Broken Tektronix TDS 320 PSU

[fox8091]

I've recently been given my first two oscilloscopes, a BK Precision Model 2120 and a Tektronix TDS 320. The BK Precision works fine (They thought the 2nd channel was broken), but the Tektronix TDS 320 does not work. I would like to get the Tektronix working because it's 100MHz compared to the 20MHz on the BK Precision. However, it seems the PSU has failed, and I'm not sure where to start on repairing it. I've called Tektronix to inquire about a replacement, and when they said they no longer had it, I traced down Zytec's current company (Artesyn Technologies) to ask about surplus, they no longer have it either. So a repair or a redesign are my only options. Help?

[james_s]

I'm working on one of these too, mine is from a TDS 460 but it's the same PSU. There is a LeCroy scope that uses a nearly identical power supply and the schematic is in the service manual for that one, I can look it up when I get home if you like.

For the most part the usual SMPS troubleshooting techniques apply to this one. Check for bad electrolytic capacitors with high ESR. Check for bad solder joints, particularly on the output connectors and the vertical PCB. I've heard the optocoupler can fail too, and the LM324. What is the symptom with yours?

Mine experienced a flashover on the primary side which appears to have had an external cause. I fixed a couple of vaporized traces but I believe the green hybrid was damaged. I found a reverse engineered schematic and am working on laying out a replacement, more for fun than anything else.

[fox8091]

Mine simply won't turn on. It makes a hum from the PSU when I try. Just a btw, I don't have much experience, nor do I have very good tools.

[lowimpedance]

As james_s indicated a good start would be to check the electrolytic caps in the supply as these are most likely the issue. (dried out).
From memory that was the case for a TDS320 at work so that's the first thing I would check.
Do be careful to make sure the supply is not retaining any stored charge on the main input cap etc. when you start poking around.
Maybe find some one you know with more experience nearby to guide you.

[james_s]

Mine simply won't turn on. It makes a hum from the PSU when I try. Just a btw, I don't have much experience, nor do I have very good tools.

Be careful working on switchmode power supplies then, they are unforgiving devices that pose a significant challenge for a beginner. You don't need a lot of fancy tools, a decent multimeter and soldering iron is about the minimum, ESR meter is really nice to have but you can also do the shotgun approach and replace all of the electrolytics.

Test it on the bench without the rest of the oscilloscope connected. All you need is mains voltage to the input and you can probe the output pins with a meter. Only once you have verified that it works should it be reinstalled, you can potentially completely destroy the scope if the power supply is wildly out of spec.

[fox8091]

I still don't have a good multimeter really. I have some free Harbor Freight thing. A proper multimeter is on my to-buy list though. I just recently bought some flux.

[james_s]

Ok well you may want to pack up the scope and set it aside for now, or send it out to someone who can go through it because you're going to destroy it beyond repair, injure yourself or both. I'm not trying to rain on your parade but this is not a beginner friendly project, at least not the power supply and I would never get near a line operated switchmode power supply with one of those crappy little HF multimeters. I have one of those I keep in my car for emergency diagnosis and they work acceptably for that but poking around a 340VDC bus with a cheap little toy meter is a deathwish. Meters like that can flash over internally and when that happens they tend to explode. You don't have to spend a fortune on a Fluke but you do need something properly designed and rated to do this kind of work safely.

Another thing that will need attention, these scopes are full of surface mount electrolytic capacitors, there are a bunch of them on the acquisition board, attenuator board and the front panel board and I guarantee they are leaking and damaging the boards. You will need to remove the boards and wash the leaked electrolyte off ASAP and then replace these capacitors, don't even try to power it up until this has been done. The capacitor replacement does not have the safety risk of the power supply but it's a fiddly job that requires careful technique and a temperature controlled soldering iron or the pads will rip off the board and then the instrument is scrap.

I also recommend removing the two lithium memory backup batteries as they will often leak eventually and when that happens it makes a huge mess. This picture is a computer rather than an oscilloscope but it's the same type of battery.

http://www.vintagecomputing.com/wp-content/burst_battery_small.jpg

[fox8091]

I've been injured enough times to know it's not beginner friendly, but still want to try it anyways. The best way to learn is to get in way over your head. Also, regarding the battery, mine doesn't seem to have one, instead relying on NVSRAM.

[james_s]

The TDS300 and TDS400 series scopes have two batteries, one plugs into the CPU board and one plugs into the DSP board. They both sit in clips at the end of the card bay behind the front panel.

If you insist on working on this, replace the electrolytic capacitors in the power supply, all of them except for the two big bulk filter capacitors, and *be careful* with those bulk capacitors, the DC bus on those can knock you dead before you know what hit you, there's no second chance if you screw up there.

And clean up the leaky capacitors on the acquisition, attenuator and front panel board, the fluid is corrosive and will destroy the boards beyond repair, greatly accelerated rate if powered up. I'm rebuilding one of these instruments myself so I'm familiar with the problems. It's a nice design that is pleasant to work on for the most part though.

[fox8091]

Like I said, I checked. There's no batteries in mine. And the DSP board and the CPU board aren't separate, but a single board.

[ArcticGeek]

Fox,

I think people are getting the TDS300 series confused with the TDS400 series.  The TDS400 series does indeed have a couple of batteries in it, the TDS300 series does not.  At any rate, the state of the battery will not cause the unit to not power up.   

Also, the TDS300 series was not plagued by bad electrolytics like the TDS400 and TDS500 series.  While the TDS320/340/360/380 were built at about the same time, they didn't have the huge number of caps that the 400/500 series had and thus did not have near the problems.  I've repaired dozens of TDS300, TDS400, and TDS500 series scopes, and while it might be a good preventative measure to replace the caps on the 300 series, in general they don't cause power on issues.  Also, the TDS400 and TDS300 series of scopes do not use the same power supply.   There is a schematic available for these on the Tek website though, that might help.

As others have pointed out doing repair of off-line power supplies is not a beginner project, and can be very tricky (and dangerous) to perform.

For what it's worth, I do have an old power supply from either a TDS320 or TDS340 that I'd be willing to let go pretty cheap, contact me off line if you're interested.  Good luck!

[james_s]

Well shoot, now I feel like an idiot, I could swear I had worked on a TDS340 before but I think now it was a TDS430. They look almost identical on the outside so I assumed they were very similar on the inside.

At any rate I bought a dead TDS340 to play with, should have it by the weekend so I'll be a bit more familiar with the 300 series. It will be interesting to compare to the TDS460 that I'm already working on.

These are rather old scopes and a bit limited in terms of sample rate but they're still nice instruments. Well designed, easy to work on, everything is accessible with the cover off and subassemblies come out easily. As long as you understand the limitations of an older DSO with limited sample rate and memory they can still be powerful tools when used properly. Some don't like the Tek user interface but I find the interface on my TDS784C to be pleasant to work with and these are similar. Perhaps it's just a matter of what one is used to.

[james_s]

Ok well the TDS340 arrived already, turns out it was coming from fairly nearby. Wow, no kidding these things are different! The chassis and monitor are exactly the same as the TDS460 I already had on my bench but that's about where the similarities end. The whole works is in a small PCB on the bottom that is 1/4th the size of the capture board in the TDS460, the whole card cage is completely empty, there are modern DSOs I could store in the vacant space inside this thing. It's too bad they didn't put a door in the lid to make a compartment to store the probes and accessories inside.

When plugged in I could hear a soft rapid ticking sound from the power supply. Pulled it out and did some poking around and found a FEP16DT ultrafast dual rectifier diode with one side shorted, replaced that with a similar part I had in my stash of salvaged parts and that resolved the problem, it fired right up and works perfectly, just needs to be cleaned up. That was much easier than I was expecting.

The diode is clipped to one of the two heatsinks in the power supply, if yours has this symptom that might be a good thing to check first.