Last spring I bought a "used" VDO CHT gauge (from a
unnamed member 
)at the GnR flesh roast. At $20 this seemed like a good deal. I was unable to locate a VDO sender so I ordered another brand from Dakota. Once installed I discovered that it did not work. 
I had opened the ring terminal to 14mm and believed that my doing so might have ruined the sending unit. So I ordered another another from Dakota but this time with the proper end (14mm). I avoided this one previously because of the 14' lead. It is made long for the bus application. But before I could install this one I got side tracked by some erroneous advice or article (I forget) which convinced me that I needed a matched set. I contacted the VDO supplier and found that they were still available. I had been told they were no longer available originally. Once my VDO arrived it was installed. Much to my chagrin the gauge still did not register anything. I realized that my "used" gauge was trash at this point and ordered new gauge. Once installed and coupled to the VDO sender the new gauge showed signs of life. 
Or so I thought!I thought that I would take a quick run up the road to Ron's (aka Slits) and enjoy my "now" working CHT gauge. It was a coolish California morning but it appeared the car was running a tad bit warm. After talking with Ron, I believed that I needed to recheck the timing, fuel pressure, valve gap and cooling flaps. The return trip was tagging 400F if I ran over 50 mph. OK, what did I expect? I built this engine w/o any prior experience in type. I contacted Geoff (aka Bleyseng) whom had rebuilt the MPS regarding the mixture. After a few pm's I decided to look at the CHT a little closer. Geoff had asked if I had ever calibrated the gauge. Well, no. Who in the right mind would mfg a CHT gauge that was not accurate? The answer; everyone. More especially VDO!
I put a pot of water on the stove and got it to the boiling point. I checked the temp of the water with a mechanical candy thermometer against 2 digital ones on different multimeters. Almost 25F difference between the lot.
I finally realized I should determine what the boiling point was at my altitude ([212F]-[elev x .0018] which was 209F. Once that was determined I began to take readings from the gauge with all the different configurations that I could assemble from my nest of parts. e.g. VDO sender only, Dakota sender only, VDO w/ factory length leads then with factory leads cut to length (which they say not to do ) then the same set of tests w/ Dakota sender and different lengths of Dakota's factory leads. What I found were temps that ranged from 250F with the VDO gauge, sender and factory leads, to 210F with the VDO gauge, Dakota sender and Dakota leads. I had a set of Dakota leads which were cut to length that gave the same reading; 210F. To summarize, I found that;
It doesn't matter if you enlarge the ring terminal. It is there to transfer heat to the thermocouple. The thermocouple consists of a cold junction and a hot junction. The ring terminal is the hot junction and the other end is the cold junction. The two wires are of dissimilar metals and creates a difference in voltage that is measured between the hot & cold junctions with a determined set point. Herein lies a problem. The thermocouple and gauge are calibrated using 72F as a set point. Is there a place in the Hell hole that stays at 72F. As such the gauge will proportionately register higher than actual temps when ambient temps are below 72F, and lower than actual temps when ambient temps exceed 72F. Simple math but something you should be aware of.
Cutting the leads between the thermocouple and gauge has no effect on the readings, as far as I could determine w/ 50 degree increments.
Each thermocouple has it's own particular set of values, though Dakotas only varied 2 degrees between themselves.
VDO's sender is a hunk of junk.
The best news??? My CHT's are normal. Whew!
