Fixing Tek TDS684A PSU

[siggi]

Hey all,

I'm attempting to repair a dead-as-a-doornail TDS684A I just bought off eBay - does not power on.

The PSU is very much like the one documented in the 520B service manual, but not identical in all respects. In particular the standby power circuitry looks quite different.

It appears the (one?) problem mine has is the standby power supply. I've located a pair of 1.5KE220CA/1.5KE250CA devices that have clearly suffered some trauma. The PCB was discoloured, and the devices are one-way conductive at 2-3V, so presumably have shorted in one direction?
Now, these guys go across the standby transformer primary, and mainly serve to clamp flyback voltages to protect the switch transistor I'd think (a BU508A - which seems massively over-provisioned). These 1.5KEs are however conductive in the other direction, e.g. from the 400V DC supply, and so effectively short the standby primary with a 5V zener.

Does this seem like a reasonable primary failure, e.g. some sort of fatigue in the devices, or a one time transient? I can see how there would be a cascade as soon as one gives. Else am I likely to find that there's e.g. overvoltage on the 400V DC supply that caused this?

Also I wonder if it's OK to run these supplies without any load?
I know some older Tek switching supplies didn't like that. I'd expect this one to be OK, though, as it regulates against a dedicated 5.1V secondary tap...

Siggi

[TVman]

Can you post some pictures of the inside please?

[siggi]

Can you post some pictures of the inside please?

Inside of the scope, or the PSU? What are you looking for? There really isn't much to see in the way of damage, just slight discoloration in the PCB under the TVS devices.

[nctnico]

Tell us what happens if you try to power the scope with the CRT section and acquisition board disconnected. However, keep the FAN connected. The 24V rail needs a minimum load.

There is an alternative power switch on the CPU board so you can also disconnect the front panel.

[siggi]

It appears the (one?) problem mine has is the standby power supply. I've located a pair of 1.5KE220CA/1.5KE250CA devices that have clearly suffered some trauma. The PCB was discoloured, and the devices are one-way conductive at 2-3V, so presumably have shorted in one direction?
Now, these guys go across the standby transformer primary, and mainly serve to clamp flyback voltages to protect the switch transistor I'd think (a BU508A - which seems massively over-provisioned). These 1.5KEs are however conductive in the other direction, e.g. from the 400V DC supply, and so effectively short the standby primary with a 5V zener.

I guess I should have mentioned that these guys are connected in series across the transformer primary. As far as I understand, this ought to clamp 1mA flyback to ~210+230V or ~440V. The clamp voltage is still a good 10% above the 400V supply, so they ought not to have any stress in that direction? Incidentally, The switch transistor is 1500V rated, so 840V flyback looks fine by it.

Everything else I've looked at so far seems fine, e.g. the standby switch transistor measures good, there's no shorting anywhere, etc.
I've seen clamp zeners fail short without apparent cause. Will TVS devices do that too, or am I definitely looking at a secondary failure here?
Might a failing reservoir cap cause this? I suppose the standby supply would see a jagged 400V rail when the PSU is on...

[siggi]

Tell us what happens if you try to power the scope with the CRT section and acquisition board disconnected. However, keep the FAN connected. The 24V rail needs a minimum load.

There is an alternative power switch on the CPU board so you can also disconnect the front panel.

Thanks, will do, though I need to get some replacement TVSes first.

[siggi]

This loks like the same exact PSU and problem: http://www.jvgavila.com/tds524a.htm.

[tom66]

Some TVS devices are unidirectional. This means they act like a normal diode in the "forward" direction. Others are bidirectional with high-standoff in both directions.  So I wouldn't say conducting one way is necessarily indicative of a fault.

[siggi]

Some TVS devices are unidirectional. This means they act like a normal diode in the "forward" direction. Others are bidirectional with high-standoff in both directions.  So I wouldn't say conducting one way is necessarily indicative of a fault.

According to the data sheet, these guys are supposed to be bidirectional. See e.g. http://www.littelfuse.com/~/media/electronics/datasheets/tvs_diodes/littelfuse_tvs_diode_1_5ke_datasheet.pdf.pdf, and note that "CA" indicates bidi.

Also going by the circuit they're in, and the fact that there's no direction marker on the packages (see data sheet again) I'm pretty sure they're bidirectional goners...

[siggi]

I did some work on this last night.

I measured 72.5 ohm resistance from the 400V rail to the collector of the BU508A switching transistor. By the data sheet it can take 8A max, and so for this failure mode of the TVSen, it's never even close to limit. With the TVS gone, it'll see max 2.3A from 120V rectified to 170. Even during failure it'll see "only" 400V/72.5 ~= 5.5A, which is still well in spec. Assuming the driver saturates it, post-failure it's only burning ~2.3W with whatever duty cycle it's PWMed, and it's easily heat sunk against that. Looking at the PCB, there's a bank of 1 or 2W resistors on the 400V rail that have clearly been warm for a long time. That's my 72.5Ohm.

As the BU measured good and I didn't have a replacement on hand, I replaced the TVSes and tried turning the PSU on. This was unloaded on my bench.
Got that startup twang, and measured 400Vish on the DC rail. There was a nasty ~10V .5-1Hz wobble on it though, perhaps hiccuping due to over voltage on the 24V rail, which is monitored, I figured.

All told, it seems the TVS gave, and the standby supply transistor now looked directly into 170V at 72.5Ohm instead of the primary, until the hard power switch was turned off. The resistors will have borne the brunt of the dissipation, then the TVS, then the transistor. Nothing else seems to have failed but the TVS.
I figure it must have been over voltage on the 400V bus that did them in, and apparently these devices will fatigue and fail if you repeatedly ask too much of them.

I stuck the PSU back in the scope, and she booted fine. There was even a saved reference waveform on the screen from the previous owner.

I played a bit, got her to display the calibrator and such. There seemed to be a nasty line wobble on the captured traces.
Then the PSU started hiccuping. This seems time or temp related, if I let her chill for a little, I get more play time until hiccup.
So I'm guessing/hoping there's more trouble in the PSU. Maybe the main reservoir cap is going, and this may have been the cause of the standby supply failure to begin with - any other likely causes?

Apparently this PSU will hiccup under overcurrent on the primary, or under over voltage on the 5.1V or 24V rails. Guess I'll have to load it down on the bench to see what's up.

I'm waiting on an ESR meter, it's on the slow boat from Japan, so I guess it'll be a while until I'll look at this one again.

[nctnico]

I'd try to run the scope without the CRT section connected (using an external monitor) first.

[Jay_Diddy_B]

Hi,
This might help:



I fixed mine by changing VR4, VR5 and Q9. But also check RT4 and RT5.

This is from one of the TDS 5xx, 6xx or 7xx scopes. There are several variations of this circuit.

Regards,

Jay_Diddy_B

[vaualbus]

Long time ago I've post a message about mine repair of a TDS scope. In my case turns out to be a problem on the main switching IC (I don't rember the name now but it is on the center bord).
Check If there is the 5V ref voltage from that board.
Also the TDS520B schematics is different from the A revision so not trust that schematics, instead buy the artek manual for the TDS520 that have the real schematics.

[siggi]

Hi,
This might help:



I fixed mine by changing VR4, VR5 and Q9. But also check RT4 and RT5.

This is from one of the TDS 5xx, 6xx or 7xx scopes. There are several variations of this circuit.

Awesome, thanks. Had Q9 failed, or did you replace it as a matter of fact?

Long time ago I've post a message about mine repair of a TDS scope. In my case turns out to be a problem on the main switching IC (I don't rember the name now but it is on the center bord).
Check If there is the 5V ref voltage from that board.
Also the TDS520B schematics is different from the A revision so not trust that schematics, instead buy the artek manual for the TDS520 that have the real schematics.

Thanks, ordered the Artek 520 manual. I hadn't realized it was different from the TDS520B one. I've been keeping Dave in beer lately, I think .

[Jay_Diddy_B]

Siggi,

The schematic that I showed is from the TDS520 manual. (You may find the rest of the manual useful  )

I cannot remember if or why I change Q9. I do have a 576 curve tracer, so I would probably have removed the transistor and tested it out of circuit. Q9 is BU508A ( I recommend getting one from Digikey or somewhere that can be trusted).

Here is the bias supply schematic from the TDS520B manual:



It use TOP200YAI for U12.

Link to the datasheet: http://www.powerint.com/sites/default/files/product-docs/top200-204214.pdf

Regards,

Jay_Diddy_B

[vaualbus]

When i repair my psu i have also change q9 because if the tvs diode go bed the transistor could blow from overvoltage come from the transformer.
I have not test the transistor but i change it anyway.
 You are luck to not have revision b. In  it the transistor  U12 is a custom made switching transistor tha have a pwn controller inside.

[tbsmith]

Resurrecting an old thread...Anyone know where Jay Diddy got the schematic with the test values listed?  I have the TDS520B schematics, but they don't show test point voltages.

Thanks!

[siggi]

Artek Manuals. You can also try the TDS544/644 scans here http://hakanh.com/dl/kits.htm.

Sent from my Nexus 5 using Tapatalk

[tbsmith]

Awesome, thanks!

Tom