Tektronix 475 scope: repair required.

[Skywave]

Originally Posted by dave cox

Nope. I can't figure that upper driver.
Am I missing something or is the schematic incomplete ?
dave

Funny you should say that, Dave: I came the same conclusion.

Al.

[dave cox]

i think i've figured it out !!

The upper output tr is in constant current - about 3mA.
The lower is passing the same current.
So, there must be an extra ~11mA in R1207 from the input followers T1172/1174
Any current change will make the output swing - ie its a folded cascode.

How clever is that !!

[Skywave]

Hi Dave,

R1207? I don't see an R1207.
I see R1235, R1233, R1174 . . . moving to the 'left' on the cct. diag.
And an R1209 in the emitter of Q1204

Al.

[Red to black]

Hi Al,
Although not specific to this particular circuit, I have had similar problems with DC coupled PA amplifiers in the past.
I noticed on this scope diagram at least 3 feedback/stabilisation loops (DC wise) which could upset the biasing of the output stage.
With DC coupled amps, I found it quicker and easier to just replace all semiconductors (including diodes/zeners etc.) in one go, believe me it is far quicker than spending days getting tied in knots trying to follow the tortuous feed back paths.
At least one or more of your "loops" contains a preset/pot (check for mechanical probs first).

Just a different view for what it is worth.
Cheers,
Baz.

[Skywave]

And here I am - again

I've put the DAC11 overhaul aside pro tem - I discovered why it was very quiet - a broken valve-holder pin, mixer anode

And so to the Tek. 475; voltages as requested . . .

First of all, there does not appear to be a problem with the deflection cctry. to the LH plate (the 'lower' part on the cct. diag.). The collectors of Q1212 and Q1214 correctly sit at +56v.

The problem is in the deflection amp. to the RH plate, ('upper part' of cct. diag.); Q1202, Q1204, etc. Voltages under fault conditions follow:

Q1202 / Q1204 collectors at +4.3v (should be +52v)
Q1204 base = -4.9v ('should be' not determined, but this sounds correct)
Q1204 emitter = Q1172 collector = Q1174 collector = +1.0v (should be -5.4v)

Q1174 base = +1.0v (shoud be +0.6v)
Q1174 emitter = +2.0v (should be +1.3v)
R1172 / R1173 junction = 2.5v (should be = 2.5v; therefore, correct)

Q1172 base = +0.3v (should be -0.1v)

Q1234 base = R1237 / R1239 junction = +4.2v (should be +4.3v)
Q1234 collector = +2.5v (should be +1.4v)

The 'should be' is the 'nominal correct voltage'.

HTH

Al.

[Skywave]

Baz - thanks for your contribution.

Initially had two 475 'scopes on the bench: one good, one bad (the described fault). I swopped over all the relevant semi-conductors between the 2 'scopes. The fault did not 'go over'. I checked all pre-set pots: NFF

Your suggested technique is the one I always use when repairing d.c. coupled A.F. power amplifiers. I learnt that technique the hard way, having lost (read: destroyed) too many expensive O/P power transistors trying to be 'logical'.

Al.

[Red to black]

Hi again Al,
the upper and lower sections are cross connected in a couple of places though, via R1169, 1163 for example (sensing voltage drop for stabilisation purposes ?).
With this type of cicuit it is not immediately obvious (the biasing arrangement that is), for example it, could be sensing a voltage drop across a resistor and applying the biasing to a collector or an emitter even (common base mode ?).
sometimes there are "loops" within "loops".
I absolutely detested DC coupled audio gear for this very reason, the voltage readings made no sense and was a pointless exercise, one stage upsetting another and not always in the same direction polarity wise.
just my take on this type of circuit.
Cheers,
Baz.

[Red to black]

Just a quick thought Al,
If you have an identical scope, have you tried swapping the panel across (I have no idea how feasible this is !) just in case you are chasing a ghost.

Failing that, taking "empirical/relative" resistance readings from the two panels (signature readings).
Although not scientific, I have used this technique in the past when faced with the "defies logic" /"well and truly stuck" type faults , on this type of circuit (when you have one working channel), and it has paid dividends.

Like you I learned the hard way/destroyed expensive silicon on DC coupled amps .

Ps. I have had "trimmer caps" go leaky resistance wise.

Cheers,
Baz.

[glowinganode]

Originally Posted by Skywave

And so to the Tek. 475; voltages as requested . . .

Thanks.
Rob.

[Skywave]

Originally Posted by Red to black

If you have an identical scope, have you tried swapping the panel across (I have no idea how feasible this is !) just in case you are chasing a ghost.
Baz.

Rgeretfully, that's not feasible. On account of the mechanical construction of this 'scope, there is a very good chance that I will end up with two 'scopes that do not work!
Besides, there is so much circuitry on this main pcb that it is extremeely unlikely that any new info. would be revealed by doing this.

---------------

Originally Posted by Red to black

Failing that, taking "empirical/relative" resistance readings from the two panels (signature readings).
Baz.

Tried that (in the relevant areas): no significant discrepancies observed.

-----------------

Originally Posted by Red to black

Ps. I have had "trimmer caps" go leaky resistance wise.
Baz.

Yes - now that is possibility - albeit remote - there is one fitted in the 'appropriate area'.
But then what was that about "drowning men . . . ?"

-----------------

Thanks for your ideas.

Al.

[kvimbo1]

Dear Skywave,
Let me join the interesting discussion. I'll have my try in logic sequence, based on the information already given.
First, if the emitter of Q1204 is +1V , while its base is -4.9V, and the collector is at +4.3V, means that the transistor (if not defect, as been quoted) is absolutely OFF.
Second, the point where C-Q1204, C-Q1202, RHPlate, and C1235//R1236 are connected, is high impedance, and so the only path possible to pull down the potential of the RHPlate is C1235//R1236.
Third, The left side of C1235//R1236 is connected to the base of Q1234, which sits approximately at +5V-0.6V=4.4V.
Very close to the potential of RHPlate, isn't it?
This leads to a doubt for SC of C1235. But if it is OK, as declared before, the equity in potentials at both sides of R1236 leads to conclusion, that no current is flowing through.
Well, the other path that may contribute for a low voltage ( which is available through C1235//R1236) is open circuit at Q1202, or respectively OFF state.
As known from the discussion, Q1202 must be OK. Then what can switch it off?
I haven't seen information till now for the potentials at the base and emitter of Q1202. Maybe I missed it, but never the less, i'll investigate the possibilities.
There are few elements around Q1202, that can be the reason for the malfunction-
- If R102 is open, Q1202 will be OFF
- if R202 is open, Q1202 will be OFF
- If C1205 is leaking, Q1202 will be OFF, also will help for the +1V at Е-Q1204
- If C1201 is leaking, Q1202 will be ON, which eliminates it from suspicion
- If B-E of Q1202 is SC, than Q1202 will be OFF.
There are some other possibilities, that can be caused by some constructional things - the PCB, the dust, any liquid and so on.
Analyzing the above mentioned assumptions, may help in solving the problem.

N.B.
I beg to be excused for any errors while using the language, as I am from Bulgaria, South-East Europe.
Best wishes to all!

[Skywave]

Thank you, kvimbo1, for that contribution: plenty to think about and investigate.

However, there are two cct. references that you make which do not appear on my cct. diag. (the Tektronix manual) . . .
I don't have an R1236 in parallel with C1235; I have R1235 (100 k-ohms) in parallel with C1235.
R102? (A typo.?)

From memory:
Q1202 base = +104v.
Q1202 emitter = +108v.

Yes: C1235, if leaky or S/C, will cause the fault. I'll check it out now!

Aside: I had no trouble at all in understanding your English: it's good.

Al.

[Skywave]

Nope: not C1235: it isn't fitted! It's shown on the cct. diag. with an *. I can only deduce that this is an 'only fit if necessary' component.
Whilst there, I checked out R1235 (the 100 k-ohm R that is in parallel with it): it reads 100 k-ohm on my Fluke 77 DVM.

Anyway, it's late now: time for bed. There's always another day.

Al.

[kvimbo1]

Originally Posted by Skywave

...
From memory:
Q1202 base = +104v.
Q1202 emitter = +108v.
...

Al, if there are 4V at the BE junction, Q 1202 is not OK. But you must be sure in your Voltmeter - if it has enough high input impedance. Sometimes a relatively low Ri may cause errors in tne reading.
Sorry, I saw your DMM must be OK.
If after all Q1202 measures OK, the problem is somewhere around it ...

Tony.

PP. Check also the pin positions, as You wrote, exchanges of the components have been made. If the B and E were reversed, such voltage can be obtained.

[dave cox]

Al, yup i meant R1209

And there is far too much current in it.
If you can figure out where that current is going i think you have your answer.

[kvimbo1]

Originally Posted by Skywave

...
R102? (A typo.?)
Al.

Sorry, R102-R202 > R1201, R1202
In the base and the emitter of Q1202!

My keyboard is crying for change ...

The opinion of Dave is correct, and coinciding with my thoughts till now.
The low voltage at RHp "lies" that Q1204 is ON, and Q1172 and Q1174 are shutting off Q1204, raising the potential of its emitter.
"Tutte le strade portano a Roma (All roads lead to Rome)..." in this case I hope to Q1202 or around it.

[dave cox]

PS I'm a bit puzzled by CR1173 - is it drawn the wrong way round ?

[kvimbo1]

No, its function is to limit the positive potential at E of Q1172 to approx. half of the 5V supply. The same is CR1163 at the other side.

PS. CR1183. My copy of the schematic is of low quality and I often make mistakes...

[Skywave]

Originally Posted by kvimbo1

Al, if there are 4V at the BE junction, Q 1202 is not OK.

Tony.

Never mind 'memory': I've just made the measurement . . .

Q1020/emitter = + 110v. The supply to R1202 = +110v. Therefore Q1202 is not passing any current.

Q1201/base = +104v. The supply to R1201 = +104v. But, of course, the base current will be very small.

Nevertheless, Q1202 is cut-off. Its collector is at +4.3v (should be +52v).

It seems that the real issue is why is there so much current through R1209? The excess current is not coming from the emitter of Q1204. (R1209 checks-out at the correct value). This leaves Q1174 & Q1172. But both of these check-out as O.K - by diode-check on DVM and by substitution with those from a functional 475. They are both low-power PNP. I even tried swopping over Q1174/Q1172 within the faulty 'scope: made no difference.
The 'other side' of R1175 is all a.c. coupled: there do not appear to be any s/c capacitors in that bit either, which would upset the d.c. conditions.

And that brings me right back to where i came in. I now have a substantial headache! ← No, that's not a wall: it's a Tek. 475!

Al.

[kvimbo1]

Al, Q1202 has OC of the BE junction. If you replace it with regular transistor, the problem may be solved. But! there is possibility to blow the new one again, if the mounting would not be correctly made. For instance, if anything makes SC - heat-sink, screw or anything else.
Just be careful with the installing.
The large current is due to the attempt of the circuit to regulate itself. Nothing special!

OOps: The OC may not be in the Q1202, but at the pcb . Check for crack in the tracks from R1201 to base and from E to R1202. Little to expect , but ...