Author Topic: HP 54602B oscilloscope problems repair (Read 13070 times)

bd139 · « **on:** February 09, 2018, 01:50:00 pm »

Another scope crossed my hands today. Few issues with it with resolutions for the interested

Display squashed

Usual C608/609 on CRT board replacement fixed that. vertical was rolling after so the CRT went through the calibration steps in the manual.

Before:

After:

Button 2 not working

Cal and ground ports are right next to the track here which was damaged the board. This has been repaired and buttons are all working. Fix was reroute track with some kynar:

channels 2 & 4 not working

TBD

Noisy fan

Needs replacing.

bd139 · « **Reply #1 on:** February 09, 2018, 03:29:04 pm »

Channels not working are interesting. No sign of any physical damage after extracting the main board. The bottom is dirty as anything though. There were some rubber mounts on the bottom which have disintegrated into slime. These are coincidentally underneath channel 2 and 4. Also the bridge that connects the shield and connectors together was dirty on channel 2 and 4. Hopefully related.

don.r · « **Reply #2 on:** February 09, 2018, 04:13:22 pm »

Very nice sleuthing. I have a 54645D that needs some small front panel work, how hard are these to get into?

bd139 · « **Reply #3 on:** February 09, 2018, 04:19:15 pm »

Really easy to get into. Two screws on the back and you're done. To get the panel off, undo the ribbon cable from the main board and then pull it back at the base and it'll unclip.

Not having any luck getting channel 3 and 4 up so far. I'm working from some 54600A schematics which I have. Seems to be the same front end.

don.r · « **Reply #4 on:** February 09, 2018, 04:34:43 pm »

Super! Thanks for that. Mine is a mechanical issue so hopefully won't need schematics. Good luck with the repair.

bd139 · « **Reply #5 on:** February 09, 2018, 04:40:05 pm »

Fingers crossed for you

I've narrowed this down to missing signal entering the programmable amplifier. Hopefully just a problem with the control circuit around that as it's all discrete components, mmbt3604, mmbt3906, mmbt4416 transistors and an LT1097 which are all really easy to get hold of.

bd139 · « **Reply #6 on:** February 09, 2018, 05:04:58 pm »

Found channel 4 problem. Nuked transistor due to a short on the board. This means that the vertical position is stuck near the bottom.

Made a right mess. This has burned 0.5mm into the board surface at least. Cleaning hole out. Was marked P43 which looks up as a P channel MOSFET.

Also found a broken trace that is killing the erase button.

This thing has taken a beating.

bd139 · « **Reply #7 on:** February 09, 2018, 05:17:43 pm »

oh this is not pretty. Has been dropped. There's multiple fractures and track breaks on the front panel board. Extensive repairs. Fun.

bd139 · « **Reply #8 on:** February 09, 2018, 06:00:49 pm »

Boards repaired. Now to play “guess what component that was”. All others are parked P and 43. They’re all SOT23.

Specmaster · « **Reply #9 on:** February 09, 2018, 06:46:54 pm »

Have you found the cause for that yet?

bd139 · « **Reply #10 on:** February 09, 2018, 06:59:37 pm »

Board cracked and shorted.

don.r · « **Reply #11 on:** February 09, 2018, 08:23:47 pm »

Must have been a heck of a shock to damage the board like that. Was there any noticeable traces of damage on the outside?

bd139 · « **Reply #12 on:** February 09, 2018, 08:48:46 pm »

Yeah the top right hand corner of the front panel is cracked. Didn’t notice that to be honest. Ah well that’s how it goes

bd139 · « **Reply #13 on:** February 09, 2018, 09:00:03 pm »

Looks like it was once a bas40 diode pair which makes sense. It’ll be being used as a switch.

I will pop one of the I assume same parts near it and test that as a dual diode.

Specmaster · « **Reply #14 on:** February 09, 2018, 09:33:23 pm »

Quote from: bd139 on February 09, 2018, 09:00:03 pm

Looks like it was once a bas40 diode pair which makes sense. It’ll be being used as a switch.

I will pop one of the I assume same parts near it and test that as a dual diode.

Well you're making good progress with it, should be working tomorrow with a tailwind.

bd139 · « **Reply #15 on:** February 09, 2018, 09:41:15 pm »

Fingers crossed. Although if I'm honest I'm dreading looking at the channel problems. There's just no signal on either 2 or 4. The plan is to compare the voltages on channel 1 and 2 when in the same state and see if there are any differences anywhere. There are large defined objects such as opamps which I can check voltages on. If not, it's old fashioned signal tracing with a generator plugged into the channel and work back from the front end to the ADC and see what happens. I've got the schematics for an earlier revision of the device so I can trace it roughly and go from there.

On a positive note, I didn't have any bas40's or any schottky diodes. I probably do somewhere but I couldn't find them. So I just jammed two 1n4148's across the burned hole where the bas40 was.

Bingo. Whole panel works now

Shitty but it works:

Fortunately with these, apart from a couple of usually reliable ICs, they're chock full of standard parts. All SMD but easy to work with as they're quite large parts. Smallest is 0805.

Specmaster · « **Reply #16 on:** February 09, 2018, 09:45:39 pm »

Neat, the board also look pretty clean as if its not seen much use, maybe the damage happened a few years ago and its been stuck in a repair queue and then the owner decided to get rid of it?

bd139 · « **Reply #17 on:** February 09, 2018, 09:48:01 pm »

That's only because I cleaned it

. It was covered in shit and the other half still is. The outside has been cleaned before he sold it. Seller was these guys: http://www.tics.co.uk/

Specmaster · « **Reply #18 on:** February 09, 2018, 09:58:33 pm »

Nice, but it would nice to them put some prices against things rather than call us for a price all the time. Surely you want to know if the price is within your budget before wasting time and money on phone calls?

bd139 · « **Reply #19 on:** February 09, 2018, 10:08:55 pm »

Yes indeed. I was having exactly that discussion earlier with a colleague about software as well. Certain types of software require you to call someone to set up a trial or get access to it to install locally. I have a rule of if you can't download it, then you can fuck off!

Think I might have isolated the channel issue at least in theory looking at the schematic. There's two hold ICs in the analogue signal path. Each of those has two analogue inputs and one output. A digital line called ch_bank_sel switches between the inputs on both ICs so it'll digitise pairs of channels 1,3 and 2,4. This flag is controlled by a 74act574 tristate DFF off the CPU bus. If the output of this is hosed or the resistors in the level converter between it and the hold ICs are hosed then it could be feasibly why alternative channels are working i.e. channels 1 and 3. I will check that tomorrow.

Hopefully the hold ICs aren't hosed as they are unobtainium and also bastards to get off the board.

Specmaster · « **Reply #20 on:** February 09, 2018, 10:12:49 pm »

Fingers crossed they are OK and its something else then.

bd139 · « **Reply #21 on:** February 10, 2018, 02:27:28 pm »

Well it wasn’t that. Confirmed hold ICs and ADCs are good and the entire front end. Phew. The problem lies in a CA3246 transistor array between the hold and the ADC This appears to be used as a summing amplifier that sums a voltage from an DAC and the reference voltage with the input voltage. Now there’s a big offset on channel 1 as well so investigation is now heading towards voltage reference and DAC. Channel 2’s offset pushes the signal out of the range that the ADC requires so there’s no data coming out hence no signal. It’s pegged off the screen.

Will resume again in a few days. Have other things to do at the moment.

Specmaster · « **Reply #22 on:** February 10, 2018, 02:30:51 pm »

Good work, looks like I might be getting my meter bits and bobs tomorrow now, assuming he see's my reply to his message asking him to confirm that tomorrow is OK and also for a contact number just in case....

bd139 · « **Reply #23 on:** February 10, 2018, 04:09:45 pm »

Fingers crossed for you

rob.manderson · « **Reply #24 on:** February 10, 2018, 06:34:12 pm »

Quote from: bd139 on February 09, 2018, 10:08:55 pm

I have a rule of if you can't download it, then you can fuck off!

bd139 · « **Reply #25 on:** February 11, 2018, 04:49:16 pm »

Before I forget, I broke the board down into rough logical functions before I attacked it. This is not exactly the same as mine. I didn't have photos at the time so I was reverse engineering it from schematics.

Disclaimer: this isn't necessarily exactly the layout - this was "where to look for shit" to start with. Lots of pissing in the dark here

bd139 · « **Reply #26 on:** February 12, 2018, 04:07:40 pm »

Further debugging. Very interesting so far. I did some checks on the circuit design and there's a 100k/0.1uF between the DAC and the LM324 which controls the summing amplifier. If I inject a voltage into the input of the LM324 then I can bring the trace back into the range of the ADC and get a trace back on all channels. This confirms this is entirely a problem with the DAC controlling the vertical position. However it is pulling 25mA when I do this. I think this might just be a ceramic cap or two that have fractured short when it was dropped. If I had an IR camera I could probably see it. The DAC is well protected from output short however which is good.

Dac is an HP 1SJ2-0102. Worst case not totally unobtainable if I have to replace it.

Found an interesting technical article on the architecture and design of the scope series as well which helps with the understanding: http://www.hpl.hp.com/hpjournal/pdfs/IssuePDFs/1992-02.pdf

The DAC and filters are in the top half of the image above labelled "clock gen". They aren't clock gen. That's the DAC. Remember what I said about pissing in the dark - I was completely wrong

bd139 · « **Reply #27 on:** February 12, 2018, 04:27:55 pm »

I'm an idiot. Been staring at the wrong LM324. There was another one hiding under one of the DL marked packages (have no idea what the hell they are).

This one does NOT look healthy - note the burn marks. I will trace it out and order a replacement.

bd139 · « **Reply #28 on:** February 12, 2018, 04:33:24 pm »

Ordered 2 from RS. £2.88 including delivery. 2 because if I don't work out why it blew up correct the first time it saves me another day of waiting

Good excuse to fire up the T12 here and use it

I reckon it might be voltage rail collapsing causing the input / output to be driven past them.

Specmaster · « **Reply #29 on:** February 12, 2018, 04:37:15 pm »

No you're not an idiot, that much we all can vouch for. Hmm pins 1 and 14 are outputs, thats give away then, 1 for each of the affected channels.

Specmaster · « **Reply #30 on:** February 12, 2018, 04:39:03 pm »

Would that T12 be the mains switcher or are we still on the external supply one?

bd139 · « **Reply #31 on:** February 12, 2018, 04:43:19 pm »

External supply one. Mains one hasn't arrived yet

Found a TH LM324 lying around In the junk box. Resisting wiring up a bodge impatiently with kynar

Specmaster · « **Reply #32 on:** February 12, 2018, 04:46:29 pm »

go ahead and do it, whats the difference between that and the M version I wonder?

It would give you more debugging time if there was another reason for the old one failing, so that if this one failed, you might find the reason before the new arrives from RS?

bd139 · « **Reply #33 on:** February 12, 2018, 04:58:18 pm »

Can't find a reason. No shorts, no bad components. Am wondering if it just overheated. it's wedged underneath DL1 which looks like a large thin film array which is kicking out some serious heat. The other LM324's on the board are quite warm.

Specmaster · « **Reply #34 on:** February 12, 2018, 05:01:18 pm »

In that case, I'd be inclined to pull that sucker out and install the other one and try it, if it works OK, when the RS delivery comes, you'll have 2 spares in stock.

bd139 · « **Reply #35 on:** February 12, 2018, 05:04:39 pm »

Yeah looking like a plan.

Going to persist a little longer and see if I can trace out a bad cap as well as there are a ton of ceramics across the output of this amp in various places which could cause this. If they are low impedance suddenly, then the opamp will still try and correct its output and eventually blow itself up again by trying too hard. They're only rated for max dissipation for a set amount of time.

Specmaster · « **Reply #36 on:** February 12, 2018, 05:08:31 pm »

Well pin 14 certainly has been getting hot, but there looks like there might be a burnt hole on pin 1, at the bottom of the pin 1 ident marker, is it a hole or burn mark? If so than you are going to be looking for more than 1 cap?

bd139 · « **Reply #37 on:** February 12, 2018, 05:16:54 pm »

That's the orientation mark. It's pitted and burned.

For future reference, here are the output pins for the DAC:

DAC pin 23 - channel 4 vertical position
DAC pin 1 - channe 2 vertical position
DAC pin 4 - channel 3 vertical position
DAC pin 11 - channel 1 vertical position

Specmaster · « **Reply #38 on:** February 12, 2018, 05:20:56 pm »

What are pins 1, 7, 8 and 14 doing on the op amp?

bd139 · « **Reply #39 on:** February 12, 2018, 05:38:34 pm »

Hang on a min. I have traced this further out.

It goes from U14, the DAC at 0-3v approx to U1 which is an OP-400 instead of the expected LM324. applies a -2.5v offset to it giving it an output range of +1.6 to -1.6v on pin 14.

Tracing that output now with continuity tester. Low tech but it works. This is a 6 layer board apparently so YMMV with that method

The schematics I have are for an HP 54600A which are vastly different apparently. There are no schematics for the 54602B

Specmaster · « **Reply #40 on:** February 12, 2018, 05:45:12 pm »

PM me your email address, I have a copy of the HP 54600B which also covers the HP 54602B service manual.

bd139 · « **Reply #41 on:** February 12, 2018, 05:52:21 pm »

Ooh interesting. PM on the way

bd139 · « **Reply #42 on:** February 12, 2018, 07:22:06 pm »

Found it at last.

The LM324 is responsible for buffering the voltage reference from the main AD588 reference for the input voltage on the hold ICs. The hold ICs are expecting 0.5v as the reference voltage but the LM324 has blown up and the output is being pulled down. This is saturating the transistor which is holding it at -1.2v.

The only reason channel 1 and 3 works are because they are exceptionally high gain amplifiers in the control loop on that side of things which are saturating providing offset. No sign of anything nasty around the LM324 so will just swap it out and see what happens. There's a tant across the output I will test however as if that thing goes short it will sink 54mA constantly through the opamp which will eventually blow its arse out.

This is working off an HP54600A schematic but it traces out the same:

There are a couple of 100nF ceramic decoupling caps as well on the 0.5v rail so will check that too.

Specmaster · « **Reply #43 on:** February 12, 2018, 07:35:51 pm »

Ok, what are you waiting for, whack that lm324 in and lets see....

bd139 · « **Reply #44 on:** February 12, 2018, 08:09:10 pm »

Just tested the one I have and two of the opamps in it are dead. It was a pull out of something else. Grr. Will have to wait until tomorrow. Might yank the old one out tonight though.

Specmaster · « **Reply #45 on:** February 12, 2018, 08:21:26 pm »

I would, the more prep you can do today, the sooner you will know tomorrow when the new ones arrive, if it works. The suspense is killing me....

bd139 · « **Reply #46 on:** February 12, 2018, 08:31:05 pm »

Killing me too. Knowing ParcelFarts it will turn up at 6pm

Ran the circuit up in LTspice and actually looked at it properly this time. It's not applying any offset. the output should be close to 5v. What it's doing is removing any digital noise picked up from the reference before it hands it over to the hold ICs. I think that either the hold ICs are holding it at -1.2v (two transistor BE drops) or the extra mmbt3904 depicted above has died.

To quote the LM324 datasheet: "Short circuits from the output to V+ can cause excessive heating and eventual destruction. When considering short circuits to ground, the maximum output current is approximately 40 mA independent of the magnitude of V+. At values of supply voltage in excess of 15 V, continuous short-circuits can exceed the power dissipation ratings and cause eventual destruction. Destructive dissipation can result from simultaneous shorts on all amplifiers."

Specmaster · « **Reply #47 on:** February 12, 2018, 08:38:00 pm »

With it possibly not turning til late, then more prep done tonight the better, then all you have to do is drop it in and run up the T12 again, a few dabs of solder applied and then BANG it will either go again or hopefully fire right on all channels

bd139 · « **Reply #48 on:** February 12, 2018, 08:41:55 pm »

Will do that shortly. Curry-o-clock

Specmaster · « **Reply #49 on:** February 12, 2018, 08:43:13 pm »

Haha beat yeh to it, just finished mine.

bd139 · « **Reply #50 on:** February 12, 2018, 11:05:36 pm »

Found another LM324 and tacked it in and no difference. Have returned the old LM324. It appears that the original wasn't blown up as the burn marks rubbed off after washing post-reattaching it

... got two SMD LM324's going in the parts drawer though

Channel 2 still doesn't work however. I have traced the vertical position throughout the entire front end offset, control, everything and I can't find this one.

I think I will pause for a while on this one and consult the experts in the HPAK yahoo group.

Specmaster · « **Reply #51 on:** February 12, 2018, 11:39:48 pm »

Your stock is increasing all the time, mine is as well, which in the end can't be a bad thing because it will pay dividends at some point in the future I'm sure. Yes thats not a bad idea, assuming someone on it already has experience of your scope. Problem I find on it is that it does seem to be very much a closed circle of friends and is difficult to get anyone to read a post from a newcomer. But then, that might just be my take on it, whereas in reality it might be that at the time I post, there is no one active with the required knowledge that I was seeking so I may be wide of the mark with my comments.

One thing for sure, its not as friendly as the EEVblog forum

bd139 · « **Reply #52 on:** April 14, 2018, 03:33:16 pm »

Meant to conclude this thread a long time ago. I found a crack in the board on the channel 2 attenuator where it had been dropped. This is therefore irrecoverable. I have split it for parts for now. Display assembly and knobs are gone which made up the purchase price.

If anyone comes across this thread, I still have the main board, minus a couple of generic parts I nicked, most of the panel and the power supply. PM me if you need parts.

If anyone has another 54602B they want to shift in EU let me know


EEVblog Main Site	EEVblog on Youtube	EEVblog on Twitter	EEVblog on Facebook	EEVblog on Odysee

Author Topic: HP 54602B oscilloscope problems repair (Read 13070 times)

Share me