Tektronix 2430 Repair (possibly damaged CCDs/preamplifiers?)

[nsd_c]

Hello everyone,

I picked up a lovely Tektronix 2430A 150 MHz dual-channel DSO back in November or thereabouts at a hamfest. It's a 1987 vintage and looks to be cosmetically in very good shape--opening it up for repair, I hardly even saw any dust anywhere. At the time I purchased it, the seller informed me that the second channel was non-functioning, which was the case. I used the scope for maybe two months, and then, *poof*! There goes the first channel. At first I wasn't even sure what was happening--I was seeing noise on the display but way off-screen (below) and only at high sample rates (so many nanoseconds for the timebase) and low vertical scale voltage. 

I've done some reading and came across this thread elsewhere on this forum where another user had a 2430A that needed mostly just being cleaned up on the inside. Also, I found this page containing a lot of good information as well as an OCR copy of the manual (direct link for those curious).

Anyway, cutting to the chase, the scope shows no errors in the 1000-6000 range, so I believe it's safe to assume that the NVRAM and such are all fine (right?). The scope shows tons of errors on the 7000-9000 range, however. I would have posted the whole table but it's pretty long. For anyone interested, I uploaded it here.

So, here's what I know/understand:

• Again, NVRAM seems fine. Also I poked around through all of the menus and such and everything seems fine. (I'm not really sure what portions of the software reside in NVRAM and what reside in, oh I don't know, EPROMS?)
• Attenuator relays click when transitioning from 1V - 500mV and from 100mV -50mV. Relays also click when switching between input coupling modes(AC/DC/GND). I did a little bit of poking around with a logic probe U511 and U221, which are the shift registers that drive U520 and U220, which act as buffers between them and the attenuators, and what little poking around I did seemed to jive with what's supposed to happen.
• Some have suggested (iirc in that other thread linked to above) popping the CCDs (and perhaps other hybrids?) off and cleaning the contacts with isopropyl alcohol. I haven't tried this but I don't really have reason to suspect that all of this is just due to the contacts being slightly corroded/dirty. Is this worth a shot? I'm a little hesitant to remove them if I don't have to. Also, I have some DeOxit; would that be worth a try if I do go that route?

On a whim (and an eBay gift card), I bought two CCDs for this unit, though I don't know if they're functioning. Now here's my real question (well, aside from some others ): is it a safe bet to try and put one (or both) of these possibly functioning CCDs in this scope and see if she works? I don't know a whole lot about troubleshooting these scopes (or really any scopes for that matter, quite frankly), so I'm a little apprehensive that perhaps there's a short somewhere and that blew... whatever is actually blown (if that's possible) or something along those lines. And, if that's the case, if I try and stick the "new" CCDs in there they'll just get blown, too (provided that they weren't already also broken).

Now, there is also the preamplifier stage, though I'm not exactly sure how I'd test that. I'd guess that I'd go about testing the PA by injecting a signal into the input and watch the output using another scope. I do have a crummy 2MHz Tek R5103N that works as well as a MTS 410 function generator that works (sorta) well enough, so this is within the scope (pun not intended) of things I can do.

It may be worth noting that I don't quite have the necessary equipment to calibrate the scope: I have DMMs at my disposal but the only DC power supplies I have are 0-6V units. Floating outputs or no, 3x6 isn't high enough for the 20V needed (I can't seem to find this in the manual atm but see this thread), unless I go ahead and purpose-build a shoddy linear DC PSU just for this purpose ().

Thinking about the errors that are being displayed and looking at the schematic for the input/attenuator/preamp section of the scope, I'm starting to think that likely the CCDs are all fine and dandy and the real issue is the preamplifiers: I can't quite imagine that the PAs, CCDs, and trigger circuitry all decided to go kaput at the same time. I don't know how the scope does its self cal/diagnostics routine exactly, but I'd suspect that with the PA toast, it can't test the CCDs or trigger.

Anyone know if this plausible/correct? And where I might go from here?

[PaulAm]

Although there's not a lot on the 7000 errors, the service manual does have a procedure for identifying whether or not the CCDs or other components are at fault.  The 8 and 9000 errors might be related to the 7000 errors.

Did you try following that?

[nsd_c]

Although there's not a lot on the 7000 errors, the service manual does have a procedure for identifying whether or not the CCDs or other components are at fault.  The 8 and 9000 errors might be related to the 7000 errors.

Did you try following that?

No, I haven't, as I didn't see that while browsing through the service manual. I'll take a look and see what I find. Thanks for the heads up!

Also, an update:

I found that the channel 1 and channel 2 attenuators (AT400 and AT300) are connected to the (channel 1 and channel 2) preamplifiers U420 and U420 by a 470K resistor (R1005) in series with a "selectable" (i.e., chosen at time of manufacture) resistor R1015. R1005 is in parallel with a 0.001 (uF?) capacitor C1005. For anyone trying to determine if their attenuator is properly functioning on their 2430A, tying into the output, right before R1005 is probably the best spot. I looked and looked on the board layout and couldn't find it, then realized that it's hidden under a small piece of shielding held in with two torx T-10 screws. (The photo was too big to upload to the forum, but I uploaded it here; channel 1 is shown with shielding removed and channel 2 with shielding in place.)

Tying in here with my other (working) scope and applying a sine wave with a DC offset to the input, I was able to verify that the attenuator relays (1x, 10x, and 100x attenuation) were working as well as the input coupling relays (AC/DC/GND). I didn't try 50 ohm termination.

I also tried probing pin 1 of A10U230 (which is the output of a MC1458U dual op-amp between the preamplifiers and the trigger circuitry) and found a steady 1.85 VDC, so I figure either that op-amp is toast (probably relatively unlikely) or the preamplifier(s) are dead. Just for reference I re-drew (part of) this part of the schematic.

[siggi]

Hey there,

You should start at the beginning by checking the power supply voltages. The two channels are separate past the CCDs, so if both failed, you likely want to look for common things to the channels. The op amps are probably just doing sample and hold on control voltages, perhaps the DC offset. There will be a way to play with the DAC outputs from the front panel, check the service manual for the details....
On the bench, you MUST run a fan on the hybrids on the underside of the scope or they may cook. themselves.
It won't hurt to re-seat the hybrids, so long as you observe good ESD hygiene, and it may well help.

You may also want to check out the TekScopes yahoo forum.

Good luck!
Siggi


Sent from my iPad using Tapatalk

[nsd_c]

Hi siggi, thanks for the reply!

You should start at the beginning by checking the power supply voltages.

That would be the sane and rational place to start, but I didn't start there. Luckily for my self-respect ( ), though maybe unluckily for my ease of fixing this, I did check the rails via the side board (A13) and they all matched what was printed on the silkscreen.

The two channels are separate past the CCDs, so if both failed, you likely want to look for common things to the channels. The op amps are probably just doing sample and hold on control voltages, perhaps the DC offset. There will be a way to play with the DAC outputs from the front panel, check the service manual for the details....

Yeah, I'm seeing something now about a "FORCE DAC test to determine if the DAC system can control the PD-OFFSET voltages (PD11, PD13, PD21, and PD23)"; I'll have to look into this some more. Thanks for the tip!

As for the U230 op amp, the manual says:

Common-mode signals are rejected from the trigger signal by the circuitry composed of operational amplifier U230B and associated components. The inverting input of U230B (pin 6) is connected to the common-mode point between +PICK (pin 12) and -PICK (pin 15) of U420. Any common-mode signals present are inverted and applied to a common-mode point between R133 and R235 to cancel the signals from the differential output. A filter network composed of LR421 and a built-in circuit board capacitor reduces trigger noise susceptibility.

(It also says elsewhere about these being used for the triggers, and errors related to the trigger can be related to the op amps. But they're probably fine and I just don't understand quite how they're being used.)

I had tried probing these because at the time it seemed to be the most expedient way to see some kind of signal coming out of the preamplifiers, but really I guess what I need to look at would be the differential outputs, wherever they're hiding. (Unfortunately those big heatsinks and weird packages make it so you can't just probe right at the thing.)

On the bench, you MUST run a fan on the hybrids on the underside of the scope or they may cook. themselves.

Thankfully I did see that warning elsewhere! (I believe it was the thread that you started?) Yeah, lucky for me I did see that warning beforehand and I've been using a small 10" or so fan a couple inches from the main board to keep it cool. I used a temperature probe and it seems like nothing gets over the mid 90 deg. F range or so with this setup.

It won't hurt to re-seat the hybrids, so long as you observe good ESD hygiene, and it may well help.

Duly noted, although I wish the eBay seller who sent me the replacement CCDs had heeded that and packaged them in antistatic bags! I will try re-seating them at some point.

You may also want to check out the TekScopes yahoo forum.

Will do, thanks for the suggestion.

[...] the service manual does have a procedure for identifying whether or not the CCDs or other components are at fault. 

I didn't find a complete procedure for this, but I did find some information on using the CAL PATH "special function" to try and isolate the issue as either being on the peak detector/CCD or attenuator/preamp side of things. I removed J156 from A13 to enable the special functions and turned CAL PATH on and the signal did not return to the display, which according to the manual implies that the issue is on the peak detector/CCD side.

I'm noticing as well that I see a garbage signal at 50 microseconds and below. These signals appear as a sort of very narrow pulse train, and the timing between "pulses" appears the same width in terms of horizontal divisions on the 1uS, 2uS, 5uS, 10uS, 20uS, and 50uS scales, really only changing when going down to the nanosecond scales. (Tomorrow when I have a chance I will take a picture to show these signals). Also, I notice intermittent strangeness when trying to adjust the vertical position of both channels A and B: the vertical (DC) offset of either displayed signal does not move, but weird lines appear sometimes where the pulses are, which do move up and down. Also, these signals appear the same with CAL PATH on or off and regardless of the attenuation/input coupling settings.

I should note that when the first channel died, I noticed what seemed like a waveform juuust off the edge of the display, but unable to move it back up into view. This sounds like the sort of issue that the service manual talks about with a faulty peak detector (or some related component, I don't really know without studying the manual some more). I noticed that currently the trigger symbol is stuck on the lower part of the display, possibly implying that the displayed signal (whatever might be there) is clipped negative.

I'll have to read up on the role of the peak detectors and testing those with the FORCE DAC functionality. Also, possibly acquiring a DC PSU to perform the EXT CAL routine.

[siggi]

I should note that when the first channel died, I noticed what seemed like a waveform juuust off the edge of the display, but unable to move it back up into view. This sounds like the sort of issue that the service manual talks about with a faulty peak detector (or some related component, I don't really know without studying the manual some more). I noticed that currently the trigger symbol is stuck on the lower part of the display, possibly implying that the displayed signal (whatever might be there) is clipped negative.

I have a 2430 that came to me with ugly battery corrosion that had damaged a bunch of stuff on CPU board.
The repair was a bit of an odyssey, but after I'd finally got'er running smoothly again with all self tests passing, I left'er for a couple of hours for "burn in". When I came back, one of the input channels had drifted off screen. This was initially intermittent, depending - apparently - on the temperature of the peak detect hybrid. After a little while, however, it became consistent fault. At the time I figured I'd pushed the poor hybrid over the edge with poor ESD hygiene during my repair.
In my case I was able to isolate this to the CH1 peak detect hybrid, as the fault was only on one channel. This'll be harder to sort with both channels gone, alas, as assuming you have the same fault on both sides you won't be able to play switcharoo.

I'll have to read up on the role of the peak detectors and testing those with the FORCE DAC functionality. Also, possibly acquiring a DC PSU to perform the EXT CAL routine.

For the EXT cal routine, you can easily rig up something with - say - a stack of 9V batteries, a suitable resistive divider, a bit of trimming and perhaps some bypassing on the output. The scope input has 1M input impedance so there aren't any currents here to speak of. You'll probably want to filter and trim the output from your average low-cost DC supply in any case...

You can, of course, also use this as a pretext for buying a decent DC supply .

[siggi]

That would be the sane and rational place to start, but I didn't start there. Luckily for my self-respect ( ), though maybe unluckily for my ease of fixing this, I did check the rails via the side board (A13) and they all matched what was printed on the silkscreen.

While it's unlikely that this is a power supply problem, in your shoes I'd also see whether the ripple on the rails is within spec. The PSU in this scope is switch-mode, and getting switching spikes into your power rails could cause arbitrary modes of failure.