Omega. Simply no way to disconnect it from ground. It's virtually identical to the TC that you're supposed to use from Autometer with two exceptions. First, the Autometer TC uses a copper crush washer meant to replace the spark plug washer. Second, mine is fraction of the cost, $23 vs $90 The guys at Autometer told me Omega made their TCs and I could order one direct from them if I wished since Autometer doesn't sell direct.
When I said that insulating the TC wasn't an option, I mean it's more of a personal choice, as far as I'm concerned it isn't an option. The head does not have provisions for a temperture well so the only means of attaching the TC are pre-existing bolt/screw locations. Only three are in a location where acurate temps can be read. The only two that I wish to work with are head stud and exhaust manifold stud. The teflon washer or similar insulating device you mention would affect the clamping pressure or simply by cut in half by the clamping pressure in either of these locations. Not to mention that this is a high vibration race engine, I would think that a teflon washer would be prone to loosening the nut. Loctite is out of the question for either location as the operating temps are beyond the release temp for loctite. This is why I don't think that insulating the exisiting TC is possible. And finally if the response time is affected by more than 2 seconds, it's an issue. I've personally witness my head temps fluctuate by a margin sure to cause damage in less time than that, did I mention it's an aircooled engine? No water here to affect heat soak time.
Back to isolating the power source for the amplifier. I've just confirmed this to work. I ran jumpers from another car and powered the amp off it. Readings were accurate and very stable. This seems to be the easiest route for me to take and least expensive. Most isolating converters that I've come across on the web (thanks mechie) are half the price of a new insulated TC, even if I did find a way to use one.
New question, does anyone know of a company State-side that sells these things? --Ryan

PS, thanks for all the time and effort you guys have put into helping me with this.

This circuit put forward,
does not quite have an independent supply,
but almost.

For it to work as intended, if there are any other points
within it that require grounding,
they should be taken to the 'local ground for thermo-
couple' so that the circuit can not be affected by
any ground currents, or odd Pd's.

I hope the circuit is clear to you, it is partly

If you think this may suit your needs,
then i or someone else will try to approximate the
component values for you, based on what you need.

Best of luck with it,
John :)


  • Possible_Th-Cple_Cct.jpg
    32.7 KB · Views: 950

Ive just seen your last post,
the circuit i have offered should operate as effectively
as a separate supply,
see if someone else chips in with a comment ....

Does this unit have multiple temperature sensors ?

John :)
Seebeck Rules!

john1 said:
Both sides of the thermocouple cannot be grounded,
or it would not work.
Not strictly true !
If a steel washer is used, a copper wire is welded to one side and a constantan wire is welded to the other side you will have a stack of three dissimilar metals, the engine block is assumed to be steel (to match the washer); which lead is 'grounded' ?
Maybe neither as both will have a potential difference from 'ground' due to thermocouple action, maybe both - depends on how you argue it - either could be bolted to a steel 'ground' elsewhere without affecting the final pd.

john1 said:
... but you may not be aware that wherever dis-similar metals are touching there will be a potential difference.
As I created a stack of three metals in the above example what is the result ?
It would be a copper/constantan (type 'T') thermocouple - the law of intermediate metals allows me to totally ignore the steel washer !
The actual 'couple here is Chromel/Alumel ('K') but the theory is identical.

The modules I suggested must be available in the US, surely Digikey or someone will have similar products?
Just look for the spec to claim something like 1kV input to output isolation and a regulated output (unregulated devices are cheaper but very load-dependant!).

Been chatting about your engine ...

Are you concerned about the head temps from nossing it ?
Or is it just straight fueling?
Are you doing bracket racing against the clock,
or normal racing ?

Others here are interested too ...

John :)
It's an ACVW engine (aircooled VW for anyone not 'into' the scene - gotta be wierd like me). Not throwing nitrous at it yet, but the bottom end can more than handle it, should take up a 100 shot w/ no worries, more if I feel lucky. Just not in the budget right now. Franky it's just a sleeper, got chipping paint, rust bubbles, haven't washed it in about a year, oh and did I mention it'll embarrass most 'stangs, 'maros, any import where the owner payed more attention to how it looks than how it runs. I've got plans for Auto-X, but not untill I have something else as a daily driver. Not stroked yet, but bored over as far as it can, nice set of custom flowed heads, dual Weber carbs, fat cam, plenty of compression, close ration tranny, yadda, yadda, yadda. It's just a hobby that gets too much of my budget.
Problem is the heads are aluminum, 356-T6 I believe, not positive though. Anyways, 350F is borderline high, 400F is scary, 450F and valve seats start to move, much past that and you alter the heat treatment. Going WOT for say 30s can produce a 50F gain in temps, and since I commute up into the foot hills, it's likely I'll drop a gear and pass traffic on long grades. I have too much invested in this engine to just hope it'll be ok. I need to know when it's safe to go, for how long, and if I should stop short. I'll give up in a heartbeat to save the engine if the need arises.
Back on topic, the TC is welded to a SS washer, the washer is bolted to the exhaust manifold stud right now on top of the manifold. Manifold is ceramic coated so the only contact is through the stud/nut. Sooo, I don't think there would be any further TC action involved in the circuit.
The unit utilizes a single input, but I could purchase aditional 'black boxes' and run more inputs, then simply swith the output from the boxes to the same gauge.
Let me see if I have this straight. Run a new ground from the TC, say maybe an eyelet wire connector on the stud with the TC washer. Run the other side of this through a zener diode (why 10v, only half way through my CE degree) and I'm guessing the cap if for filtering? Also recommendation for the cap value? I think I can adapt that to work with what I have. Like I said, the amp is part of an IC of which I have no clue as to the makeup. The gauge isn't your normal 90 degree sweep gauge that returns to zero with no signal. It's a free floating, full sweep gauge. When the signal is removed, the needle just stays where it was. I presume this is why there are three wires going to it, ground, v+, and v-, for to sortof push the needle on both sides and hold it, figuratively speaking. Thanks again. --Ryan
Hi Ryan,

the more i ponder this, the less i think you will need
the zener, or dropping the voltage, that was for giving
the amp and gauge a stabilised supply of its own.

now i don't think thats necessary,
but just to be sure,
could you fix up again with the supply from another car,

and this time, watch your nicely behaving Th/Couple gauge,
while someone else starts this other car, and tries the
lights on this other car, and so on.

Just to check that your Th/Couple set-up does behave nicely,
even when the other battery is giving little variations of

If it does behave,
and i hope it will,
then the zener and cap will not be needed.

OK ... ?

Oh, and my mate says to be careful with ally heads,
i said i think he knows that ...

Cheers, John :)
Yeah, Al heads are a great, but you have to be careful with them. Though all ACVW heads are Al, side effect of being aircooled. Steel heads just can't shed the heat.
New results. Hooked it up again to the other car, ran headslights, stereo, engine, etc, no errors. So tried running the ground for the amp to the same location as the TC, no erratic readings but completely inaccurate, as in 200F low. Not only that, but inconsistant, non-linear. So am I back to an isolated PS? --Ryan
Hi Ryan,

it might be easier to fit the independent supply you speak of
than to try to make it work on the existing one.

you say you tried wiring the ground for the amp to the same
location as the TC.

Well, the ground for the amp, and the ground for the gauge should
be connected together, and taken to the body or casing of the TC
and not connected to any other ground except at the thermocouple.
If there are lamp(s) fitted into the gauge, to light it at night,
check that the ground wiring for the lamp(s) is not connected to
the circuit for the gauge.

There are some types on this page:

Could you see which is most like the one you have?

The gauge you describe is interesting.

Could you have a good look at the wire from the thermocouple
and describe it please ?

I am getting very curious about this set-up,
Cheers, John :)
K, let's see. I'm not really sure where the ground for the gauge itself is. Need to confirm this, but I seem to remember checking each of the three wires leading to the gauge from the amp with my MM and found no continuity between any of them and the ground lead. The gauge does have a light, but it is electrically isolated from the gauge itself. The light is wired separately and insulated in a plastic housing from the rear. While the gauge may ground though it's body, it's not likely as the plate it's mounted in is powdercoated. I will check into this though.
On the page you linked to, I'm using the TC listed at the bottom, HD washer style. Bought it from them about a month ago. I'm not sure I understand what you're asking about the TC, what is you want a description of?
For reference, here's a link to info about the CHT gauge at Autometer's website. **broken link removed**

Not open for further replies.

Latest threads

New Articles From Microcontroller Tips