Tektronix DSA602A blows fuses at turn off

[alexd1]

I have recently acquired a Tektronix DSA602A mainframe from an Ebay listing.

Once I received it, I turned it on twice. The first time I ran the internal self test and kept the oscilloscope on for no more than a minute. The second time I let the scope perform the Enhanced Accuracy initialization routine as well set the correct time and date. I turned the unit off for a few minutes then back on to see if the time settings were saved. This third time, the unit did not start at all. I checked the fuse and it was blown.

I took the power supply assembly out of the instrument for a visual inspection but nothing stood out: no burnt components or traces, no bulging/leaking electrolytics and no weird smell. There are however 2 X2 rated EMI suppression caps across the power line that have slightly cracked cases. Those are about to get replaced, but if they were to fail, they would most likely open or short when voltage is applied at startup, I think.
After replacing the burnt fuse, the oscilloscope comes back to life, but something blows the fuse when the switch is flipped to the off position. I noticed that, for very short operating durations, like 30 sec to a minute, the unit can be turned off and back on, but if it runs for more than ~1 minute the fuse gets blown at shutdown.

I think the culprit is somewhere on the power supply assembly. Based on these observations, I assume there are some strong transients or pulses that have either very high amplitudes or the components that are meant to suppress the voltage spikes fail, thus blowing the fuse.

What issues could cause such behavior in an SMPS (or in case of a DSA602 if you had any similar experiences)?
Any tips will be appreciated.

[Johnny10]

Wrong Fuse type?

Manual shows 12A 250Volt

[alexd1]

No, sadly. All the fuses were 12A, 250V. That's what is indicated on the backpanel as well.

[Johnny10]

Still sounds like something in Fuse or Switch. Those ceramic Fuses take seconds to Blow. Your quick on and off switching could be heating them again before it has chance to cool?

If the scope works and doesn't bring up any errors? Makes me wonder?

I have both the DSA602 and DSA 602A. These things are beasts but I use them mostly for the Differential Plugins.

They also have the dreaded RIFA caps in them so watch out for that. I changed them in one of the units.
I think I lost the 602A somewhere in storage :-)

The 602A has FFT capability that the 602 doesn't. If you have the room I really like the scopes.

[alexd1]

It's a beefy oscilloscope, indeed. Wanted to get one for 3 years. Finally found this one for $200 and ended up paying twice that price for the shipping from US + customs fees.

As for the self tests, it passes all apart from the NVRAM option, which throws the error code E6111. I didn't get to look into that error. I hope that I won't have to replace the whole memory board.
The Enhanced accuracy routine passes as well, every time. However, I don't have any plugin modules to further test the scope beyond its self diagnostics. Other than this, I noticed the powered-on status LED on the front panel doesn't light up.

It's a good point, fast cycling the unit on and off could fry a fuse. But isn't a fuse not supposed to heat up during normal operation?

I decided to test the scope, by letting it run a few times for 30 minutes and a 20 minute break in between power on and off. It worked only for the first 2 attempts.
At first, I plugged it, turned in on, and it ran until I powered it off half an hour later. Unplugged it then, checked the fuse for countinuity and it tested ok. 20 min later, I put the fuse back, plugged it to the mains, powered it on again. After its 30 min run, I flipped the switch to off and 2-3 seconds later a flash occured on the power supply board. The fuse was fried upon inspection.

After the scope has cooled down for an hour, I replaced the fuse and plugged the unit to the mains without powering in on.
With the instrument still off, the fuse blew in 15 seconds. Maybe, are those RIFA's now shorting out the mains? (https://www.eevblog.com/forum/repair/smoke-und-fire-after-switching-on-old-electronics/msg1279950/#msg1279950). Fortunately, the other 2 fuses (F520 4Amps & F130 1.5Amps) on the inverter board are intact, which could tell that the issue is closer to the line input.

At this point I think the problem is most likely on the power supply as opposed to other modules inside the mainframe. Is it a good idea to recap the entire power supply section, including the rectifier and regulating boards?

PS: That's a nice setup you got there.

[Johnny10]

Oh Boy that's an old pic of my bench. 2017, Right before hurricane Irma came through and plopped a tree on my house!

If I remember I checked the caps on the power board and they were fine.
I used my Sencore LC unit to check for ESR also.

Changed those RIFA caps and then put that unit away after getting the TDS7104. It was a bit smaller and lighter.
I do miss the differential plug-ins though.
While checking floating grounds they are a must.

[tooki]

Still sounds like something in Fuse or Switch. Those ceramic Fuses take seconds to Blow.

That’s not how fuse ratings work. Any fuse, regardless of type, will blow in “seconds” at some particular current. How many seconds depends on the fuse’s time-current curve.

This fuse, ABC-12 (or its predecessor, 314012), blows in about 60 seconds at 20 amps, and is faster only with significantly higher currents (e.g. 1s at 40A, 100ms at 75A, 10ms at 130A). So if it’s blowing in 15 seconds, then it’s got around 22A flowing through it, far more than it should be seeing in use (never mind while off!).

Your quick on and off switching could be heating them again before it has chance to cool?

If a fuse is being operated so close to its thermal melting point that this matters, then the selected fuse rating is significantly too low.

The Enhanced accuracy routine passes as well, every time. However, I don't have any plugin modules to further test the scope beyond its self diagnostics. Other than this, I noticed the powered-on status LED on the front panel doesn't light up.

It's a good point, fast cycling the unit on and off could fry a fuse. But isn't a fuse not supposed to heat up during normal operation?

No, that should never blow a fuse. A fuse might heat up a bit — I mean, heating is how they blow — but as I said above, its design operating point shouldn’t be anywhere close to a value where thermal effects can cause spurious fuse blows.

I decided to test the scope, by letting it run a few times for 30 minutes and a 20 minute break in between power on and off. It worked only for the first 2 attempts.
At first, I plugged it, turned in on, and it ran until I powered it off half an hour later. Unplugged it then, checked the fuse for countinuity and it tested ok. 20 min later, I put the fuse back, plugged it to the mains, powered it on again. After its 30 min run, I flipped the switch to off and 2-3 seconds later a flash occured on the power supply board. The fuse was fried upon inspection.

After the scope has cooled down for an hour, I replaced the fuse and plugged the unit to the mains without powering in on.
With the instrument still off, the fuse blew in 15 seconds. Maybe, are those RIFA's now shorting out the mains? (https://www.eevblog.com/forum/repair/smoke-und-fire-after-switching-on-old-electronics/msg1279950/#msg1279950). Fortunately, the other 2 fuses (F520 4Amps & F130 1.5Amps) on the inverter board are intact, which could tell that the issue is closer to the line input.

At this point I think the problem is most likely on the power supply as opposed to other modules inside the mainframe. Is it a good idea to recap the entire power supply section, including the rectifier and regulating boards?

No, not until you know what’s actually failed!

[tooki]

P.S. If I’m understanding the service manual correctly, this unit has two power switches: the “principal power switch” on the rear, and the “on/standby” switch on the front. When you say it’s blowing the fuse when turning off, which switch are you using? What happens when you use the other one?

[Johnny10]

Of course you are correct tooki.
I should have been more specific. It was my notion that since the unit worked unless there was switching being done. That the fuse type could be the issue or a short in the switch itself may have been the first place to look.

I looked up the manual and the fuse datasheet and reviewed the open time chart.
Also since the OP only was a newbie I should have asked more questions.

[alexd1]

P.S. If I’m understanding the service manual correctly, this unit has two power switches: the “principal power switch” on the rear, and the “on/standby” switch on the front. When you say it’s blowing the fuse when turning off, which switch are you using? What happens when you use the other one?

I only used the on/standy switch on the front panel. I have no more spare fuses at the moment to check what happens if I turn it off from the rear panel.

Prior to your post, I tested the power supply outside the mainframe, with no load, running in the standby state. 30 sec after applying mains voltage, the gas discharge tubes / voltage surge arresters on the inverter board flashed and the fuse went out. Those 2 are placed between the diode bridge and bulk caps (image 1 - above and to the right of the 4 diodes). There is one more identical tube that doesn't flash (image 1 - top left of the 2 pink capacitors), and a neon one (image 1 - right above the white capacitor, left of the transformer), with an orange glow, flickering at a few Hz during and after operation. What could possibly cause such a voltage spike?

The fuse F520 (4A, 250V) (image 1 - above the bulk caps) is still alright, as well as F130 (1.5A) that is inline with the secondary of the transformer.

[Swake]

You have a fusion reactor :-)

Are you feeding that board with the correct mains voltage?
US is 110V. What mains voltage do you have in your place and is the power supply in the mainframe fit for that?

[alexd1]

You have a fusion reactor :-)

Are you feeding that board with the correct mains voltage?
US is 110V. What mains voltage do you have in your place and is the power supply in the mainframe fit for that?

I supply it with 220V and the voltage selector is set to 210-240V.
I'll try troubleshoot the power supply unit in the coming days.

[tooki]

You have a fusion reactor :-)

Are you feeding that board with the correct mains voltage?
US is 110V. What mains voltage do you have in your place and is the power supply in the mainframe fit for that?

Nope, USA has been 120V since the 1950s.