Help needed with - Racal Dana 9081

[3pinplug]

Hi Paula
Wow! thanks for taking the time to check the waveforms on your 9082 that is really good of you. I am not an expert on the oscilloscope so will do what i can with measuring the waveforms, one question if i may are the likes of Q6 , Q7 etc fitted in a holder and can simply be pulled from the board or do they need desoldering, not seen the black housing around them before?
Thanks again
Mark

[frsimen]

Hello Mark,

I'm not sure of the answer to that and I've just put the 10,000 screws back into the shields of mine! I suspect they are mounting pads, which just space the transistors from the PCB but I could be wrong. A gentle tug will answer the question for you.

You don't need to remove the transistors for a quick test. All you need to do is to use the diode test facility on your DMM. Check from base to emitter, and base to collector, both ways around. An NPN transistor will conduct when the base is connected to the positive prove of your DMM, an NPN transistor will conduct when the base is connected to the negative probe. The other way around will show a higher resistance.

The tests aren't entirely sure fire but is better than removing the transistor with the risk of damaging the PCB. If the transistor shows a short or an open both ways around it is faulty for certain.

Paula

[PJL]

Impressive progress, seem to be nearly there ...but where did you find the full circuit as the one linked to in post #5 appears to be missing some circuit drawings?

[CambridgeWorks]

Paula, you are an absolute mine of information and expert at passing it on in an easy, understandable manner.
I keep out of discussions on these sig gens as it is well over 25 years since I worked on them as part of my self employed work.
I must say, I never saw sockets fitted on any 9081/9082 at all.

Stock faults istr are the reed relay (fm board??) that if faulty gave no fm mod indication on the meter, despite being present. The other was the flickering fm indication on the meter with no mod present. This was never cured, but reduced often by just cleaning up the main vco pcb screw fixing positions (earthing) with a fibre tip brush.
One peculiar fault I do recall is a mains modulation hum. Rotating the toroidal transformer maybe 30 degrees helped in reduction.
Why? I don't know!
Have fun. I am sure this will eventually be working to spec.
I still have my 9082, not used maybe 5 years? Plenty of tants to explode yet!
Rob

[frsimen]

Hello Rob,

Thank you for your nice comments. Your notes on this forum in the past have been very helpful in trying to tame some of the problems in my 9082. The flickering FM indication remains a problem, it's very fussy about how tight some screws are. Too tight makes it bad, too loose and it's worse. I've got it to the point where it is acceptable now, but it barely meets the specification in the manual on that parameter. The other area where it isn't meeting the specification is the stability when using the fine tune. The spec does seem to expect an awful lot from a crystal oscillator which uses an ordinary crystal.

I had some tantalums fail in mine probably 20 years ago, the smell is still lingering inside the case!

The copy of the manual I'm using can be found here:

https://www.g7syw.com/files/racal-da...aintenance.pdf

It has more pages than the Electrotanya version.

Paula

[3pinplug]

Hi Paula

I attach a pic of the circuit with the relevant traces I obtained at the test points, TP15 and TP17 appear to be dead! I then popped the board out and conducted the DVM measurements on the two transistors (Q6&Q7), with the following results:-

Q7
Positive probe on base
B - E = no reading
B-C no reading

Neg probe on base
B-E = 0.747
B-C = 0.741


Q6
Positive probe on base
B-E 0.619
B-C 0.717

Neg probe on base
B-E = No reading
B-C = 1.120

So a B-C short on Q6 - does that look right from the measurements and the traces obtained?

Q6 appears to be a 2N2369 and the more readily availbale of the two transistors so hopefully it is that, but i will await your thoughts.

Regards

Mark

[frsimen]

Hello Mark,

Is the voltage on TP19 0V? If it is, there should be pulses at TP17, so Q6 could be faulty and it would be worth changing it. The switching speeds aren't that high in that area, so you could substitute another NPN small signal transistor if you have one. A BC107/8/9 BC547/8/9 or 2N3904 will all work in that position.

If TP9 is +12V, I would leave Q6 for the time being, as there should be no pulses on TP17 in that condition.

I've spent a while trying to understand how the circuit around ICs 5 and 9 is supposed to work. I've concluded that it's probably a victim of a fault elsewhere, rather than the cause of the fault.

Its purpose is to control the switching of the divide by 160/161. It counts the incoming clock from TP17 and when a particular value is reached it causes IC8a's output to go HIGH. That switches off Q7 and stops the clock pulses at TP17. It also switches the divide by 160/161 to divide by 160. The output from this circuit ends up at TP18 and has been shown to be working.

What should happen next is the problem. The other three ICs, IC21, IC14 and IC25 should still be counting. At some point in the count, they should trigger the J-K flip flop IC10, which will put a 12V pulse on its Q output and, via Q8 onto TP16. This pulse is applies to the preset enable pin on all the counter ICs. The count on ICs 5 and 9 will change and Q7 will be switched on again, returning the clock pulses. The divider will switch to divide by 161.

My earlier measurement of at TP16 looks suspicious, I will have to repeat it later as I think it should be a 10V or more pulse, not what I reported previously.

Even with IC5 and IC9 not working, the rest of the counter should be working, with an error in the divide ratio which will make the final locked frequency off somewhat from what it should be. This means we should turn our attention to IC21, IC14 and IC25 and their associated circuitry (IC6 IC8, IC15, IC22 and IC24).

It's probably best to start with the basics. Check that the 12V supply is present as follows:
IC6, IC8, IC15, IC22, IC24 Pin 14
IC5, IC9, IC21, IC14, IC25 Pin 16 - It will be useful to check IC5 and IC9 for completeness.

Next check that the 10V p-p clock waveform is reaching pins 15 of IC21, IC14 and IC25.

Check that pin 1 of IC5, IC9, IC21, IC14 and IC25 are at 0V, there may be a brief pulse if the circuit is working.

Check for a 10V p-p signal at pins 6, 14 and 2 of IC21. (the output from pin 11 isn't used).

Check for a brief 10V pulse at pin 7 of IC21.

If all pins are good, move on to IC14 and check for a 10V p-p signal at pins 6,11,14 and 2, plus the brief 10V pulse at pin 7 of IC14.

If all is good, move on to IC25. looking again at pins 6,11,14 and 2. The output from pin 7 isn't used.

These tests should narrow down the search area a little more. Report back with anything that isn't as described above.

Paula

[3pinplug]

Hi Paula

Test resulst as follows:-

Volatage on TP19 is 12.26v and no pulses on TP17.

TP16 has 11.55v measured with DVM but when i put scope probe on pin i get nothing.

IC6, IC8, IC15, IC22 ,IC24 Pin 14 = 12v present on each pin

IC5,IC9, IC21, IC14, IC25 Pin 16 = 12v present on each pin

Checked for 10v P-P on IC21 on
pin 2 scope set 5V amp/Div and 2us
Pin 6 scope set 5v amp/div and 0.2us
Pin 14 scope set 5v amp/div and 1us
see pics


10v P-P reaching pin 15 IC21, IC14 and IC25 - yes it is


Pin 1 IC15, IC9, IC21, IC14 andIC25 are at 0V, they were when checked.

10V p-p signal IC21 pins 2,6,14 - it was seen.

brief 10v pulse at pin 7 of IC21 - it was there all the time, no brief pulse observed. - see pic

IC14 Pins 6,11,14 and 2 - all have the 10p-p waveform.

IC25 tested as above.

[frsimen]

Hello Mark,

Firstly, the waveform that I posted yesterday for TP16 was wrong. It should be a 12V positive pulse, every 400us. It just so happens that it was missing when you tested pin 1 of the ICs. It looks now like the fault is somewhere in the logic between the outputs of the counter ICs and TP16.

There are three test point to check. TP7, TP8 and TP15.

TP7 should show 10V positive going pulses, in a mix of single and double pulses.


TP15 should show something approaching a fairly wide pulse with a period of 3.3ms.

If TP15's waveform looks good, check at pin 12 of IC8. You should see an inverted version of the waveform there.

Check pin 2 of IC21, which you have seen working. Move to pin 10 of IC22. You should see a similar waveform there.

Check pin 11 of IC22. Assuming TP7 was as described, you should see a series of negative going pulses, rather irregularly spaced, about every 250us.

TP8 should show a single brief (600ns) pulse every 330us or so.

With these checks it should be possible to home in on the culprit.

Paula

[3pinplug]

Hi Paula,

Latest test resuts as follows:-


TP7 showed 10V positive going pulses, in a mix of single and double pulses - see pic TP7 latest

TP8 single brief pulse was observed but over top of the waveform shown in attached pic, managed to catch both in the pic - See attached pic TP8 latest


TP15 should show something approaching a fairly wide pulse with a period of 3.3ms, see attached pic TP15 lastest showing waveform seen.

If TP15's waveform looks good, check at pin 12 of IC8. You should see an inverted version of the waveform there.

Check pin 2 of IC21, which you have seen working. Move to pin 10 of IC22. You should see a similar waveform there - couldn't see anysort of waveform here - so looks to not be there. Checked pin with DVM and recorded 12.32v on pin 2

Check pin 11 of IC22. Assuming TP7 was as described, you should see a series of negative going pulses, rather irregularly spaced, about every 250us - yes did see this and they were the same!

Thanks again

Mark

[frsimen]

Hello Mark,

The waveform at TP7 looks fine, it's not quite the same as mine but you don't have the PE (pin 1) pulses from TP15 which will change the appearance if present.

The waveform at TP8 isn't great, nor is the one at TP15. Both look like you have a poor contact with the probe.

I thought earlier on that you saw signals at pin 2 of IC21. Occasionally, I've found that the counters stop when probed, so perhaps that's what happened here. If it happens, the frequency readout drops to zero too. Switch the generator off and try again after a few seconds.

A further test for IC2c lo check that pin 9 is at 12V, then check pin 8. If pin 8 is at 0V, the output should be at 12V. If pin 8 is 12V, the output should be 0V and if pin 8 has pulses, the output should be an inverted version of the pulses.

Unfortunately, this was all rather inconclusive. I can't believe that all these ICs are bad, so there could be something wrong with your test set up for that series of tests. Could you run this last series of tests again, please plus the additional test for IC2c.

With CMOS, you should see voltages which are very close to zero or very close to 12V or switching between those limits. You should not see any voltages which appear to go below the 0V line. If you do, it hints at a poor connection somewhere.

Paula

[PJL]

Are all scope pictures taken with the same scope channel/probe and gain? Based on the TP7 result it looks like it is set for 5V/division. As all of these TP's are at CMOS logic levels, they should all be near 0V or +12V or transitioning between them.

TP15 looks like noise, can you check it again. If you are sure it is like this then put your scope on IC24B, pin 13 to check it is the same as TP15, if it is then put your scope on pins 9, 10,11 and 12 looking for similar noise.

TP8 also looks bad but that will be affected by the output of TP15.

[3pinplug]

Hi all
Firstly apologies for my last post, i thought i had typed more but it didn't appear.

Hopefully this one is back to normal!

All set at 5V Ampl / Div

TP15 - No trace observed or movement on oscillocope, measures 0.00v on DVM when checked.

Pin 2 of IC22 - yes i get the series of negative pulses

Pin 2 IC21 & Pin 10 IC22 - see the same waveform present on both pins.

TP8 - No trace observed or movement on oscilloscope, measure 0.00v on DMM.

Pin 9 IC21 at 12v, Pin 8 at 0.00v

IC26 Pin 13 no trace observed or movement, 0.00v on DMM.

The trace you suggested may be noise i would agree with, as I really had to hunt to display that.
Thanks again to you all for your continued help
Mark

[PJL]

Looks like the counter is not running properly but I will leave it to Paula. Do check your scope is setup correctly, I suggest DC input mode 5V/div (or 500mV with the preferred x10 probe), adjust the vertical position so you know where 0V is, then connect to TP7 and adjust trigger. All scope readings of CMOS logic can then be done without any further adjustment to gain or position.

[frsimen]

Hello Mark,

Things are looking a little more promising with this set of results. There are still a few gaps that need to be filled in for a positive diagnosis but you're getting there.

TP15 should have a pulse waveform, but you're measuring a permanent low. Earlier you reported that pins 6, 11, 14 and 2 of IC 25 all had a pulse waveform on them. Can you confirm that is still the case. You may find that pins 14 and 2 don't change state, they were fixed low on my 9082, your results could be different. Any outputs which are stuck at +12V will stop things working properly.

If the pulses are present, check that they turn up on pins 10,12,11 and 9 of IC24b. If any are missing there, check for bad connections, broken tracks etc. If they are all there, but there is no output on pin 13, IC24b is faulty.

If the outputs (pins 6,11,14,2) on IC25 are not counting, check for a pulse at pin 5 of IC25 and pin 7 of IC14. If the pulse is present at IC14 but not IC25, check for a crack in the print or a poor connection.

If the pulse is present on IC25 pin 5 but it isn't counting, IC25 is faulty.

To get a pulse at TP8, all the inputs to IC8b must be low at the same time. You're measuring a low at TP15, so the output of IC6d will be high and that will stop IC8b going low even if everything else is working.

I hope that by sorting out the problem around IC24 and IC25 you should be a lot nearer to success.

Paula

[3pinplug]

Hi Paula
when you say the pins are “counting”, not sure what that means can you give me a bit more info on recognising that please. As mentioned am still learning and learning a lot from this so this would br something else i would have learnt.
Thanks again
Mark

[frsimen]

Hello Mark,

Counting was intended to mean pulsing from 0V to 12V repeatedly. Going a bit further, the frequency of pulses from pin 6 should be higher than those at pin 11. That pin will be at a higher frequency than pin 14, which is at a higher frequency than pin 2.

Paula

[3pinplug]

Hi Paula
Latest results as follows:-
All at 5v ampl/div

IC25 pins 6,11, I see a square wave, pins 2, 14 i see 12v on scope.

IC24 pins10,12 I see a square wave, pin 11 12v on scope, pin 13 is dead, nothing?

IC25 pin 5 pulsing seen

IC14 pin 7 pulsing seen

Found a dry joint on underside of board IC21 pin 15 but it is just an isolated pad so not sure it affect anything but will redo the joint.

Does the above mean IC24 needs replacing?

Regards

Mark

[frsimen]

Hello Mark,

The isolated pin on IC21 is probably pin 11 rather than 15 but, as you noted, isn't connected to anything.

From your measurements, IC24 is doing what it should do given the inputs. IC25 is the problem, as pins 11 and 14 should either be at 0V or changing state. Permanently high is not the state that's needed. There's not much else to check, other than pins 3 and 13 of IC25 are both at 0V

If pins 3 and 13 are at 0V, IC25 is looking like it is a good candidate for replacement. It may be possible to prove it if you can desolder pins 11 and 14 only so that they are left disconnected. I've not had to change any ICs in my 9082, so I don't know if the holes in the PCB would allow that.

If it is possible, you can connect pins 11 and 9 of IC24 to 0V via 10k to 100k resistors (use the resistors to prevent a nasty problem if IC25's pins aren't isolated fully!). With that done, the output frequency of the generator ought to drop and there should be pulses at TP15, TP8 and TP16. If there are no other problems you should be able to tune the generator, although the step sizes may not be what you are expecting.

Paula

[3pinplug]

Hi Paula
I will check IC25 pins 3 & 13 in the morning and let you know the results. Its a double sided board so even if i desolder pin side the component side is also soldered so think it will be difficult to isolate the pins. If IC25 is suspect following these tests and the IC’s appear to be readily available i will change it.
Thanks again, your electronics knowledge is amazing and a credit to yourself and the forum.
Regards
Mark