Tek 2465 no horizontal sweep

[dan3460]

This is another project that I'm working on. I bought this scope with problems. I also bought a donor scope as I was looking for the attenuator to replace one on my bench scope.
This scope has no horizontal deflection, if I apply a signal to any of the channels you see the signal moving up and down in the same position making a vertical line.
When selecting a times division I can see one sweep going across and then return to the same position, if done at a very slow rate you see the point slowly going across. I had repaired scopes with horizontal problems before but on this one I declare my ignorance as to how the system works and where do you begging troubleshooting.
So far this is what I have done:
- Replaced Q154 and Q155, because they where smashed with leads bent out of shape.
- Replaced U650 from the donor scope, no changes.
- Replace U500 from the donor scope, no changes.
- replace U700 from the donor scope, no changes.
- Replace the front panel with donor scope, no changes.
After replacing the original part was put back on the scope.

I understand that the circuitry should wait for certain amount of time and release a sweep with or with triggering, how do I troubleshoot that. Any ideas?

Thanks guys.

[tautech]

You don't say if you've got the service manual or if it's a B model or not. 

http://elektrotanya.com/tektronix_2465.pdf/download.html
125 Mb, 317 pages.
Wait for "processing to change to "get manual"

Here's another handy Tek troubleshooting manual:
http://www.sphere.bc.ca/test/tek-parts/troubleshooting-scopes.pdf


So it would seem there is some connection to the CRT plates from what you describe.
I've not studied the manual as yet but there will be some specific setup to make DC values in the schematics valid....follow that path for now for the sweep generator, then the Horizontal output stages. Tek are usually good at providing waveform test points too, do you have another scope to examine them?
Simplistic explanation of the sweep is a triangular waveform of varying frequency (set in timebase by RC networks) in order to electrostatically drive/deflect the electron beam across the display. At slow settings this rising ramp sweep can be seen with a moving coil meter, usually DMM's are too slow.

[dan3460]

Thanks, I do have the manual, this is my 4th 2465 that i'm working on. I do have a few other scopes and I have been using them, I even open up my personal 2465 to check for signal that are not provided in the Tek waveform page. After I posted this I worked on it for a little bit more and I staring to suspect that U165C may be defective. The signal that I get on pin 9 is the same on both scopes, but I get the waveform #20 and #21 on my bench scope not only a flat line at about 3.5V on the sick scope.
The manual calls for a LM324 which I don't think will be hard to find, but changing it seems to be a pain on the posterior as the whole main board has to come off, so I would like to confirm my suspicions.

Here is a clip from the diagram showing the circuit in question, this feeds the display sequencer.

Thanks again.

[MarkL]

Q154 and Q155, although drawn in a non-typical fashion, form a current mirror.  The inverted output from Q154's collector is being fed back to the opamp's non-inverting input.

What voltages are you seeing around the pins of Q154 and Q155?  The emitters should be the same.

And what voltages/waveforms are you seeing on pins 8 and 10 on U165?  We can assume pin 9 is ok from your comparison.


If Q154 and Q155 were squashed, I would also inspect very carefully for a broken trace.  Hairline fractures can be difficult to see.  Trace integrity should be verified with an ohmmeter.

Do other things on the scope work ok, like XY mode, the horizontal position knob, and the on-screen readout?  I'm assuming "yes" since you're focusing on this portion of the circuit.

[dan3460]

Thanks Mark, the emitters of the two transistors are sitting at about the same voltage, around 4V.
I get a flat line on pins 8 and 10, don't remember exactly the voltage did not pay attention.
I did check the connections from the 2 transistors to the resistors and checked the bases of the 2 transistors to make sure that they were connected together.
I do have readout. I did not check the functionality of the X-Y position, but the dot responds to the position knob vertically and horizontally.
I will check the x-y position to see what happens.

[MarkL]

It's probably not necessary to test XY mode if readout and position is working.  Based on your previous posts it wasn't clear exactly how much of the horizontal was still alive.

Anyway, it would help if we knew the exact voltages down to 0.1V around the opamp and transistor pins.  I wanted to see if we could definitely point to the opamp before you go to the trouble of pulling it out.

[dan3460]

Thanks Mark, I will not going to be able to look at those voltages until Tuesday.

[dan3460]

Here are the voltages around the circuit:
U165- Pin 8 1.369V, Pin 9 1.329V, pin 10 1.359V
Q154 - Em 4.07V, Ba 3.454V, Co 1.359V
Q155 - Em 4.07V, Ba 3.454V, Co 67mV

Let me know what do you think.

[MarkL]

Nothing unequivocally says to me that U165 is dead, but it's not looking good.

If U165C was operating properly, pins 9 and 10 should be the same, or no different than 9mV which is the max input offset on an LM324.  Your readings show they are off by 30mV.

And if U165 is being presented with 30mV on the non-inverting input, the output should be swinging to the top rail.  U165 is being supplied with +/-5V, so the output should be no less than about +3.5V (+5V - Darlington output drop).  You're getting 1.359V, which also points to a sick U165.

However, because we're looking at a changing signal with a DMM, it could be throwing off the reading slightly.  Do you have a waveform on pin 10 also?  It will be the same as pin 9 if U165 is working.  And you should see an inverted version of the waveform on the output, pin 8 (although at a different magnitude).

Here's something else:

Waveforms 20 and 21 on HRR (U650 pin 10) are supposed to go between 0 and +3V.  That says Q155 supplies from 0mA to no more than (+5V - Vce,sat - 3V)/1.50k = 1.2mA (approx) to pull up HRR to +3V.

You're reading 4.07V on Q155's emitter, so that says we're already providing (+5V - 4.07V)/1.50k = 0.62mA.  Assuming HRR is a resistive input (and it might not be), the voltage should be around 0.62mA/1.2mA * 3V = 1.55V or more.  But you're getting 67mV on HRR.  This is not making sense unless I've missed something.

The current calculation is further confirmed by looking at the current in Q154, the other side of the mirror.  With your reading of 1.359V, the collector current is (1.359V - -1.25V)/4.42k = 0.59mA.  Close enough to 0.62mA.

I'm not bought into a dead LM324 yet because I can't explain the low voltage on HRR.  I'm suspicious of a short to ground on HRR.  Since you've already swapped U650, maybe it's C660?  Or something from the mechanical damage?

[dan3460]

Thanks Mark for the detailed information.
I did looked at the waveform on pin 10 and it was a flat line, with the volt/div to the max I could only see a little of ripple there, the same as pin 9.
I have ordered a bunch of LM324 but in the mean time I will look at your suggestion to look for a short on HRR.

Thanks again.

[MarkL]

I'm also noticing that you previously reported for #20 and #21 a "flat line at about 3.5V on the sick scope".  And in your most recent measurements it's now 67mV (Q155 collector)?

What I said about the opamp before is valid, but I made a bad assumption about the origination of the signals #20 and #21.  I should have looked beyond the schematic snippet, specifically what "Holdoff from U170-2" is doing.  It turns out that's an output from the DAC multiplexer.  Along with holding capacitor C180 it creates a DC voltage that varies with the holdoff setting.

In other words, there is no waveform similar to #20 and #21 on the opamp input pin 9.

The DAC voltage controls the amount of current used to charge up C660 which creates a linear ramp.  Although we don't know the details inside U650, it appears that when it reaches +3V, it triggers U650 to reset the capacitor to 0V again to create a sawtooth.

These pulses appear to be going to a counter inside U650 which counts a set number of cycles for a particular sweep speed (100 @ 50us/div, for example).  When the count runs out, the holdoff is over and the sweep starts.  Since the voltage controls the slew of the ramp, and therefore the frequency of the pulses, the voltage is actually controlling the length of the holdoff.

So, you should see a varying DC voltage on Q154's collector as you vary the holdoff.  I measured +1.36V at minimum holdoff, and -1.11V at maximum.  If you get this, U165 is ok.

It's also worth noting that the waveforms at #20 and #21 only appear when sweeps are occurring, such as with auto-trigger.  If there is no sweep, such as with single sweep or normal sweep with no trigger, HRR stays low.  I measured that level to be around 65mV, same as yours.

So that lack of the debug waveforms at #20 and #21 I would say is just a symptom.  I think there's something wrong with the trigger or sweep.

What other debug waveforms have you checked?

[dan3460]

Thanks Mark, I followed your advise I checked the voltage at the collector and varied the hold off and I got voltages very similar to yours. So I started to look in another area.
Started to check waveforms around U650, form 19 flat line at 0v. Form 22 and 23, similar to chart but much faster. It supposed to be 2.4ms between pulses I got 6.4uS at the same high of 5V. Same thing with 24 and 25.
Form 26, only on with x10 mag, got similar wave for but as above is much faster than described.
Then I came to form 33, this should show at pin 17 of U650 got a flat line at 0v. Started to trace this back, first to the jumpers, made sure that where in the same position as the working scope. Check both sides of the jumper there is no signal. This line goes into U500 pin 44. So I decided to look at form 28, got a very small distorted signal, tested the base of Q645 and I have very clear pulses. So here is something that is definitely wrong.
I checked the resistors around Q645, R645 checked ok at 5.78K, R643 checked ok at 75. R644 showed about 1/2 of the resistance or about 2.4k as did R646 with 2.2k. Q645 is a 2N3906 (which I have a ton), so I decided to pull the transistor and checked, it checked ok. Tonight I will lift on leg of the offending resistors and really check their resistance and replace if neccesary.
I'll keep posting.

[MarkL]

It's quite odd that you're getting pulses of 6.4us on #22 and #23.  Can you confirm you have the sweep set to 200us/div?  And you also have AUTO TRIGGER set?

Waveform #33 is the A trigger, which goes low to indicate a trigger has occurred.  What do you see for waveform #29, which is supposed to remain low until the sweep has completed?  (This is from the Theory section, "A/B Trigger".)


It might be useful to step back for a moment and see if we can characterize the problem a little better.  First, I would try to get the scope to trigger at all.  You said in your opening post it would sweep one time when the horizontal was changed, so some things are working.

If you set a slow sweep rate, say 10ms/div, trigger to NORMAL, turn the brightness up, and then move the trigger level from one extreme to the other, can you get a single sweep to occur?

How about using LINE trigger input and then moving the trigger level from one extreme to the other?

If this scope has no options, are both P101A and P101B in place?  It doesn't seem to matter on my 2465, but maybe it does on yours since there's two versions of the schematic.

And, by the way, what serial number do you have so I can be sure I'm looking at the right schematic?

[dan3460]

Thanks Mark again, I have a digital scope and used the cursors to measure the periods. I will measure the same thing on my working scope just to make sure.
I will follow you instructions when I'm back home this afternoon.

[MarkL]

I found more jumpers that should be installed if you don't have options:

  P100: Pins 1 and 2, pins 9 and 10 if no TV option.
  P101: Pins 1 and 3, pins 6 and 8 if no CTT option.
  P102: Pins 3 and 4, pins 7 and 8 if no TV or CTT option.

This is from the CTT/WR Option Service Manual (pg 5-11).


Regarding waveform #28 and Q645, that sounds unusual so let's take a look at that for a minute.

What signal levels do you see on the base of Q645?  And what levels on the emitter and collector?  Do you see the +14V on the other side of R645?  How about the -5V on the other side of R646?

Perhaps some posts of the waveforms you're seeing would be helpful.

I should point out that I was unable to see anything that looked like waveform #28.  At that test point I have a signal that's 0 to -150mV.  Maybe the manual is wrong, or I'm overlooking something in the test setup.  I have to look at it more to figure out what Q645 is supposed to be doing.  Unfortunately U500 is another black box, so we don't know what kind of an input "CHN5B" is.  There's only a glancing mention of it in table 3-2, and that it's used by the "Diagnostic Monitor".

Q645 could be another red herring.  I wouldn't burn a lot of time on it.

We need to figure out why it won't start the sweep.  What are the logic levels on pins 1, 2, 3 of U700?  These must all be low for a sweep to occur.  See below from theory section, pg 3-22.  You should find it very useful to read the whole section on "Display Sequencer, Triggers, and Sweeps" starting pg 3-17, if you haven't already done so.

Have you worked through the "Sweep Troubleshooting" flowchart?

[dan3460]

Wow Mark, still in the office but I will be in my shop in about an hour. I really appreciate your help. I will post the waveforms I'm getting so you can see. I will follow your advice, the only thing that I will do is to check the two resistors that I mentioned as there are bothering me.

[dan3460]

I checked the two resistors and they were ok.
Following your lead, checked the levels at pin 1,2 and 3 of U700. The AUXTRIG in pin 3 is always low and I think that is the way it should be.
THO in pin 1 is connected to U650 pin 13 and is always low.
TRIG in pin 2 is connected to U500 trough a series resistor and is always high (2.75V aprox)
I will be reading the sections you have indicated and I will try to follow the flow chart.
I will post.
Thanks

[MarkL]

Actually, 2.75V on pin 2 would be considered low.  Take a look at the range of waveform #29.  It goes from +3V to +4V.

2.75V is maybe a bit too low, but not outrageously so.  I get 3.3V to 4.1V.  Something to keep in mind.  I wouldn't chase it right now.

So, a sweep should occur, or has already occurred.  It would be interesting to look at SG-bar on pin 45, and at testpoint (waveform) #31.  According to the theory section, if SG-bar is low (#31 range = -0.2V to +2.6V), a sweep is in progress.

Please continue to work through the flowchart.  I'm just interested in poking around some of the inconsistencies you're finding along the way.

EDIT:  And also, were you able to get any sweeps to occur by moving the trigger level back and forth (described a couple of posts ago)?

[dan3460]

Thanks Mark, I did go trough the flow chart yesterday but this morning when I was about to post I realized that I missed one step of the flow chart. After it says "Press beam finder" did not test if with the beam finder pressed the dots move, I went trough the yes branch, then stuck low, then yes, then yes and yes. I replaced U700 using the one from my bench scope and replaced U900 from the parts scope. No changes. But I will repeat the flow chart process from the beginning.
One thing that I wanted to mention is that the trigd light on the front panel lights up when a signal is connected and goes off if I take the signal away. I think this means that the triggering circuit is working.
Yes you get one sweep when you move any of the controls including the level. Changing the volt/div also triggers a sweep. I can see a prefect square wave when quickly push up the mode leaver.
Will go over the flow chart this afternoon.

[dan3460]

Here is where I stand as of now. I did go over the flow chart and did not change anything. I have been following the theory of operation in how a sweep is triggered.

I checked the signal CCA (control clock A) on U500 pin 23, this is coming from U2596 pin 14. There is a series of pulses when the source selection is changed, see picture.  So I'm assuming this is correct.

The A trigger source is selected by inputs SR2A, SR1A and SR0A, this HIGH and LOW signals respond to de selection of the source. So this part checks out.

Now, as the theory says to initiate a trigger the hybrid compares the input level TLA U500 pin 13 to the channel selected in the previous step. The level for channel 1 is picked up from LR180 and I can see the input signal there, I also checked that there is a contact between LR180 and U500 pin 6.

So what I think is happening is that the input level TLA is wrong and U500 cannot compare them to output a trigger gate TGA to the A sweep IC.

Well, I was about to measure the change of TLA as you move the level knob and compare it with my bench scope, turn on the scope and I could  not hear any of the relays clicking. Activated the lever to change the coupling on each channel, no relays. I feel that the whole problem lays here with the main control. If this failed now it may have been putting the wrong TLA level.

Too late tonight, will continue tomorrow.

[MarkL]

You should see pulses on CCA (A TRIG CLK), CCB (B TRIG CLK), and the two clocks for the sweep chips too, A SWP CLK and B SWP CLK.  The data line for all of them is CONT DATA, on J512-1.  I would check that CONT DATA is doing something, and that all chips are getting a clock.  And make sure all signals are valid levels (these are all TTL), and all clock lines are idle high.

There's 8 pulses, so it looks like the processor is loading one byte into each of the chips' internal register when trigger and other settings are changed.  Maybe something is getting corrupted in this process.

Looking at TLA would be a good thing to verify (range = +/-1.0V).  But in your testing when you moved the trigger level back and forth you could get a sweep, and also you said the trigger light was on and responded to the presence of a signal.  So, I'm thinking it's not TLA.

But now you have a different problem to chase.  I agree that it could be a related control problem.  Time to go back and check all the supply levels on the processor board.  Are any of the front panel lights on?

Since you have a working scope of the same model, you could try swapping the processor board to see if that fixes the current and previous issue.  It's not too difficult to get that board out, and it might give you a fast track to at least focus on the right board.


EDIT: Fixed typo on signal name.

[dan3460]

Thanks Mark for bearing with me.
I was taking the A1 board form my bench scope when I saw your note. Switched the A1 and response is the same, so it is not the controller board. I check the voltages against the manual, all within the range.

 I wanted to see if the pins on U500 are receiving data as you suggested, previously I checked that they are receiving power. I took U500 of the circuit and got a message on the screen that one of the tests had failed.

I want to make sure that there is communication between the pin on the IC and the connector, do you know how the pins are numbered on those connected, talking about J512. My wife is calling me so I didn't have time to play that much.

[MarkL]

Below is the pin numbering for J512.  There should also be a small "1" next to pin 1, but you might have to unplug J521 to see it.  On other connectors there's sometimes an arrow for pin 1, and on other equipment you might find a different shape pad (like a square).  Another way to find pin 1 is to look for a stripe on the ribbon cable or an arrow on the housing, but it's not fool-proof since it relies on whoever assembled the cable.

On my 2465, the connector pins stick through the back of the housing, so it's easy to probe the signals.  Not all connectors are made this way.

Another trick, but not as easy, is to probe the end of the ribbon cable where it has been cut.  Usually the conductors are exposed enough to touch them with a small sharp probe (plus good eyesight and a steady hand).

Of course this doesn't say the signal is getting all the way to the destination.


But more importantly, are the relays clicking now on boot?  Any lights or error codes?  Is it making it through self-test?

[MarkL]

I just noticed that "CONT DATA", besides serially setting up the trigger and sweep chips, also feeds serial data to latches U140 and U150.  These control the attenuator relays and a number of other functions.

That's very interesting since you say the relays stopped clicking.

Included in U140's other functions is controlling the BDC (Bypass Delay Comparator) pin on the A & B sweep chips.  I don't know if a corrupted BDC hi/lo could cause a sweep failure but it's at least suspicious.

How about front panel LEDs?  CONT DATA loads them too.  Are they behaving ok?

[dan3460]

Thanks, there a lot of questions. I will try to answer them all:
- There is a little arrow indicating pin 1, my question was on the order of the pins (clockwise, counter clockwise) I discover that they go in order 1 on top, 2 on the bottom, 3 on top next to 1, etc. And you are correct is very easy to get signals there.
- I made notes on the signal of J521 on my working 2465 and compare them with the broken scope, they match. I was hoping for a signal that did not show on the connector indicating that I had a short on that pin but no luck.
- All the lights in the front panel lights up as supposed to. When you move the coupling selection leaver up and down, the 1Meg, ground, cd, 50ohm, the corresponding light is lit but there is no clicking of the relays. All the other lights are find. The only thing that I notice last night, before I was going to bed, is that the A/B trigger, shows "run" but it doesn't stay on the B trigger, did not check if the behavior is the same on the good scope.
- The only time that the scope has not passed the self test is when I have taken one of the hybrids to check the voltages inside the socket. Once the hybrid is back in place doesn't complaint at all. It does not click the relay on the first startup.

I was reading and deciphering the functioning of the relays and it is not very complicated, my suspicion is that somewhere I'm loosing signals on the main board. This evening I'm going to check all the voltages to all the chips on the main board to make sure that there are all working. I will do the same with data lines, making sure that they are connected to where the control board connects.