Continuing to live my life as an object lesson for others...

I recently had to pull my engine for some repairs. For me, part of pulling the engine includes removing the thermocouple from under the #3 spark plug and fishing it out of the engine tins. This runs to the CHT gauge in the cabin of the car.

And we all know you are not supposed to cut the wire. Everybody knows that.

So I figured I'd cut the wire, add a nice weatherproof automotive plug, and see what happens. All in the name of convenience and curiosity.

I drove 300 miles last weekend. Including going fast up a really long couple of hills (like miles long hills) in 5th gear.
So the results.... At cold start, the gauge reads the same as ambient temps.
Under normal driving, the CHT gauge read slightly cooler then before I added the plug, but within 10 degrees of what I expected.

However, when pushing the car, the error rate grew exponentially. Where I would expect the car to be running 340-350 head temps, the gage only showed about 300. And pushing 90 up hill in 5th for over a mile I would expect the gage to read 360+ the gage only ever got to 325. I did not push past what I know this engine required to get to 375 (which is my back off point).

The result of this experiment is that if I did not know and expect bogus returns on the gage, I could easily roast an engine and start dropping valve seats, while a quality CHT gage was reporting I was WELL under the 400 degree danger zone.

I have another cable I can swap in. Kind of a pain, but not a huge deal. I'll add it o the maintenance schedule.

But, yeah. In this case the group think is accurate. The very slight resistance that the plug will add to the cable will be enough to throw off your CHT monitor.

Do not cut the line between the thermocouple and your CHT gage.

Zach

I'm guessing you didn't have a setup similar to the dakota digital? That thermocoupler has disconnects already built in. With that in mind I'm surprised your setup would be much different. I don't know much about the inner workings of thermo dynamics just that heat is transferred via these two wires to the gauge. I guess if the metals were dissimilar in your connection points that would cause a discrepancy. Good to know though about other types of CHTs.

We just temperature tested a unit we're designing using 7 thermocouples. The logging device we connected to just had terminals with screws to attach the thermocouples. The thermocouples only produce millivolts of voltage.

You should be able to put a connector in series with the thermocouple as long as the connector doesn't add any resistance. It's only the tip of the thermocouple which produces the voltage. If you jiggle (official EE term) the connector while running does the gauge move around?

It is a Dakota Digital actually. But the screws are under the tins. In the past I've always just fished everything out after pulling the #3 spark plug.



I'll have to see. Easy enough test to do tonight after work. Driving over bumpy roads though did not seem to have an effect.

Zach

This was the first thing that crossed my mind. I would think this is what happened. Spoke, please publish your report on the accuracy of those senders. I tested the Dakota vs. the VDO years ago and found the Dakota to be very accurate and the VDO not. I had suspected the VDO gauge but as it turned out the VDO gauge performed well when coupled, no pun intended, to the Dakota sender.

I have the DD as well, and I swear they had a little note with it about cutting the cable, as in you could but there was something about not increasing the length I think.

Free white paper for the day:

Cylinder head temp gauges are just a thermocouple gauge designed to read a specific type thermocouple. There are many different types of thermocouples available depending on what type of wire is used in its construction and what temperature range is being measured.

Dakota Digital uses K-type based on images of their sender and extension wire that I see online.

In K Type Thermocouple positive leg is composed of 90% nickel, 10%chromium and a negative leg is composed of 95% nickel, 2% aluminum, 2% manganese and 1% silicon. These are the most common general purpose thermocouple with a sensitivity of approx 41µV/°C.
https://tempsens.com/blog/k-type-thermocouple

Each type of thermocouple can be identified by its wire colors.

There are two different types of K-type wire that can be used to route to the gauge.

1) Thermocouple Grade wire -- this can be used to make the sensing end and could be used all the way to the gauge -- but it is more expensive wire so it is rarely used all the way to the gauge.

Thermocouple grade wire has a brown exterior insulation and for K-type the internal wires will be red (actually the negative wire -- UGH!) and a yellow wire (the positive wire)

2) Thermocouple Extension Grade wire -- this cannot be used to make the sensing end of the cable. Extension Grade wire can only be used to extend the sensor to the gauge or data acquisition rig. It's wire is made of material that closely mimics the sensing wire but is not exactly the same. It is however, vastly different than copper wire.

K-type Extension Grade wire will have a yellow exterior insulation. The internal conductors will be red and yellow.

Note: There are exceptions (aren't there always). The thermocouple wire can be obtained as single strands - without an exterior insulation holding both together. And if I recall my old VDO - it had wires that were covered with a braided material and had a non-standard color code. Likewise there are huge color code variations between the US and Europe regarding color coding and identification of wire type. If you look online you can find identification charts that cover all the wire types, as well as the US and International color coding.

Thermocouple extensions can be spliced.

However, they are spliced with a connector that mimics the behavior of the thermocouple sensing end and the extension cable. The connectors are specific to the type of thermocouple wire that is used for the senor / extension cable.

Here is an example of a K-type connector
https://www.farnell.com/datasheets/2339540.pdf

There are several different styles and sizes of connectors but each has to be matched to the thermocouple wire type.


Moral of the story: If you replace your standard wiring connector (copper / tin terminals) with the proper K-type thermocouple connector, you'll be back to a fully operational system and won't have to rewire the whole thing all the way back into the cabin.

What type of connector did you use? Since thermocouples/CHT setups are sensitive to small changes in resistance, you might just need to use a high quality connector like a Deutsch that's intended to minimize such issues. One of my cars has an old school EGT which uses thermocouples, and it's set up with Omega connectors that are intended for thermocouples like these.

https://www.amazon.com/T-Type-Thermocouple-...66697&psc=1

This type might work. Give a call when you know the specs on the TC you have. They may have just what you need.

Verify you have K-type (red and yellow conductors). If that is indeed what you have, I just went though my stash and I have a K-type connector pair (male and female) I will send for free if you want to repair what you have.

FYI - you can also buy silicone boots that will go over the thermocouple connectors to make them more weatherproof but I don't have any of those.

Deutsch (some styles) connectors can be ordered with special pins to accommodate the special conductors needed for thermocouples. Think $20 per pin. Standard Deutsch automotive connectors are not going to work as they are still copper pins.

Continuing to live my life as an object lesson for others...


We thank you

Question…if the ends of that red/yellow extension wire were flipped, what would that do to the readings?

If you accidentally flipped the wires of a thermocouple . The readings will just go opposite. When it gets hotter it will read out a colder number, thats all.

The IC for a thermocouple is just an differential op amp (with some other things like cold junction compensation). All it does is take the voltages from the two wires, subtract them, and then amplify a signal to a readable output.

No damage

Yes, its a K type thermocouple.
and Thank You!
Zach

Here's what I used, extension is long enough to route thru center tunnel up to under dash gauge. A very cheap solution.

https://www.ebay.com/itm/203517821362

Saying of the Day:

"Do or Not Do, There is no try"

---Yoda

Did this work with a Dakota Digital gauge?

I have a similar question ... will the cheap thermocouple work with a VDO CHT gauge.

Went and looked at the ad. It says: Type Ungrounded K Type Thermocouple
So if you look up your gauge it should say what type of TC you need.
I think it will work with the VDO.

Well I asked the vendor and got this response:

Yes, the VDO CHT gauge normally use K type thermocouple as input and it is fully compatible with our sender. Please make sure that the yellow wire is connected to the positive and the red wire is connected to the negative. It should work.

Is there any difference in the makeup of that extension wire? I mean, is the positive wire a different metal, etc. Or are the red and yellow sides the same and just color-coded to make sure it hooks up to the thermocouple correctly?

K Type Thermocouple positive leg is composed of 90% nickel, 10%chromium and a negative leg is composed of 95% nickel, 2% aluminum, 2% manganese and 1% silicon

Yes, the thermocouple extension wires are made of different materials between the positive (yellow) and the negative (red) wires.

Thermocouple extension wire is very similar to the sensing wire, but uses less of the expensive metals (Nickel, Chromium,etc.) content in its construction to keep cost lower. However, the overall characteristics of the extension wire are still very close to the sensing wire and are VERY different from simple copper wire. Because the extension wire is slightly different than the sensing wire - its use comes with some constraints such as a more limited range of operation temperature. Example: Extension wire needs to be kept below 200C (not a problem for 914 application) and you want the temperature across the thermocouple extension (end to end) wire to be as uniform as possible.

Introduction of a third metal (like copper extension wire or copper/tin connector terminals) into the circuit will add a 3rd temperature junction that is uncompensated and will affect the accuracy and calibration of the thermocouple.

@Iankarr

Thermocouple extension wire also has different temperature limits and tolerance than the sensing wire. There are more expensive grades of extension wire that are used when higher accuracy is needed (doesn’t apply to our application).

When I used to do data acquisition, it wasn't uncommon to have to run long thermocouple extensions from the test cell into a control room - sometimes 75' or so. It would get expensive stringing sensing wire the whole way - especially when you may have 40 or 50 thermouple channels under test. In the case of our 914's that distance from the engine to the cabin in really only 5 feet so using sensing wire the whole way wouldn't be unreasonable.

Question would be how far away from the joining of the 2 dissimilar metals can the 2 leads change to one type of conductor maybe not made of the same materials as either dissimilar metals?

When the thermocouple is connected to the gauge, both metals change to the type of metal found on PCBs. So somewhere away from the dissimilar joint it is ok to use different metals. Where is that point?

Anywhere two dissimilar metals join will become a new temperature sensing junction with it's own EMF produced at that junction.

In the case of attachment at the gauge, this is the cold junction and it is compensated.

If a dissimilar metal is introduced somewhere between the hot junction and the cold junction it too could be compensated (if temperature at that junction is known) but often that junction will be highly non-linear and would make it tough to compensate for.

I think the thing that throws people on thermocouples is that the sensing is not due to resistance change but due to the Seebeck effect that generates its own voltage potential between the two dissimilar metals as temperature changes. Resistance only begins to come into play when extension cables become very long (100' or so). I'm not saying resistance doesn't matter . . . just that resistance change isn't what drives the operation of a thermocouple.

Yes. See post #5

Let's flip things around. Guess what happens if power is applied to a thermocouple? If you figured that out guess what happens if you reverse that current???

Wait are you saying that the analog VDO gauge reads correctly (it’s got a reputation for being off) when using the Dakota thermocoupling wire & ring? Meaning that the wire is actually the compensated part of the puzzle? Or are you just saying that it fires up and “works” in general? I was wondering this a few weeks ago when I purchased the complete dakota setup while really wanting the VDO to be a more viable solution. Sorry to be so dense, but can you confirm?


.

Both use the same K-type thermocouple sensor.


VDO is an uncompensated cold junction - its just basically a milli-volt meter with a gauge face that is calibrated to cylinder head temperature. It assumes the cold junction ambient temp is 68F (maybe 70F??) if I recall. This is why VDO has more error.

Dakota Digital is compensated cold junction and is more accurate. Internally, it measures the temperature of the cold junction and applies a correction to the displayed value based on the ambient temperature at the gauge (i.e. the cold junction).

Getting the spark plug and ring out has always seemed such a major PITA. If you come up with a disconnect solution it will be a service to us all.

Yes. I had a post 10-12 years ago here showing the results. I used heated oil since it won't cool as quickly as water. IIRC, I compared the VDO setup vs the Dakota setup and finally tested the VDO with the 12ft long Dakota lead & sender. Try it and let me know your results.

The OD of the ring can be gently filed down so that it slips into the round plug receptacle. I used a flat file from a needle file set because it was less aggressive. The material is very soft so do so with care.

I’m confused as to why not leave the ring under the sparkplug and just unscrew the two small connecting screws that are about a ft. down the line
Here’s my DD lead wire:

There are a couple reasons:
1) it’s common to put some heat shrink tube over the screws to protect them and keep them in place when driving. Which just adds another barrier to getting at them.

2) I also have mine routed under the tin (fairly common) so you can’t get to the screws anyway. They are hidden under the cooling tin. There tend to be a lot of wires and hoses on the 3/4 side of the car next to the battery, and this gets the TC wire out of the way.

3) those tiny screws are a PITA to handle and tend to get swallowed up in the engine bay when you inevitably drop one.
Zach

Uh Hu....

Note to self: route TC wire above the tin and use a magnetic screwdriver, got it, thanks

.

Is there a way to join a K type thermocouple to a proper extension lead, other than an Omega plug? Are there butt splices made of the correct material?

I’ve never seen them in years and years of data acquisition testing.

Google search for Chromel butt connector turns up at about $10 and some change per splice, min order 2500, at a cost of $26k, 182 day lead time.

And then you’ll need the other Alumel splices for the negative conductor - let’s assume another $26k minimum order for those.

how bad do you really need butt splices for thermocouple wires?

Looks like Mouser stocks the Chromel & Alumel ring terminals that Dakota Digital is probably using - you could use those and the screws like Dakota does

Click to view attachment

Well that makes the Omega plugs looks like a better option

One other thing - I have a VDO CHT that's missing the harness and thermocouple. It has what appear to be standard male blade connectors at the gauge. I'm not seeing chromel or alumel female connectors. I guess using copper ones would create an uncompensated junction (if that's the correct term)?

This is making the $60 VDO thermocouple and wiring kit more appealing haha

If I recall the VDO did use plain copper female spade at the gauge. Yes connection to the back of the gauge that is considered the uncompensated cold junction for VDO so that transition from Chromel / Alumel to copper right at the gauge “should” be part of their gauge face calibration.

It’s been a long time since I had a VDO gauge in my hand but in all honesty - it serves a purpose which is to get general trends in cyclinder head temp without fretting over 20F one way or the other due to gauge error.

You’re right though - $60 for an extension, sensing end, and connectors isn’t outrageous in this day vs. sourcing all the bits yourself.

I used a digital temperature compensated gauge bought off E-Bay. I'd bench measured temperatures of room temp & with spark plug fitting immersed in boiling water , got correct readings.


Click to view attachment