Tektronix 2230 Oscilloscope Repair: Part 1

Another eBay find - a Tektronix 2230 oscilloscope with a strange fault symptom! Also appears someone had a go at this before. Can I fix this one? Not looking good so far...
Part 2 here • Tektronix 2230 Oscillo...
0:00 - Overview
6:00 - Someone has had a go at this!
6:30 - First impressions/visual inspection
9:18 - Storage board lift instructions
11:18 - dodgey soldering and desoldered jumper wires!
14:00 - Attenuator shield removal/further visual inspection
17:44 - Broken knob for Channel 2 attenuator revealed
20:08 - Faulty Cursor select pot/leads ripped
21:37 - visual inspection of power supply section
23:55 - power supply voltage rail testing
28:11 - cursor pot wiring repaired
30:07 - jumper wires reconnected on CH1 & CH2 logic boards
31:16 - verifying ‘probe adj’ operation w/ external scope
35:04 - Isolation of analog circuitry from digital/conclusion

[@NickNorton]

That is a nice Scope. It is old, but very worthy of repair. If you own and use a 2230 you are a clever dude.

[@andymouse]

There are loads of channels out there like 'Buy it Fix it' who buy broke stuff and mend it and they are a good source of practical fault finding techniques, the stuff you can't learn from books. Another good one is 'Learn Electronics Repair' or LER and 'Curious Mark' or 'The Signal Path' there are tons. Good luck with this puppy I shall watch with interest, I think your gonna have to break this down into smaller parts or end up down so many rabbit holes we will lose you for good ! you know how to eat an Elephant ?...one bite at a time....cheers

[@fmashockie]

Thanks as always for your feedback! Yep this one is definitely going to take time. Already working on part 2 with new issues revealed 😂

[@mirkobruhn9841]

This Oscilloscope has no deflection plates, just a magnetic deflection. So it must be fully digital connected to an old fashioned crt monitor. I guess, there will be very little analog things.

[@fmashockie]

not necessarily. As I show in the video, the service manual provides a procedure for isolating the analog section from the digital. The digital portion of the unit is mainly for storage of captured waveforms.

[@thetechgenie7374]

It is a true analog scope, the digital board is multiplex in with a multiplexing IC and a few diodes. the two digital logic board as he refer to them as on the vertical attenuator is just a set of switches with passives on it and use the same cam gear to switch vertical. You literally can isolate the whole top digital logic board from scope and scope will work in fully analog mode. It pretty much two scopes in one, the digital works independently pretty much for storage

[@thanhhuynh272]

Oh, this one looks like a bit of a “dozie”. All the usual culprits, power supply, EHT supply, tube Z-Axis amp and intensity DC restorer seem O.K. As does the timebase. Note how it gets dimmer a slower settings, this is normal because a slow moving spot is bloody bright they turn the brightness down internally.
The cal square wave connector is not marked, “1KHz” probably for a similar reason, because its frequency changes with the timebase setting to maintain one cycle per two divisions on the graticule. I am working on a Tek 2445 here, and although quite different inside, its cal out does indeed track the timebase setting.
I also have the “baby brother” of this CRO here, the 2213, its power supply and vertical inputs/attenuators look very similar. I see that bloody Rifa X2 under the IEC connector and, If I recall correctly it is fitted before the power switch. On just 120v RMS it will probably hold out for quite some time….here in 240v RMS Australia thay tend to hang on for less time. The 2445 emitted a jet of smoke withinn 90 seconds of first turn-on and I got to them in the 2213 well before the smoke release….but did take them, (the dodgey Rifas) out into the yard, on the secondary of a 240-500V transformer preceeded by my Variac and 500w worth of series incandescent globes….still, with 750v, (the 500v secondary in-phased on top of the 240v primary they still held out for about two monutes…but when they went, what a jet of destructive distillates!, a quick blast with the MAP gas torch and we had a ten foot flame for a few seconds! Us Aussies have to “extract all the fun” from them before they head to the bin!
Anyhow, that is another story…ba k to your 2230. I think the squares on the screen are a test mode for screen geometry and it looks like it has somehow frozen stuck on this test. I went to the Tekwiki and got the manual and saw the bit where it tells how to isolate the digital and analog parts of the ‘scope…glad you found that.
Here’s what I see…the previous owner was “testing” the shift controls and cursor control out of circuit which may go part way to explaining why they were disconected like that. The way screws were just put half back in and one was rattling around loose inside tells me that thay got so exasporated with it, they just wanted to be rid of it and dump the problem in someone else’s lap. We might respect such nice gear, but thise who can’t tell an Oscilloscope from a TV couldebt care less. A lot of owners of this sort of gear are governmet departments, schools or corporates and the users of the gear are not its owners, so they don’t give a damn if they missuse or damage the gear as it ain’t their problem….and generally such gear is/was far beyond the budgets of any individual to be able to own and cherish. But now, as the gear gets older and somewhat either more fault-prone or even, like your unit there, fault burdened..they become cheap enough for individuals like us to “reach” them, fix them and look after them.
The fact nothing is coming through on either channel would lead me to test the attenuators and take a close look at the Y-Amp input j-FETs….if is was from a university, students could have probed the mains with it whilst on 5mv per devision. Just use the HP 54600 to go anomg the signal path from the input jacks to the deflection plates…it is very likely the signals loop via those lonks you had to fit when isolating the digital section. I would guess from gut instinct the break in the signal path is down near the inputs. You could try injecting signals at those loops (to the input side only) and see if they show on the screen…that would eliminate the Y-Output stages and all stages after those loops. The one thing I hope ain’t buggered is the channel alt-chop multiplexer stage…older, lower soecd CROs simply used biased diode switches, but these high bandwith Tek’s tens to use a proptretsry analog MUX chip, if that is damnaged you might need help from the Tek forum or buy another 2230 to cabbage and biard jockey.
I’m not too sh!t-hot with firmware defined digital stuff, bat can give you a few general pointers. Digital stuff is usually pretty robust, it either functions fully or not at all. I think the kernel is working here because the squares appear and I’d suspect the micro has to be running to generate them, (again, as I suggested with the HP PSU, check the microprocessor clock crystal with the 54600 and see if there is RF there, if so, the micro is going. If not, the micro will be buggered.
When I was studying Chemistry st the University of Tasmania, there were some electronics techs down in the basement of the Chemistry building, one of them told me how to think of how a mocroprocessor works fo the purposes of repairing mocro containing gear…the program running in the micro and associated RAM and ROM can be thought of aca toy trainset, the little train starts at grand central, (i.e. bootup) then goes round and round on the “main track”. If an instruction arrives at an input of the micro, it can be thought of as “changing the points” in the “track” and making the “train take the mountain route via the tunnel”, but it eventually goes back to grand central and starts again. If the “train” derails, i.e. the program gets into an illegal state or lockes up, uou hsve to reboot to make it appear back st grand central. It wil never pick itself up back onto the rails at the point it derailed!
Somehow, your train is getting stuck at the screen geometry test. The EPROMS might need to be redone…their contents are available on the Tekwiki 2230 page…w140(dot)com(backslash)tekwiki(backslash)wiki(backslash)2230.
Cheers and good luck, I will be watching with interest.

[@fmashockie]

Thank you as always for your input Globe Collector! I will definitely be checking those areas your mention. But this one will definitely require me to sit down and read 'theory of operation' portion of the service manual. Unlike HP manuals, it seems Tektronix doesn't give as thorough a troubleshooting section. HP (not referring to Agilent/Keysight because these days they won't even give you a schematic) will give you troubleshooting trees and step by step procedures for each portion of the circuit. Tektronix makes you work for it lol. Which might be good for me in the long run because it may help better understand how it works. Thank you for the analogy on how micros work! I did probe the 5V rail by the 8088 and it is not short so that is a good sign! And I will have to take another look at the manual for this calibration mode - I scanned the manual for an example photo of these 'squares' but it was nowhere to be found - must be described in the text then. But I'll save that for later. I think I'll start with troubleshooting the analog first and then work on the digital. I'll keep you posted!

[@thetechgenie7374]

Yep shock he didn't get any, power supply, EHT supply, tube Z-Axis amp and intensity DC restorer and also multiplexing issues that most of these have, especially considering it condition it came in, as had better condition one with one, or a few of those issues? I seen it on every 6 out of 10 I repaired on this series to the point modified the focus and intensity on this series of scopes. Yes frequency track with time base setting on theses as well as the 2400 series I worked on. These have a PU diagnostic it will run before bootup complete if digital board was at fault you get errors before the Cal aids screen. Typically if you get multiple issues like Rom, CPU and a few ram chips coming up as bad, it likely not the digital board and somewhere on the local supplies.
He figured it out and was a bad switch that caused the cal aids to stick as that is another way to pull up cal aids is hit one of the display control switch when it boots. The shame about this one, it seem to be working good before eBay seller scavenge parts from it as alignment and calibration was OK from looking at the cal aids which is not common as most are way out, even low hours ones. All it likely needed before eBay seller got to it was switches cleaned, or replaced and would have been a good working scope.
I do have a few of those proprietary analog MUX chip I donate to him if he needs one as have a few I keep as they do fail, as well as a few of the missing knobs. It nice to see a older piece of test equiment put back to use and working 100% when most people nowadays don't even bother and throw them away.
Old Rifa X/Y safety capacitors don't get me started, I replace them before even turning on the equipment. I have umbrella cockatoo's even though have a good air cleaner in their room far away from my bench area now, I don't chance it. Yes they hold out at 120v better but will leak current and blow at 250v.
The 2230 was a great scope back in the day, used one back in the day for over 10 years and cost over 5k new. It is a shame people just don't care and toss it as they consider it outdated, or don't know what it is. It still a quite useful oscilloscope due to it being a hybrid and can handle higher voltages 50v/div better then newer DSO's that get 10v/div that work on vintage electronics.

[@hopper7755]

with those unsoldered wires I think they tried to swap the board with the attenuator switch from ch 1 to ch 2. the marking ch1 and ch 2 also confirms this. They look identical

[@fmashockie]

Good point! Maybe they broke the channel 2 switch off in the process 😂

[@kiltrash1]

At 14:26 the white resistor links should go from the blue variable gain pot to the corner hole in the upper PCB. There's one on each side. I have photos of the Y attenuator in various stages of disassembly if it helps.

[@fmashockie]

thank you! I did find that out eventually. Part 2 will reveal additional issues with the attenuator board 😂

[@fmashockie]

Since you own one of these, any idea what those 'squares' stuck on the display are? I've heard some people speculate it is some calibration mode, but I can't seem to find an example figure of this display in the service manual anywhere!

[@OrBiT257]

​@@fmashockieyeah, something like that. I messed around a bit with those squares on my 2230 and I concluded they serve to allign all the storage/readout information, ... Maybe 2 months ago, when I was surfing trough eevblog forum I found several threads about this problem on 2230. I think one mentioned that this happens when you start the scope with some memory switches pressed or something like that. The best would be to go through it and maybe you find some more information, that comes handy.

[@fmashockie]

Thank you so much I'll check the forum!@@OrBiT257

[@mirkobruhn9841]

I do not know about the loose wires, but I have seen often things in Tektronix build “dead bug style”, like the resistor and diode.

[@fmashockie]

Gotcha thanks for that info! I started seeing similar soldering on other parts of the board and thought maybe that was the case

[@maebeans]

this scope has had the bad touch, oh no

[@maebeans]

do you suppose they desoldered those points on the back of the knobs to inject a signal? sort of a brute force way of making it do *something*?

[@fmashockie]

possibly! I don't know enough about the damn thing to wager a guess 😂I have a lot to learn with this one if I want to get it back up and running! @@maebeans

[@jleonardoperez5402]

I'm going to need to teach analog electronics. Because this generation is not up to speed

[@fmashockie]

by all means. Look forward to hearing something more constructive