BK Precision 2160A repair

[ZaUcY]

So with some wires boged (is that how you spell it Dave?) on the pre-amplifier board for remote runtime, the pre-amplifier board seems to be doing it's job.  The AC/GND/DC coupling switched are good.  I checked the input's to U102/202 and the signal de-amplified as I increased the V/div knob and amplified as I decreased the V/div knob (which I believe is correct).  I also tested the signals coming off 16 wire bundle at the end of the board (which leads to the mid-amp board) and every part of the vertical amp switched seems to work correctly on both channels (5X magnification, all V/div settings{atleast relative to each other}).  So I just soldered everything back together and re-installed the pre-amp board.  I was going to measure on the diodes, D308/310/304/306, but one of my probes broke and it's hard to get a good measurement.  I have ordered a couple more (cheap Hanteks)probes that should hopefully do the job for now.

The main reason why I'm posting before I have tried everything you guys have suggested, is because I managed to get the scope to do something interesting.  I turned on the "CHOP" setting in the scope, which makes channel 1 and 2 sweep simultaneously, and it seemed to show the bottom end of the square wave along with the top end (shown in picture).  This is with only one channel plugged into the CAL signal, but I can get this to show with either channel 1 or channel two.  The only difference is the trigger needs to be set to the correct channel, and the position switches need to be inverted when you switch channels.  I have re-tried to recreate this with "CHOP" off, but I cannot.  Does this give any more clues?

[ZaUcY]

I also want to point out that I can only get both of these parts of the wave to show when V/div knob is set to 50mV/div, no other setting will show both.

[tautech]

So with some wires boged (is that how you spell it Dave?) on the pre-amplifier board for remote runtime, the pre-amplifier board seems to be doing it's job.  The AC/GND/DC coupling switched are good.  I checked the input's to U102/202 and the signal de-amplified as I increased the V/div knob and amplified as I decreased the V/div knob (which I believe is correct).  I also tested the signals coming off 16 wire bundle at the end of the board (which leads to the mid-amp board) and every part of the vertical amp switched seems to work correctly on both channels (5X magnification, all V/div settings{atleast relative to each other}).  So I just soldered everything back together and re-installed the pre-amp board.  I was going to measure on the diodes, D308/310/304/306, but one of my probes broke and it's hard to get a good measurement.  I have ordered a couple more (cheap Hanteks)probes that should hopefully do the job for now.

The main reason why I'm posting before I have tried everything you guys have suggested, is because I managed to get the scope to do something interesting.  I turned on the "CHOP" setting in the scope, which makes channel 1 and 2 sweep simultaneously, and it seemed to show the bottom end of the square wave along with the top end (shown in picture).  This is with only one channel plugged into the CAL signal, but I can get this to show with either channel 1 or channel two.  The only difference is the trigger needs to be set to the correct channel, and the position switches need to be inverted when you switch channels.  I have re-tried to recreate this with "CHOP" off, but I cannot.  Does this give any more clues?

The trigger always needs to be set to the correct channel!
Have you experimented with the Trigger Level?

It seems by your tests that the pre-amp and attenuators are working correctly. 

[What_NZ]

If you want to try a little experiment.

Desolder and lift out of the PCB, one end of diodes D304, D305, D308 & D310. Then recheck the display using Channel 1 with the CAL signal.

Then can you measure the voltage on U301 pins 5 & 6 with the Chop function both Off and then On

[alanb]

This may be a stupid question but do we know that the probe is ok?

[What_NZ]

This may be a stupid question but do we know that the probe is ok?

I couldn't think of any way a passive probe could cause a fault like that and post #25 seems to confirm that now.

[ZaUcY]

This may be a stupid question but do we know that the probe is ok?

I was using the probe on my USB scope and getting a clean signal.  For some reason my other prob recently died   China made crap.  The probes I just ordered are also China made (crap?) but they are atleast branded.

I was getting an open circuit on the bad probes positive lead compared to 228ohms on the good probe's positive lead, in 1X setting.

[ZaUcY]

Hey guys,

Again sorry for I have not completed everyone's recommended tasks yet.  However, I have created another video (Shorter) showing something that I feel needed to be posted right away:

https://www.youtube.com/watch?v=uZ6FOrh_naI&feature=youtu.be

Not really sure what is going on, I have a feeling this is just a coincidence.

If you want to try a little experiment.

Desolder and lift out of the PCB, one end of diodes D304, D305, D308 & D310. Then recheck the display using Channel 1 with the CAL signal.

Then can you measure the voltage on U301 pins 5 & 6 with the Chop function both Off and then On

This is my next step

I really don't think the problem is on that PCB you would be better off quickly checking this (from my other post)
" You could also measure (with the power off) all the low value resistors on page (4) "

You can post pictures of that PCB if you want to.

This is my next next step

Thanks

[What_NZ]

After looking at your video, I'm starting to think there is something fishy with U301, half is used on page (3) and the other half on page (6). Please check U301 as per the previous post #28.

[ZaUcY]

After looking at your video, I'm starting to think there is something fishy with U301, half is used on page (3) and the other half on page (6). Please check U301 as per the previous post #28.

OK, so I have tested what you have suggested.

I de-soldered one end of D304, 306, 308, and 310.  When I tried to display a wave on channel 1, I could not get anything.  The only beam I saw was an extremely faint and out of focus vertical beam; nothing I did would sharpen the beam.  This beam would begin to sweep from left to right as I increased the time/div knob, like how a signal will slowly sweep across the screen as you increases said knob.

With the same diodes still disconnected, I continued to measure U301.

Power:
Gnd looks good, no voltage drop between the chip Gnd and chassis Gnd.  Vcc looks good too, in the schematics the chip's Vcc is connected to 5V in series with a 51ohm resistor.  Measured values are 4.93V at 5V rail and 4.73V at chip's Vcc.  According to datasheet, Vcc can go as low as 4.5V.

U301, CHOP half:
I have attached a sort of truth table for how pin 5 and 6 (Q and Q') were behaving when switching chop and the vertical mode switch.  the on voltages (output high voltages) I was seeing was ranging from about 4.3V to 4.7V, which looking at the datasheet seems to be fine.  However, the datasheet claims that at Vcc = min, the V output high should be 2.4V-3.4V, so the chip is out of spec when vertical mode = dual.  There is one instance when the voltage output can be out of spec, according to the data sheet, when pre' and clr' are both off.  However when my voltage outputs are out of spec I'm measuring pre'=4.76V=clr'.  Something strange.......  Oh, turns out it's a square wave input and output at this setting with the correct output voltage.

All the Off voltage (output low voltages) values were fine too, about 230 mV.

Another thing I just realized, how is Q and Q' on at the same time in a single flip-flop?  Looking at the truth table on the first page shows this possible, I measured pre' and clr' both off, so this is correct.

U301, Trigger half:
I was getting a couple of clks for the trigger's half of the chip, depending a couple siwtches.  I have posted another truth table setup, and snapshots of clks based the switches.

Also is a truth table for how the Q and Q' were behaving.  Everything seemed to be a DC signal and straight forward.

Datasheet:
http://www.ti.com/lit/ds/symlink/sn74ls74a.pdf

Measuring this chip is getting a little cumbersome now, but I think it is working correctly.  If the community is not convinced, I can probably throw it on a breadboard in a more controlled environment.  I also could do a thorough truth table, like in the datasheet, if necessary?

[What_NZ]

Just a quickie, is this a typo (in BOLD)?

<Quote> I de-soldered one end of D304, 306, 308, and 310.

As it should be D305. What I was trying to achieve is isolate Channel one from Channel 2 and also the operation of the CHOP (for the want of a better name) circuit.

I just re-read the whole thread cause I lost my train of thinking on this fault.
Back in Post #15 you confirmed the Component Test Function didn't work, this still is the biggest clue and rules out all circuitry to the left of S301-A on page (3). So don't worry about D304, D305, 308, and 310 or that half of U301. Sorry for that diversion.

It still would be good to see the waveforms (using the CAL signal) at the Anodes of (CH1) D304, D306 and then move probe to (CH2) and do the same on D308 & D310 on page (2)
Those signals should remain similar all the way to the Emitter of Q404 & Q405

[ZaUcY]

Oh, sorry!  I assumed you made a typo with D305, seeing how D304 and D306 correlated the same way as D308 and D310; I will fix that.  Do you recommend trying to re-test U301 after I make the correction?  (U301 seems to be an isolated digital circuit to me.)

So I have gotten around to testing page 4's resistors.  It was a little tricky, because the CRT is laying right over this circuit.  Thankfully the schematic book board layout matched the on in my scope, I was able to probe from the solder side using the schematics.  Some of these values MAY be off due to components in parallel, I haven't looked too closely to that yet.  An example is R433 and R434, I cannot measure those (unless I de-solder them) due to being in parallel with inductors :/  I have highlighted the resistances that seem to be way off, I have also included the percent error in ideal value and measured value.

Hope this will help find something!

[What_NZ]

Sorry I seem to be making a few mistakes swapping between the circuits and the forum, my apologies.

The diodes to lift one end of are D303, 305, 308, and 310.
The reason for doing that was to isolate Channel 1 from Channel 2 and also from the operation of the CHOP (for the want of a better name) circuit formed by U301. Just seeing if you get a correct display would be enough when using Channel 1 and the CAL signal.

I don't think rechecking of U301 is necessary.

Oh did you check R446 on page (4). Also (but maybe nothing) it is interesting that R419 is about 10K different than the same resistor (R422) in the identical circuit below. Could you recheck R419 and then again but this time reverse the multimeter probes while checking it. If it still measures 10K lower could you check Q407 & Q408 for leakage.

It still would be good to see the waveforms (using the CAL signal) at the Anodes of (CH1) D304, D306
Those signals should remain similar all the way to the Emitters of Q404 & Q405

[onlooker]

I am not sure if it is already asked,  did the scope show the flat trace when nothing connected to the input bnc?

If not, a dc level check using a multi-meter along the signal path should already be enough to do an initial diagnosing. At the very end y+ and y- should have a voltage difference following the vertical position knob centered around zero volt difference.

I would also give all the  knobs and buttons some hard turns/toggles to make sure good contacts.

[ZaUcY]

The diodes to lift one end of are D303, 305, 308, and 310.
The reason for doing that was to isolate Channel 1 from Channel 2 and also from the operation of the CHOP (for the want of a better name) circuit formed by U301. Just seeing if you get a correct display would be enough when using Channel 1 and the CAL signal.

This actually isolates Channel 2, the schematics are confusing.  And I now only have the listed diodes disconnected on one end.

Oh did you check R446 on page (4). Also (but maybe nothing) it is interesting that R419 is about 10K different than the same resistor (R422) in the identical circuit below. Could you recheck R419 and then again but this time reverse the multimeter probes while checking it. If it still measures 10K lower could you check Q407 & Q408 for leakage.

I must have missed R446, I measured it and added it to the list.  I have re-measured R419 and I am still getting values that are pretty far off.  If I measure the resistor in one polarity the DMM settles at about 22.5K and in the other polarity the DMM settles at about 25.8K.  The weird thing is the DMM will take about 3 times longer to settle on a value than almost every other resistor.  Here is some information I have on the transistors:

PNP:
Q407 HFE = 5
Q410 HFE = 64

NPN:
Q408 HFE = 20
Q409 HFE = 21

I am not sure how to check for leakage, but it looks like there is something fishy with the PNPs.

It still would be good to see the waveforms (using the CAL signal) at the Anodes of (CH1) D304, D306
Those signals should remain similar all the way to the Emitters of Q404 & Q405

So I have posted the input and output waveforms for the mid-amp in channel 2. 

Negative terminal
Input: Pin14In
Output: D306Anode

Positive terminal
Input: Pin13In
Output: D304Anode
(These screen shots were taken with V/div knob = 5mV/div, most amplified signal)

The signal definitely looks messed up at the diode anodes, like runt pulses maybe.  The other really off thing about the signal at the diode anodes is how the signals behave with V/div knob settings.  The signal output, past the pre-amp board, will de-amplify as I increase the V/div knob and amplify as I decrease the knob; this is correct.  Measuring the signal at the anode of the diode will begin to de-amplify as expected, but will gradually add more DC offset, and at the same V/div setting when the signal disappears off my CRT screen, the signal at the anode becomes a seemingly pure DC voltage at about 2.8V.

So if you are saying I should see a similar signal, and behaving the same as after the pre-amp board, then my problem is probably somewhere in the mid-amp circuit before D304/306.  However, like you have mentioned earlier, it is unlikely both channels get identical problems in isolated circuits. 

I am not sure if it is already asked,  did the scope show the flat trace when nothing connected to the input bnc?

If not, a dc level check using a multi-meter along the signal path should already be enough to do an initial diagnosing. At the very end y+ and y- should have a voltage difference following the vertical position knob centered around zero volt difference.

I can't get a flat GND line on the screen whether or not anything is connected or disconnected.

If you look in the attached screen shots of this post, there is a DC offset in the signal after the pre and mid amplification, but I haven't checked all the way up to the CRT yet.

[tautech]

R419, is this measured out of circuit, if so replace.
Q407 replace

[ZaUcY]

R419, is this measured out of circuit, if so replace.
Q407 replace

It is measured in circuit.  Q407 and Q408 had some burnt flux around the solder, as if it had already been replaced at some point.  So it definitely sounds like Q407 is bad?

[onlooker]

I can't get a flat GND line on the screen whether or not anything is connected or disconnected.

If you do not see the GND line, that should be the 1st priority to work on since it is easier and no need for cal-signal.

Assuming your plots are GND referencing,  then,  for the most part those plots are the wrong things to look at. Other than the 1st part of the pre amp, all other amp stages are differential amps of various types (with emitter or collector current mirror, class B (push-poll emitter follower),...). 

For differential amps, you need to check the voltage difference on the input pair and output pair of each stage. pin 13 and 14 are such a pair. I do not know the reason for the square wave yet. I guess they are due to the diode switches (D303, 305,...) are at working to alter channels for sending the input to the rest amp stags (But, in that case, the plot on D304 should be the one showing the effect??).

The right view of the ch1 input is (pin13) subtracting (pin14) and (D304) subtracting (D306).  If I do this visual math, I do not see things that are visibly wrong.

What you need to have is a DMM hooked to such pair points, then,  varying the V pos dial and  observe whether the voltage on the DMM changes and whether the voltage can be adj to zero.

Since you can trigger properly, the likely place wrong is in the "Vert Final Amp".

More of such pair check points there are
 --collectors of Q402 and Q401,
 --emitters of Q405 and Q404,
 --Y+ and Y- (or collectors of Q407 and Q410). Be careful the 140VDC.
 
The schematics is confusing. It switches around ch1 ch2, chA, chB. You may need to correct my listing accordingly.

Did you check the 140 VDC supply?

 

 

[ZaUcY]

Did you check the 140 VDC supply?

Yes, I have right off the PSU and it is measuring 145V.

I will attempt in checking the differential voltage, with no input signal, on the differential pairs as you mentioned.

[onlooker]

And make sure it is on the single channel mode.

[What_NZ]

Oh that was an easy trap, well spotted. I suppose it became obvious during your testing. An easy mistake but I should have been more careful so CHA on page (2) is Channel 2 and CHB is Channel 1.

[tautech]

I must have missed R446, I measured it and added it to the list.  I have re-measured R419 and I am still getting values that are pretty far off.  If I measure the resistor in one polarity the DMM settles at about 22.5K and in the other polarity the DMM settles at about 25.8K.  The weird thing is the DMM will take about 3 times longer to settle on a value than almost every other resistor.  Here is some information I have on the transistors:

PNP:
Q407 HFE = 5
Q410 HFE = 64

NPN:
Q408 HFE = 20
Q409 HFE = 21

I am not sure how to check for leakage, but it looks like there is something fishy with the PNPs.

Seems that way to me too.
When you get measurements that you suspect as fishy, the only sure thing to do is pull the components and test them on the bench. How hard can that be. Once done they are eliminated from suspicion.
Grab the datasheets and check hFE is to spec.

This will not be the first scope I have seen vertical output stages fail.
There are a few areas in a CRT scope where the componentry is "stressed".

In order of my experience:
1 EHT supply
2 Vertical output
3 LV PSU

Mostly voltage is the killer and the age of the components.
You can with very careful selection substitute with higher spec components.

[What_NZ]

Actually I assumed you must have pulled the Transistors to measure the hFE? Most of those types should have a gain of more than 60 so I'm not sure of your tester or testing method (in/out of circuit?)

Also check that they are what is shown in the Service Manual especially if you suspect someone has been there before you. Toshiba are a big player in the 2SA, 2SB, 2SC, 2SD types so check the brand too.

I'm not sure if this is a good idea but there is a 2SA1360 in the Horizontal circuit page (9) Q658 I can't see any harm if you swapped them. But before you do, what do the others think? If the fault changed axis it least it would be an indication.

[tautech]

First hit on a Google search: http://www.amplimos.it/images/2SA1360.pdf
As remarked on in a previous post, this device is likely to be voltage stressed, especially if any surrounding component gives problems.
Why? It's absolute max voltage rating is 150 V and in is working in a 140 V area.
Take note of: Note 2: hFE classification O: 80 to 160, Y: 120 to 240 from the datasheet.
Identify which version you have, pull it and test.
If replacement is needed, it is likely you will have to do some minor overall gain or balance adjustment, but as far as I can tell they are provided for in earlier stages.
Any component that you suspect pull and test.
It seems you are very close to the problem, let this give you the energy to crack it. 

[ZaUcY]

Alright guys,  Sorry for such a late reply, I had to put my scope away for a couple days and just ended up taking a break on it.

R419, is this measured out of circuit, if so replace.

I have pulled R419 out of the circuit, and its measuring correctly at 32.69k (I have updated the resistor table, and have added one to this post too)


So I have pulled out Q407/Q408/Q409/Q410 and have measured all of the transistors separately.  Q407 was giving me weird results, so I measured all transistors as an NPN and PNP.  It looks like Q407 is bad, but you might also notice that Q408 and Q409 don't agree when they are the same transistor.

I have tried to look for data sheets for these part numbers and could only find on data sheet for C3953 made by fairchaild semi, found here:
http://media.digikey.com/pdf/Data%20Sheets/Fairchild%20PDFs/KSC3953.pdf

I could not find a data sheet for A1538 (Q407/Q410). Is this the right transistor for where the schematics label A1360?  Actually both part numbers don't fully agree with the schematics.  I am no expert in reading part numbers and understanding the patterns, so I have posted pictures of all the writing that appears on the transistors.

It might be a little blurry to read those part numbers so here is what it all says.

Q407/Q410:
top line: "A1538:"
bottom left corner: "E" (to label emitter)
bottom right corner: "6L"

Q408/Q409:
top line: "C3953:"
bottom left corner: "E"
bottom right corner: "1G"