monitoring cylinder head temperatures
 

Is it worth doing?

I know that it's done a lot on airplane engines, but a significant portion of those are air cooled and thus don't have coolant sensors.

But I was wondering if it might be a good way of indirectly keeping track of piston temperatures to run 4x CHT probes under the spark plugs with a 4 way gauge. That way if one cylinder is running lean, it will show higher temperatures than the others and you can go "oh shit I have a problem I need to address" instead of melting a piston. Or am I completely misunderstanding what CHT probes can be used for?

Techsalvager?

4 widebands. Or just 1 in the normal spot, and watching it.

Monitoring watercooled cylinders with four oxygen sensors: "Oh shit, I'm about to have a problem I should address"

Monitoring watercooled cylinders with four EGT sensors: "Oh shit, I'm now having a problem I should address"

Monitoring watercooled cylinders with four CHT sensors: "Oh shit, I had a problem I should have addressed and now my motor is blown up"

Lol

Alrighty, so it's not a good idea. I can live with that.

I would think EGT sensors would be easier to install and probably more reliable. I've never heard of anyone running multiple WBO2 sensors.

Andrew is spot on. Both O2 and EGT sensors respond extremely rapidly to power and mixture changes.

CHT response is a damped, time-derivative type of reading. It would literally be tens of seconds to minutes before a mixture problem became evident on a CHT sensor.

And now I have. Cool.

CHT would only be useful for steady state operation anyway.

Yup. EGT used to be common for this, but with widebands getting cheap youll start seeing ^ around more and more regularly.

To be fair, the sensors don't last long sitting in a turbo manifold. It's really meant to dial in per-cylinder trims and then get removed afterwards. EGT sensors are a better option if you want something that lasts a while and still provides more data than a single sensor.

This sums it up very well and accurately. If you want the best I'd do option 1.

Also if you're wanting it perfect, I'd put some higher-octane-than-normal fuel and get a well-established baseline for your knock sensor and set fairly tight buffer so that say, 8% to 10% more knock activity than the base the per-cylinder knock detection pulls timing. Of course you already have that if you think you need 4x O2's or EGTs.

Word of warning. The output of a wideband is skewed by pressure. No idea by how much or in what direction, but something to read into if you persure this further.

What the fuck are you talking about?


I do miss techsalvager. He had the ugliest turbo setup. I remember that sea green silicone pipe.

Google found some numbers

https://www.miataturbo.net/attachmen...ine=1452898746


Source: https://www.rbracing-rsr.com/rsrgauge.htm

Megamanual mentions it too.

Well, if you are using them to trim together it doesn't make a difference what the error is.

That is actually a really good point.

True, but why would you do that when you could do it for 1/5th cost with EGT sensors?

You can leave them permanatly fitted as they last forever, are cheap, and are simpler to wire. Drive one with an AD595 with no external parts, unlike an 02 which needs a complex controller.

02's per cylinder are only worth the effort if you are pressure compensating, thus getting absolute values.