TVR Vixen - Instruments

As mentioned elsewhere the mileometer on the car stopped functioning after its visit to Back Home 99 so the speedometer had to come out for investigation. Interestingly it had been to bits before as on the back of the dial it was labelled "TVR 9.91 26250", presumably the reading when it was reset to zero in 1991.

The problem was easy to solve. There is a small reciprocating arm driven by an eccentric which turns a ratchet wheel (short arrow) and it wasn't doing so as the end of the arm was worn. A convenient abandoned Land Rover dashboard provided a very rusty speedometer with a suitable replacement part.

If you should ever feel the need to take a speedometer to pieces do not remove the needle without first marking it's position on the circular wheel (long arrow) behind the dial to aid accurate reassembly.

For months I drove around in this car and the rev-counter needle would occasionally flicker madly. An instrument specialist told me that it probably needed conversion from current sensing to voltage sensing to cope with electronic ignition. I decided to wait for it to fail properly - a wise decision as work on the wiring loom under the bonnet cured the problem - although I never found the fault.

Early TVRs seem to be plagued by problems with minor instruments and there are many which have mismatched instruments. They were originally fitted with AC Delco instruments and the only other vehicle I have seen with a complete identical set is an early Ginetta G15. Apparently owners of early Ginettas have similar difficulties.

The main problem seems to be getting the correct senders. My car had been fitted with a Smiths fuel gauge and sender. A replacement gauge was easy, I got a second-hand one from David Gerald. (The gauges are not hard to find.) I knew from the Vixen S2 that the tank sender gasket was from a Vauxhall and visits to a number of second-hand bookshops looking at old manuals revealed that the Viva had the sender in the top of the tank. A Viva manual revealed the correct resistances for testing the sender. 0-30 ohms for the Viva HB and a higher range for the Viva HC. A quick test with my gauge and some resistors showed that the Viva HB one would work. It took a couple of telephone calls to find a supplier of old Viva parts and I managed to get a sender. The arm needed lengthening slightly which I did by clamping on part of the old arm. It works quite well and shows empty very accurately. (I know to my cost.)

The temperature gauge is another instrument for which the correct sender is hard to find. From my experience with the Vixen S2 I had the part number and in 1986 I was able to walk into a Vauxhall dealer and buy one off the shelf. I tried to track down the right part. An obsolete Vauxhall parts specialist said they had a couple so I bought one. It took ages to arrive as they couldn't find it and then didn't work with the gauge - that may just mean that someone had tried to readjust the gauge before I bought the car. I still have it. Does anyone want to buy it?

The first step was to slowly heat the senders that I had and measure their resistance at a various temperatures. I did a similar exercise on the gauge using a variable resistor to establish what resistance gave particular readings on the gauge. This was surprisingly difficult as my results seemed to vary quite a lot. I ended up averaging half a dozen sets of figures.

Step two was to plot the results on a graph using a spreadsheet - essential for step three.

I used the spreadsheet to calculate what the sender's response would look like if I were to put in a resistor in series and in parallel with the sender. I used both as I wanted the gauge to work across it's whole range - it would be easy to use a single resistor to make it work at one point.

Then I soldered some resistors together and tested it. I tested the gauge in the car with the sender in a pan of boiling water on a camping stove. It wasn't quite right - when the water boiled the gauge read about 95 degrees C. Back to the computer. I think this error may be due to electrical heating of the sender but I looked at the graph to see whether I could alter the resistors to shift the response a few degrees. This worked well and also gave a much better match at the low temperature (and low priority) end of the scale. I calculated the maximum power consumption and bought the relevant resistors. They are mounted on a piece of circuit board in a ventilated box on the rear of the inner wheel arch.

Neither have given any trouble.