Racal Dana 9917A Frequency Counter

[Gradall]

Hello

I have a Racal Dana 9917A with a fault.

Using Input B in normal mode, I have no frequency reading (Input A is fine).

But if I put it into Burst mode and I hit the reset button it works fine and I get a reading, any ideas what I could check?

Does anyone have a service manual/schematic for this model or any ideas what to check please?

Regards
Gradall

[trh01uk]

Originally Posted by Gradall

Hello


Does anyone have a service manual/schematic for this model or any ideas what to check please?

Regards
Gradall

Gradall,

I have a manual for the 9917, but not the 9917A. Don't know what the difference is, so I have attached a copy of my service manual in case that helps.


Richard

[trh01uk]

Gradall,

trying to answer the question about why your B-channel input produces no reading, there are a myriad of possible failures here. Anything from a dry joint in the signal path to an open circuit wiper on the input pot. Or it could be something more obscure, such as the gate pulse going missing.

My recommendation is that you inject a large signal into channel B, e.g. from a signal generator, and then do some signal tracing with the most sensitive probe that you have available. Something like a x1 probe feeding a scope with enough bandwidth to display your input signal would be handy. And then go hunting for the point in the circuit where the injected signal disappears (if it does!).

I think the path for this signal ends at the input to IC20 (if I have read the schematic right). So if you can find the signal at that point, you can then eliminate the B-channel input circuit, and the signal gating system.


Richard

[PETERg0rsq]

Check sections 4.17-4.19

this is where the two channels are selected to go to the counter, or burst mode.

This assumes channel B is working correctly for burst mode, and Channel A is fully working.

I have a paper copy of the 9917A service manual, and will scan it when i :-

a) find it
b) get time....

[paulsherwin]

Reopened by request.

[Gradall]

I have struggled away using burst mode, but due to temp nerve damage in my left hand, decided to have another look at this as using burst with 1 hand isn't easy.

I have found a poor quality copy of a circuit diagram but not a full service manual, so to keep this easier, I will refer to the circuit and service manual for the 9916 counter as the area of the fault is identical.

Using my function generator, if I use a very low signal of around 16mV to input B, the counter actually works, if I increase the signal to around 60mV, the display goes to zero.

The signal is lost at IC21b, I have now worked out that the signal is being cancelled by the output of IC21A, this appears to be the "Low signal level inhibit" fed by 747 opamp IC31b, basically with the sig gen at 1mV, the opamp is doing as it should and killing the display, when I increase to around 12mV, the opamp stops killing the display and counting works. If I continue raising the input signal to around 60mV, the input to the opamp actually starts to drop to the level where the opamp (acting as a shmitt trigger) kicks in and kills the display. After further tests, I am looking at either D4 or D5 breaking down.

These have strange part numbers (HP 5082.2811), after some searching on here, the best match is 1N5712.

Does my explanation/diagnosis sound plausible and would the 1N5712 be a close enough match?

Also, if anyone can provide a full service manual for the 9917A, it would be very much appreciated.

Regards
Gradall

[factory]

Datasheet for the HP 5082-2811 Schottky diodes, of course no longer available thanks to Broadcom; https://www.farnell.com/datasheets/89202.pdf

David

[Radio Wrangler]

HP part numbers are separated by a dash "-" not a dot "." it can make a world of difference to search engines.

HP split off its semiconductor arm under the name "Avago" and it was largely bought up by an American financial conglomerate. Broadcom was a successful semiconductor firm specialising in digital TV chips then moved into other areas like wireless connectivity. They were a lot larger than Avago when Avago bought them out, but really it was the financiers behind Avago did the buying. After a short while, the people at Avado decided the Broadcom name was more attractive than their own, and as they now owned both, they changed to Broadcom and ditched the Avago name. Various top personnel had changed by this time and very quickly they dumped almost the whole of the HP/Avago product range of high performance diodes, transistors, mmics.

So it wasn't really Broadcom who (insert own choice of rude verb here) the HP/Avago parts. They did it to themselves.

I spent a lot of time trying to find replacements for these in the company I'd gone to after HP had pretty much pulled out of product development in the UK.

David

[Gradall]

I have sort of fixed this, but not correctly. I have removed resistor R162 on the 9916 circuit (R158 on the 9917A circuit).

This disables the low level circuit inhibit, the counter works fine now testing from 10mV up to 1V, it is very usable now.