Mixer Meter Bridge

[AS74]

Is anyone willing to point me in the right direction as to where to look to fix the channels not lighting up?

I've replaced a number of transistors found not to be working, changed all ic's and managed to get it working
except for the few channels.
Thank you. 

[Rob Strand]

I see two angles.  The LED selector (4022's) or the audio selector (4066 or whatever that switch is).

If you have an oscilloscope check for blips on the output of every 4022 output, then follow the blip through to the collector of T19.

Checking the 4066 is a little trickier.  I'd probably put the oscilloscope trigger input on the 4022 then put one channel on the input side of the switch and the other channel on the other.    If the switch is working the input will follow the output.  If not you should see something different or maybe nothing on the output.

[AS74]

Could the problem originate on the group boards?

I don't have access to an oscilloscope yet.

Chips are 4011, 4022, 4016 and lm339.

[Rob Strand]

Could the problem originate on the group boards?

I don't have access to an oscilloscope yet.

Chips are 4011, 4022, 4016 and lm339.

It's hard to know it's one of those problems where you have to try a few things to narrow it down.
If the bar display works on at least one channel then it's highly likely the LM339 and vertical bank
of transistors is working.    So that leave the problems to either the channel selection, which can
be either the 4022 (LED select) or the 4066 (audio in).   Since the 4066 is also fed off the 4022
you need to confirm the 4022 is working.

Without a CRO you would have to be more forceful.   For example, you could swap the 4022 connections
of two banks by lifting the connection of R27 to the 4022 of a working bank and a non working bank and
swapping the wiring.   The idea is to see if the non-working bank of LED starts working.  That would point
to the 4022.

There's quicker ways to do forceful checking like this but with so many banks the LED current is too
high to force a channel on permanently.    You could easily fry something, like the LEDs or transistors.
Who knows, the previous faults may have already effectively done something like this and you now
have a few fried LEDs.

[AS74]

I have checked all LED's and found only 4 (i think) not working.

They dull ones have lit up bright at earlier times.

[Rob Strand]

I have checked all LED's and found only 4 (i think) not working.

They dull ones have lit up bright at earlier times.

Perhaps doing the R27 flip thing would be a good place to start.

it's highly likely the other bad channels have similar problems so by focusing on getting one bad channel working you will know what to do to the others.

[duck_arse]

if all those cmos chips have been running on 17V supply since 1983, they might be very tired of sailing so close to the wind, and could have given up the ghost here and there.

[AS74]

Thanks for your help. 

All the IC's are new.  Replaced previous to posting here.   

I've lifted R27 and wired to non working banks of LED's with no change.  Replaced one of the transistors following the equivalent resistor of a non working bank. No change. 

I assume the problem is in a previous stage of the circuit?

[Rob Strand]

I assume the problem is in a previous stage of the circuit?

Without an oscilloscope you will need to try a few things in order to find a lead.
The fact the whole circuit is switching makes it tricky to debug without an oscilloscope.
If I knew where the faults were I would have mentioned it.  I'm really trying to think of ways of getting a lead on where the problems are using simple tools.

One weird thing I noticed on your photo is some channels have LED working in the middle but not at the ends, although they seem quite dim.   In all but one case on those channel 5 LEDs are on but dim and 5 LED are off.   If you look at the circuit each channel of 10 LEDs is split into two lots of 5 LED, one for -28dB to -12dB and the other from -8dB to +8dB.   Normally for LEDs which are off the comparator (IC11 through IC13) *divert* current via the diodes.   The corresponding transistor (T29 or T30) supplies constant current regardless if the LEDs are on or off.   If one of the transistors (T29 or T30) is faulty, or there is an open circuit along the LEDs, or the diodes are open circuit,   the transistor cannot supply current.  What happens in this case is the transistor for the off LEDs hogs the voltage set-up on the base of the two transistors and R44, then that makes the on LEDs go dim.   You should be able to find the fault on those off set of LEDs.

On a bigger scale of things.  You can compare the working an non working channels.    Using a multimeter set to DC you can measure the DC voltage at the output of each of the following,
- 4022 outputs
- collector of T19
- base of T29 and T30  ; perhaps best if you connect the meter + lead to +17V instead of 0V
- emitters of T29 and T30 ; perhaps best if you connect the meter + lead to +17V instead of 0V

You should see a pattern of voltages where all the good channels show more or less the same voltages at each of the points and the faulty channels will show some differences.   By recording the good and bad voltages it's possible to deduce what the fault is.

How confident are you that the diodes are OK?.

[AS74]

Thanks

Not really sure at all.  How would I test them?  Not sure if they can be checked  in circuit.

[Rob Strand]

Not really sure at all.  How would I test them?  Not sure if they can be checked  in circuit.

Use the diode test function on the DMM.   You should see around 0.6V to 0.7V in one direction and (ideally) open in the other direction.   It wouldn't hurt to test them but there's a lot to check.

At this point I reckon measuring the DC voltages will provide the most information.   It's going to weed out some problems with the transistors and stuff like that.

[AS74]

Ok.  Done some measurin'.  All measurements with the oscillator on full volume.  Turning the oscillator down does change some LED behaviour mainly the right channel LED's which have dimmed since the photo was taken. (?) Turning the oscillator down until the top LED shuts off brings back the brightness for the rest of the row.

4022 Outputs.
IC3 : Measures consistently around 800mV except pin 10 at 10V.

IC4 : Pin 2 at 6.35V and pin 10 at 4.55V.  Others around 800mV.

IC5 : Pin 2 at 11.82V and pin 5 at 0V and pin 10 at 0V. Others at around 800mV  (EDITED)

All Transistors in the position of T19, Collectors at 15.46V except T24 at 0V

There are 14 Transistors in the T29 - T30 positions measuring inconsistent.  Notably T47 and T48 Emitters and Base at 0V or close to 0V.

Emitters all between 118mV and 124mV
Base all between 81mV and 90mV
Except:

T31/T32 : E 43mV.           B 52mV
T45/T46 : E 15mV/10mV  B 51mV
T49/T50 : E 10mV/15mV  B 50mV
T51/T52 : E 42mV/36mV  B 74mV
T59/T60 : E 17mV/12mV  B 49mV
T63/T64 : E 9mV.             B 44mV  (This one has blown LED's)

Diodes OK.

I hope that all makes sense.

[Rob Strand]

Ok.  Done some measurin'.  All measurements with the oscillator on full volume.  Turning the oscillator down does change some LED behaviour mainly the right channel LED's which have dimmed since the photo was taken. (?) Turning the oscillator down until the top LED shuts off brings back the brightness for the rest of the row.

Looks good.

That last point looks like a good clue.  It's pretty weird behaviour for a circuit which uses a current source (T29, T30).   The current source output output shouldn't change with the number of LEDs.  There's are faults which could cause it but having those faults on many channels makes me think it's something else.  I haven't had a lot of sleep so I can quite get my head around it.

4022 Outputs.
IC3 : Measures consistently around 800mV except pin 10 at 10V.

IC4 : Pin 2 at 6.35V and pin 10 at 4.55V.  Others around 800mV.

IC5 : Pin 3 at 11.82V and pin 7 at 0V. Others at around 800mV

All Transistors in the position of T19, Collectors at 15.46V except T24 at 0V

So those voltages are looking pretty good.  If I assume the +17V power rail is 16.53V then we can estimate
the voltages

FYI, you can see the numbers below.   The outputs which go to the "T19's" get loaded down a bit do the output drops to maybe 14.4V.


                               Unloaded    Loaded
V gate high                    16.35       14.4
Total states      18     
                   number of     
                   cycles on   expected V   expected V
Selects to T19     1              -          0.80
IC3 pin 10 ("7")   11            9.99   
IC4 pin 2 ("0")    7             6.36   
IC4 pin 10 ("7")   5             4.54   
IC5 pin2 ("0")     13           11.81   
         
         
"T19" Transistors are on 1 state in 18         
but the collector voltage is reversed     
'Vc=(17/18)*(+Vcc) + Vsat (1/18)         
+Vcc   16.35     
Vsat   1     
Vc   15.50

IC5 : Pin 3 at 11.82V and pin 7 at 0V.

All Transistors in the position of T19: except T24 at 0V

These look like something is wrong with IC5.   Unless you can see some soldering problems
or unintentional shorts on the tracks it's probably easier to replace it and recheck the voltages
have returned to a good state.

Pin 3 at 11.82V is going to cause all sorts issues so you need  to fix that first.

As for T24, if T24 is driven from IC5 pin 7 then the T24 at 0V can be explained because of that.

There are 14 Transistors in the T29 - T30 positions measuring inconsistent.  Notably T47 and T48 Emitters and Base at 0V or close to 0V.

Emitters all between 118mV and 124mV
Base all between 81mV and 90mV
Except:

T31/T32 : E 43mV.           B 52mV
T45/T46 : E 15mV/10mV  B 51mV
T49/T50 : E 10mV/15mV  B 50mV
T51/T52 : E 42mV/36mV  B 74mV
T59/T60 : E 17mV/12mV  B 49mV
T63/T64 : E 9mV.             B 44mV  (This one has blown LED's)

it's possible one or both of those transistors is fried.

You can check the transistors by measuring from the base to the collector and the base to the emitter using the diode tester.  You would expect about 0.6 to 0.7V, like you get with a diode.  It's also with checking you aren't getting 0V across the collector and emitter.    You can see how it's done here,
https://electronicsarea.com/how-to-test-diodes-and-transistors/

When you test in circuit you might some odd numbers.  What I suggest is do the test on a couple of good channels to see what measurement to expect.

Another thing is because the bases are connected on T29/T30  when one is fried it will corrupt the measurements on the other one.  Usually the part which the lowest measurement is the faulty one, so replace that one first and re-check. 

Once you replace the parts you know are stuffed recheck the voltages and see which ones have come good and which one haven't. Also which bands are now working.   If you still have a few bad bands  maybe replace the other transistor of the T29/T30 pairs you previously replace - it's possible both got damaged.

[AS74]

Somehow I've typed the wrong thing.

IC5 : Pin 2 is 11.82V and pin 5 is at 0V and pin 10 at 0V

[Rob Strand]

Somehow I've typed the wrong thing.

IC5 : Pin 2 is 11.82V and pin 5 is at 0V and pin 10 at 0V

That makes more sense.  Pin 5 will have a *very* short blip which you will see on an oscilloscope but it will be pretty much zero as far as the multimeter is concerned.

So maybe look at the current source transistors.

Do you know what pin of the CD4022 the T24 transistor wires back to?   You would have to trace T24's base back through the "R27" resistor then to the CD4022.   T24 could also be fried.

[AS74]

T24 goes to IC4's pin 7.

T24 tests ok with diode test.

[Rob Strand]

T24 goes to IC4's pin 7.

T24 tests ok with diode test.

Probably worth checking "R45" or "R44" aren't open or have a soldering issue.

[AS74]

I found a lifted trace on T24.  All good now.

Everything else still the same.

[Rob Strand]

Everything else still the same.

[Click to Enlarge]


I translated your measurements by multiplying them by 18 (the number of channels).
The idea is to look at the actual base voltage and the base-emitter voltage of the transistor.

In the good case the base should sit at about 2.2V (2200mV) and the emitter sits 650mV below that at 1500mV.
Those voltages looks reasonable.

In all the bad cases, except T31/T32, we are seeing a normal looking VBE voltage but an abnormal Base voltage.
One way to interpret that is the collector of one of the transistors is disconnected from the LEDs/Diodes.   That
gives rise to the hogging issue I mentioned in Reply #8.

So just pick one of those faulty channels (not T31/T32) and check the continuity from the collector to the LED and diode.
If you find something try to fix it and see if the base and emitter voltages on that channel now agrees with the good case.
Also you might as well go through all the others the same way.

As for T31/T32, I'm not 100% sure what's going on.  It could be both collectors have a connection issue, but the VBE doesn't look like that's the case.  It could be a broken trace or bad solder joint.   Check all the traces around the resistors and 3xtransistors.  It could also be one or both the transistors has failed.   You might even check the resistors.

[AS74]

I replaced the known non working LED's in the last row.  That row is now working.

I've checked Collector to diodes/LED's and all good.

I'm going to recheck for more non working LED's in the problem rows.  Maybe something happened since I checked them last time.