Discrepancy in header temperatures between both cylinder banks

I am after some advice as there appears to be quiet a big discrepancy between the header temperatures of cylinders 5-8 compared to 1-4. A quick search of Google produced opposing opinions with some saying it is not important and difficult to tell at idle, whilst others disagreed. I bought a cheap thermal thermometer to check and discovered after 5 minutes of idling, headers on cylinders 1-4 where showing around 30 degrees C and 1-8 90 degrees C.

What alerted me to this was before the car was put away for winter in November I did a 50 mile run and I felt there seemed to be considerable more heat coming from 5-8 headers compared to 1-4. I have no idea what sort of temperatures or discrepancy would be acceptable, and the headers are ceramic coated. I believe the car is also running rich at present, but all carb settings are the same.

Any thoughts most welcome.


The temperature variation from side to side is way too much. Measuring the temperature of each pipe near the head will isolate any cylinder that's weak. Something's not right, either mixture-wise or even no ignition at all.
The temperature variation from side to side is way too much. Measuring the temperature of each pipe near the head will isolate any cylinder that's weak. Something's not right, either mixture-wise or even no ignition at all.
Thanks Bob, the temperatures were all read from the pipes near the head. I understand what you are saying but all four pipes were the same temperature so no ignition would mean 4 cylinders down and it's definitely not running on 4 hence my confusion.


Any chance your running lean on the hot side? Check to make sure you don't have any air leaks over there, and that there is no obstruction of fuel flow.
Plugs are sooty, fuel rail is common and the hot side is supplied first. The engine was dynoed around five years ago, on a rolling road and IVA tested around three years ago. Did not notice anything untoward then.

Terry Oxandale

Skinny Man
Make sure the throttle plates are equal on both sides (or between cylinders). On my individual stack injection, a very slight difference in the openings (angles) at idle position produced some temperature differences that disappeared once off idle. At idle, with the plates nearly closed at idle, an extremely small angular difference looks big until the plates are opened.

Ian Anderson

Lifetime Supporter
is it an infrared thermometer from Maplins?
i have one too and had crazy readings out of it.
my headers are stainless and blued as normal, and when testing on a blue section to a shiny one the temperature difference is huge!
That said I have not managed to get constant numbers across all the pipes, even tried sanding a section to get shiny one them all!


Bill Kearley

Do you have a velocity gauge to check balance at idle? float levels? equal jetting?
I had a big variation until I set things up with the velocity gauge, balance the air and fuel flow = even temps.
The temperature reading from an infrared sensor is heavily dependent on the emissivity of the surface. If two surface temperatures are actully the same and one is black and the other is shiny there will be a big difference in the readings. You can buy infrared sensors that have an adjustment for emissivity.

-Bob Woods

Howard Jones

Where I worked for many years we maintained very high power control systems for transit trains. I am talking in the range of 500-1000 amps at 1000 volts DC. There were a LOT of heatsinks and blowers involved. In an effort to balance current flow and on/off timing of thyristors systems a lot of effort was made to understand temperatures on different heatsinks.

The one thing that was universally agreed on was those cheap infrared handheld type sensors were unless when trying to point them at more than one sampling point and trying to use the temp reading for anything other that a comparison of same sampling point when changing test parameters.

You can check the same spot to see if the temp has changed as a result of your modifications but the temp read vary rarely corresponded to a actual thermocouple derived reading.

They are somewhat useful considering their cost but expect very inaccurate ultimate reading. You might want to use temp indicating paint as it works very well.


Ian Clark

I'm with Terry on this one. With individual runner systems like four Weber IDA/IDF carbs or an eight stack EFI system the lack of a common plenum makes equal throttle settings absolutely necessary.

All cylinders must draw equal vacuum at idle, throttle take up off idle must also be the same. The linkages connecting the banks (coupled pairs) to the main bell crank also must be equal. The bell crank pivot must be equal distant from each bank. Check to make sure the carbs on each bank are parallel to each other and the opposite bank. Get all that correct and the airflow should start and stay equal all the way up.

My suspicion is one bank is opening faster (linkage problem) and going lean hence the higher exhaust temps on one side. You'll need a synchronometer to check the airflow. Good idea to reset the idle mixture screws and put in a fresh set of plugs after the linkages are set up and before firing her up and checking the airflow.