Tek 2465 no horizontal sweep

[MarkL]

...my suspicion is that somewhere I'm loosing signals on the main board.

That may very well be.  Could also be an intermittent.  I would pay extra attention that logic signals are within acceptable TTL range: <=0.8V low, >=2.0V high.  CONT DATA is particularly suspect because it is shared amongst a bunch of chips and one of them could be killing it.

U140 cascaded to U150 forms a 16-bit shift register.  Both chips control the relays, but if U140 isn't receiving any data, U150 won't clock in any data either.  So, it's not a big surprise that none of the relays are working.

I'd love to know what U140 pin 2 is doing.  That's CONT DATA.  And the shift register clock on U140 pin 8, ATTN CLK.

If those look ok, you could check the outputs of U140 to see if they're changing, such as pins 3 and 4 when you change to and from AC Coupling.  They won't have a persistent state since other values get shifted through U140, but you should at least see something.  (The attenuator relays are latching relays, so they aren't powered except for changing state.)

Definitely check the power supply directly on the chip pins in the suspect areas, like U140 and U150.

Good luck!  Let us know what you find.

[dan3460]

First good report, decided to tacle the relays first, I started to probe voltages around U110,120,130,140,150 and Q130 and 131. All voltages appeared normal. Started to probe the signal ATTN STBR (attenuator strobe) and it showed around the whole circuit with the exception of the anode of CR130 and CR131. I checked resistance against chassis and CR131 showed about 5K and CR130 showed dead short. Pulled the other scope and both diodes anodes showed around 5k resistance.
Started to look around both lines, where is a series of diodes and only the anodes of some of them were grounded. I unscrew one side of the main board and the problem disappeared, looking with a flashlight I can see one of the screws that hold the shield for the tube making contact with the main board. I slipped a piece of rubber under the part and this problem is solved, now I get clicking of the relays.
I'll continue with my investigation of the other problem.

[tautech]

First good report, ..............

Started to look around both lines, where is a series of diodes and only the anodes of some of them were grounded. I unscrew one side of the main board and the problem disappeared, looking with a flashlight I can see one of the screws that hold the shield for the tube making contact with the main board. I slipped a piece of rubber under the part and this problem is solved, now I get clicking of the relays.
I'll continue with my investigation of the other problem.

So some muppet has been in there before.
Expect the unexpected, that is anything and everything.
Good find, keep your eyes peeled for similar stuff ups. 

[MarkL]

Oh well, an unrelated problem on the attenuator relays.  Glad you found it quickly.  Back to the original issue...

You've now been able to demonstrate that a single sweeps works.  I would look for reasons why things don't reset in preparation for the next trigger to occur.  Perhaps waveforms (or static levels) of #31, #32, #33 would be interesting.  Those are all related to completion of the current sweep and part of the trigger state machine feedback.

[dan3460]

I got some results. I put my scope at pin 9 U735 and move something on the front panel. I see the one sweep going on and one teeth of a saw tooth wave form.
So I went back and checked CONT DATA again, and I can see pulses but only in the order of around 70mV or less and weird looking. This is where I was going before I got side track with the other problem. I checked the same pin on my good scope and signal is very different nice defined pulses around 2.75V high, also the base line is around 0.25V. On the sick scope the base line is at 0V.
I pulled the connector out on the good scope and checked the resistance to found it showed open. I pushed the DMM probe hard to make sure that is open, could not get a reading. I did the same thing on the sick scope and the resistance on that pin 1 is about 5K. If this circuit requires a very high impedance that resistance can be taking the signal down to ground.
I going to investigate as there are several ICs that uses CONT DATA.
First picture sick scope, second good scope.

[dan3460]

A little set back, I know that I'm loosing the signal somewhere, but I was tracing the circuit to see where that 6.2K load was happening. There is a change on the schematics, the sick scope is an older model and it was a 6.2K resistor that in latter models was changed for a cap.
I rechecked the resistance at pin 1 and is very close to 6.2K.

I took all the hybrids that read CONT DATA out of the board and I got a good CONT DATA signal, a started to put one at the time and checking the signal, U500 good, U700 good and U900 good too  .

All hybrids in and now a good signal, this is like chasing ghosts, you think you have something then nothing. Puzzled as why this happens. resetting the hybrids cleaned the contacts?

Picture of the CONT DATA after resetting all hybrids. Will continue tomorrow.


.

[tautech]

A little set back, I know that I'm loosing the signal somewhere, but I was tracing the circuit to see where that 6.2K load was happening. There is a change on the schematics, the sick scope is an older model and it was a 6.2K resistor that in latter models was changed for a cap.
I rechecked the resistance at pin 1 and is very close to 6.2K.

I took all the hybrids that read CONT DATA out of the board and I got a good CONT DATA signal, a started to put one at the time and checking the signal, U500 good, U700 good and U900 good too  .

All hybrids in and now a good signal, this is like chasing ghosts, you think you have something then nothing. Puzzled as why this happens. resetting the hybrids cleaned the contacts?

Picture of the CONT DATA after resetting all hybrids. Will continue tomorrow.

Quite possibly. Early semiconductor scopes had all transistors and IC's socketed, it was part feature and part necessity as some transistor reliability was suspect at that time, decades ago.
It was normal to reseat all and in some way it's the same today with socketed IC's and interconnects as they're common problem areas.
Anyways, you're making progress, carry on. 

[dan3460]

Sadly this scope has just a few components that are socketed, but I have some progress (I think) to report.

Reading the theory the display sequencer is reset for the next sweep by a HIG signal from the A sweep gate SGA coming from U700 via U980A (a NOR gate with one pin set LOW). So I checked the path of SG, left side of R750 goes LOW briefly then returns HIGH. The right side of R750 stays LOW.

I lifted one leg of R751 and checked ok. I checked the resistance between pin 1 and 2 of U980 and I'm showing about 12ohms. I'm suspecting U980 but I'm going to check the other components that hang from pin 1 of U980. I did check all other components around this and they all checked ok.

I also checked on my working scope, the resistance between the pins 2 and 2 on U980 is about 2.4k.

[MarkL]

Yep, sounds like U980 is a prime suspect given the comparison with the good scope.  One other possibility is CR752, which could also implicate U950.  It's very unlikely one of the resistors.

I'll add a side note here that measuring ohms in-circuit can produce bad readings, even if the device is off.  The DMM's voltage that it uses for ohms could turn on PN junctions, and there's also the possibility of lingering voltages on power supply capacitors.

Measuring the ohms in both directions can usually eliminate this concern.  If it really is 12 ohms it should read the same in both directions.

Another way is to use the DMM diode test function test in both directions across the suspected component.  If it is a short, or almost a short, it should read the same value close to zero.


EDIT: I did a quick experiment and grounded pin 1 of U980.  I get the same symptoms you described in your first post.  The vertical is working, but the beam sits on left.

I can make the beam sweep once (most of the time) by changing settings, including moving the trigger level.  I couldn't quite replicate your behavior with the trigger light; once it was on it stayed on.

It could still be CR752, but there would have to be a second failure too, which is more unlikely.

My bet would be on U980 at this point.  Check its power pins to make sure.

[dan3460]

I did check the resistors around U980 and all check out I was aware of diodes or transistors messing up with your measurements so I always measure both ways. I also had checked CR752 with the diode mode on my DMM and checks ok.
Thanks Mark for your experiments, I think that confirm U980 been the prime suspect. I have ordered some ICs they should be coming early next week, in the mean time I will pull the main board off so I can replace the IC. I have not read the manual yet but it looks like it is not that difficult to remove.

[dan3460]

The ICs showed up early this week but did not had a chance to play with this. I replace the chip and  .
Checked the operation of the whole scope and looks good.

Thanks Mark very much for your help and taking the time to do some experiments for me.

Now I will do a complete calibration.

Cheers.

[tggzzz]

If you are replacing U800, then I strongly recommend you look at the recent thread "Likely U800 failure, ANY suggestions?" on the Yahoo TekScopes group.

It might help you avoid wrecking the replacement.

[dan3460]

I'm not replacing U800, is working just fine.

[MarkL]

Excellent!  Good luck with the cal!