HP 1715A CRT debug

[JKKDev]

Today I received a package from Farnell with all the elcos and some diodes for the PSU of my HP 1715A (one of the old elcos on the 5V rail shorted and nicely charred a pair of rectifying diodes). After replacing all the caps and diodes (and an hour or so to reinsert the PSU board (all the wires from the transformer and some more are soldered in the PSU board)) I was finally able to turn the puppy on only to be met by the next problem.

CRT lights up (intensity dial works) but there is no trace. I tried setting it up like it's written in the service manual (all the dial positions and stuff) but nothing changed. Beam finder appears to do nothing although one time I saw a glimpse of a beam disappearing off the top of the CRT when I pressed it.

I have checked all the test points on the PSU and all the voltages (5, -15, 15, 20 (unregulated at 22.5), 53 and 115) are present, within spec and no ripple present.

On a side note... I started the troubleshooting procedure and I got to step 5 where my A8TP2 was 4.4V and A8TP7 was 4.7V (eg. none of the stated) which should indicate a problem with either Hold Off, Sweep length or Trigger. Is this connected or is it a separate issue?

I have to note that I don't have any signal applied to any of the channels but I believe I should still get a trace.

What I'm looking for is a starting point to find the source of the problem as I've never dealt with CRTs before.

Thank you and have a nice day

[JKKDev]

UPADTE 1:
I think this is a large enough change that it deserves a new post.

After fiddling with the dials and buttons all of a sudden a large circle started to show. Focusing it into a dot it started responding to some of the dials (horizontal position mostly). It wasn't stable at all (the dot was wandering around at different speeds in both axis) and after a bit of trying to tame the damn thing it puffed into nothingness from which it came. I don't want to push it further as I'm not sure if I could be doing damage to some parts.

The most notable thing about it was that it remaind a dot all the time. There was no trace of a trace 

UPDATE 2:
I noticed that the dot appeared briefly while switching vertical divisions. I was able to produce a continuous dot by leaving the vertical div dial between clicks (I'll try and get some contact cleaner in there but I'll wait until I figure the problem out). It appears that the vertical axis is having problems as the dot is bouncing up and down all the time but this time there is no unexpected horizontal movement. At this point I want to turn my attention to the vertical deflection plates however I'd be alot more comfortable doing it if someone would give me a few pointer about how not to get killed

I should also note that all of this is in X-Y mode. Any other mode and there is no trace/dot.

UPDATE 3:
I found the vertical deflection plates connections. After measuring the voltages I am fairly certain that there is a problem in this area. The voltage on both plates is jumping all over the place from 30 to 50V (should be 40V on both). I do believe however that the problem is somewhere before the plates as the voltage difference does correspond to the deflection of the dot. I believe that I have a healthy CRT and something else is bad.

[JKKDev]

UPDATE 4:
After locating the vertical deflection plate connections I was able to use a jumper to common mode the two connections and the dot stabilized more or less in the center (there will have to be some adjustments later). This told me that the plates were working more or less correctly. When I moved one step further back up the output amplifier the voltages were stabile (without jumping the outputs) however there was a potential difference of about 2V (27V and 29V where it should be 28V on both). This is what I assume is causing the deflection from center when the plates are jumped.

The problem at this point is that the thing sitting between the two stages of the amplifier is a ceramic hybrid (U2). There is a diagram of what's inside but first I'll have to check the power pins (although I'm not confident that that's the problem )

UPDATE 5:
The potential difference before U2 was my mistake as I didn't center the position dial. Now it's within 0.4V (still not centered but managable).

[JKKDev]

UPDATE 6:
I was able to find a schematic of a retrofit for the A5U2. I now have to find suitable MOSFETs for the job. If anyone's got any suggestions please let me know (scope is 200MHz).

[dzseki]

That is a totaly false approach, what's in there is not even a mosfet, that is a two stage symmetrical sliding cascode circuit.

[dzseki]

Here is the schematic. It may seems easy to substitute, but the bandwidth of a cascode circuit is limited by the parasitic capacitance on the collector node. But since the circuit has considerable dissipation, this is not something you can easily overcome. Note how the original hybrid is constructed: the substrate is floating in the air, and the heat is conducted from sideways.

[JKKDev]

Hey guys,

sorry to revive a dead thread, but I just had a thought. Now like I said in the initial posts I have almost 0 working knowledge of CRT tech.

TLDR of the findings so far: U2 in the vertical axis amplifier died which is making the vertical deflection plate voltages jump all over the place.

My question:
Are the vertical axis deflection plates there just to move the trace up and down the display (in time base mode) or do they play a role in signal display as well?

My thought was to just jumper the U2 output resulting in the trace always being in the center of the display (also removing the X-Y funcionality or would it similarly just keep the origin in the center?). But then atleast I can figure out any other faults with this scope.

Would this work or am I just going to have to accept that this scope won't work again?

[dzseki]

You can jump U2 if you want.

The bias voltage of the deflection plates are not that critical, although if you can bias them to 24V it makes them very similar to normal working condition.

As for sourcing replacement vertical amplifier. I sourced one off e-bay. Although you won't find these directly. Instead I've watched out for auctions for run down instruments, those clearly needed repair efforts beyond economical scale. These machines are usualy in the US so sending them over makes them less economical. I've contacted a few seller with inquiry on selling specific parts out of the unit on a negotiated price. I made a quick guide how to remove safely the complete vertical amplifier board, luckily it is a 5 min job. One seller went into the deal. This is a gamble of course because the instruments are sold as-is, but in my case thiw worked out as the amplifier was good.

The following oscilloscopes all use the same vertical output amplifier: 1710 (not sure),1715, 1720, 1722, 1725, 1727