Tuesday, 23 August 2016

Marconi TF 2411 Nixie Counter Repair

Well I seemed to have inherited another Nixie counter.  This one is made by Marconi Instruments and is 70’s vintage.  The TF 2411 is a 50MHz counter which offers a wide range of functions, has three input channels and uses a mixture of DTL (Diode Transistor Logic) / TTL (Transistor Transistor Logic) integrated circuits.

This counter uses a non-standard XLR-LNE three-pin mains connector, so the first challenge was to try and obtain the correct mating connector.  After a bit of searching around I managed to obtain a suitable connector from a fellow UKVRRR forum member.   Having made up a new mains lead and suspicious XLR-LNE plug, I gingerly powered the unit on. It passed the smoke test, phew. 

XLR-LNE Connectors

I checked all of the voltage supply rails which were ok.  The display was incomplete and according the fault finding chart in the manual, it suggested a faulty Nixie tube.  I found that five of seven Nixie tubes were cracked and therefore rendered useless,  I had a quick look on eBay to find some replacement ZM1162 tubes.   

Dead Nixie tubes,   the cracks are outlined in red.

With the replacement Nixie tubes fitted, I proceeded to carry out the self-check as outlined in the manual, well that proved to be fruitless as the readout just displayed zeros.  I suspected the 10MHz frequency standard wasn't running.  So back to checking voltage levels at different points, I found that the +20v supply wasn’t present at pin 1 of the frequency standard unit. I followed the mass of cables to find where the wire went and discovered that it had broken at a pin connection on the amplifier PCB, re-soldering this wire restored the supply to the frequency standard.

I could now continue with the self-test, I switched through all time base positions but 1 – 4 just displayed zeros.  Hmmm well time to dig out the old oscilloscope and check the time base decades which are good old 7490N decade counters.  I could see the signal being passed from one decade counter to the next on the time base positions 5 – 8 but the signal wasn’t getting to the lower group of time base decades.  It goes through a Ferranti ZN224 (Quad 2 Input NOR gate) chip, I was probing around this chip and one of the outputs appears to be dead.  Therefore the 10MHz signal wasn’t able to get to the lower group of decade counters.

Below is a section of the Logic board schematic.

Dead logic chip outlined in red.

These chips are long obsolete but I did manage to a find some ZN346 (Quad 2 Input NOR gate) IC's on eBay and these according to an old data book are the same as the ZN224 but slightly slower, though I’m not sure if the pin outs are the same.  

Overall I'm very happy with the unit, it is beautifully made and works mostly. I may one day replace the dead logic chip but for the time being it has been consigned to the loft of plenty.

Some more pictures can be found here 

Update 26/08/2017

Having replaced the suspect ZN224 logic chip on the logic board, this seems to have cured the problem.  I switched the function switch to test and rotated the time base through it's positions, 1 - 4 now work.

Bad DTL logic chip has been replaced.

One bad chip.

Finally I checked my counter with my 10MHz OCXO module and all is good.