Tektronix 454A no horizontal problem

[belzrebuth]

Hi all,
I've got a Tek 454A with horizontal problems; seems like the display can't get past 1/3 of the screen.
The timebase settings work as expected; vertical control seems okay as well but the signal is limited to only 3 or 4 graticules.
The bandwidth switch seems broken it is always in the 20MHz range it can't stay at "normal" but that seems unrelated.
I have no clue what might be at fault here as I've never had to troubleshoot a similar issue.
I would appreciate some help with it.
See attached image for an indication of what's wrong with it.

[wasedadoc]

Looks like a problem with the horizontal drives to the CRT. Put the scope into XY mode and feed the same signal into both CH1 and CH2. If the horizontal drive circuits are OK you will get a straight line at 45 degrees and suspicion falls on the timebase circuits. If you get a non-straight line then the horizontal drives are suspect. Usually differential output to the two deflection plates. Possibly a fault on one side.

[Technotronic]

Could be low B+ to the flyback primary or the HOT has low gain.
Give the deflection board area a good eyeball for bad solder too

[belzrebuth]

While I'm able to get a signal on CH1 and CH2 in other modes I get nothing in XY mode..Maybe I'm doing something wrong?
Another thing that's puzzling is that the trace is somewhat doubling unless I'm playing with the sweep length knob.
Also, on the sweep length when I'm fully counterclockwise I get the same signal as if I'm full clockwise (full)..
I guess that's a bit weird..

[belzrebuth]

I had to crank intensity a fair bit to see the signals in XY.
Here's how 1kHz square fed on both channels looks like.

[wasedadoc]

Look like you are using the CAL output for the inputs.  A sine wave would be better - more likely to see some line between vertical points.  But the result of square wave input is suggesting that horizontal amp or output stage is faulty, not timebase.

[belzrebuth]

How would I go about troubleshooting that?I really have no experience in repairing test gear.
Is it repairable?
I was hoping that since it's discrete parts only it would be more repairable but it's pretty old still..

BTW 12, -12, and 75V rails are okay..

[colorburst]

The 454A is a fully discrete design, short of the CRT dying, everything is fixable in this scope.

Can you move the trace to the right side of the screen with the horizontal position control?

Carefully inspect the area around the connections to horizontal plates, make sure the wires are still attached and no breaks are visible inside the tube.

If you can, measure the output on the horizontal plates. You'll need a scope that can safely handle signal levels up to 150V with some safety margin. There should be a symmetric opposite polarity ramp on both plates, see figures in the manual.

Assuming this is a horizontal amp problem, check the 10.1k resistors in the output stage, R1073 and R1093, they run hot and often go high resistance.

I've also had transistors fail in this circuit, they're all socketed so you can swap them across to see if the problem follows. Be sure to put them back in their original positions once done.

[belzrebuth]

Τhat was my intention, just take photos of what goes where and test transistors one by one out of circuit since they're all socketed.

I will check the resistors you mentioned and post back!

No I can't move the trace to the right and if I try to do so it kind of compresses the signal up to a point (3rd graticule) and stays there. I can go left but not right past the 3rd or so square..

Those videos got me thinking I might have an issue as those stated there..
It seems that sweep issues can manifest as "compressed" trace..

[belzrebuth]

R1073 and R1093 are  located a bit differently than what the board layout suggests in the SM but they test not open.There is a series combination of another resistor in there I'm not sure which is which.
I attached a picture of the general area.
The glass is a bit foggy in that spot but I don't see anything too concerning.
I do have another scope; should I check the signals at the red and green wires going to the tube?

[colorburst]

R1073 and R1093 are in series with a 9.53k, so you're looking for ~20k across the string. They don't go completely open circuit but increase in resistance 10-100 fold. The green resistors in your picture can also be problematic, they likewise tend to drift high. Most can be tested without desoldering.

Yes, I would check the outputs on the red and green wires, this is to confirm that the issue is indeed in the horizontal amp and not e.g. a broken wire in the CRT. Don't remove or disturb connections at the tube unless absolutely necessary: make all measurements on the circuit board.

Since the position control can't overcome the problem, I don't think it's an issue with the sweep generators or their associated logic. You can verify you have a good linear sweep where it enters the amp at R1005. Instead, I think this is going to be a lot like the first problem in your 465 video, some resistor or transistor going bad in one half of the circuit.

The best test for transistors is the circuit itself, so I would test them by swapping the pairs. Eg. Q1066 with Q1086 and so on. If there is a fault in the pair, the issue would flip to the other side of the screen.

[belzrebuth]

I attached scope images of the red and green wires.
The red one has a squashed waveform; it's almost DC and sits at around 155V.
The green one has a sawtooth (ramp).
I also probed q1018 and q1020 which both have a ramp so the amplifier must be getting a correct signal and it might be getting lost down the line; I don't think the signal on the red wire going to the crt is correct.
I swapped q1038 and q1048, q1066 and q1086, q1018 and q1020, q1060 and q1055 and q1054 and q1053 but nothing changed.

[colorburst]

Good progress, we can be confident at this point that the problem is indeed in the horizontal amp. One plate stuck at 150V is the expected result given the symptoms.

Can you clarify what you mean by "q1060 and q1055 and q1054 and q1053"? If you have checked all nine pairs and haven't inadvertently mixed them up, that leaves the passive components.

It would be good to take some DC voltage readings and compare them with the schematic. This might provide useful clues.

However, in my experience, voltages could be wrong all over the circuit, before and after the failed stage. If that's your case, I would start methodically checking the passives, starting from the green resistors. You can pull transistors and disconnect connectors to isolate components as needed. It will take a couple of hours, but you'll find it.

[belzrebuth]

I forgot to swap q1078 and q1098, q1078 had 4.3V on its base which is wrong.
Sure enough swapping them transferred the problem to the other side..!
I'm thinking I need to find a suitable replacement for it and maybe we are close or even done..

edit :2n2369 is obsolete but I see plenty of those on eBay like this one
https://www.ebay.com/itm/324300946262?hash=item4b81d7bb56:g:n7wAAOSwLsFfZfBZ&amdata=enc%3AAQAIAAAAsFmDoJh3EqKdYskOzTG80MKvGO6Mh2fDRwI%2BuQRUykRSxstYn30ElNp3M%2Bf3vEgL7SV7SaLqjsUG3l0dUwAaGzjD%2F32d3MMWfvfO01EdybNvLtIO%2FBEIVJRWlwsutyPguS58BRxOl2f4MGD80N4GIzTezI4y6sg0ioS%2BLLCyzoZl4pOjuG8%2FdMGxa%2FH5K%2F7xF%2F2QFbgdNGezn5MiMXxUscEW5WhMdR2EL3zPuul1BJFI%7Ctkp%3ABk9SR7ix3eDtYg
but I don't like that it's selected..
Should I buy a few and then match it to the other one (q1098) or just a random one will work?

edit 2:
I also found these two :
https://gr.mouser.com/ProductDetail/Microchip-Technology/2N2369A?qs=TXMzd3F6Eyk8Gn%2F4bkeFcA%3D%3D
https://gr.mouser.com/ProductDetail/Central-Semiconductor/PN2369A-PBFREE?qs=l7cgNqFNU1hSZqEKwBR3GA%3D%3D
the plastic case one seems fairly easy to source..
I wonder if I can use something like a 2n3904 or similar jellybean part to verify that's the problem and then wait for the correct part.

[belzrebuth]

I've temporarily put a 2n2222 in there just to see if it fills the screen and voila!

[colorburst]

Congratulations! I don't see an issue with any of the linked "2369"-style transistors. Attached are the specifications for 151-0127-00. The key feature seems to be fT = 650MHz compared to the usual 500. I don't think it'll matter. Maybe you'll see some difference on the highest sweep speed, with or without 10x magnification.

[TimFox]

Mouser has 2N2369A  (in metal TO-18 package) from Microchip in stock at $3.96 in single quantity.

[belzrebuth]

Parts ordered! Thanks for all the help colorburst