Scope keeps blowing fuses

First thing to check is the rectifier; that diode bridge can fail, and apply AC to the filter capacitors, and... that'll take out the fuse. Not sure why the RIFA capa goes, though; maybe just a power surge some time last week?

There seem to be some surge-suppressors (gas discharge?) in the diagram, they might have failed short in a surge..

Yes, they're each showing about 70k ohms, so can't be responsible.

One thing I forgot to mention in my first post was that a wire had come adrift from the cooling fan's supply (poor factory soldering clearly) so this unit had been running for an unknown length of time with no fan. Not sure if that's relevant to the fault here, but it certainly *could* be....

Thanks for the links, Jeff; I'll check 'em out...

---------------------------------

** Really ? Where did you get that idea?

Plenty are found open or with low values after long times in service. X2 dropping caps used in smoke alarms regularly fail open.

But those Rifa ones love to smoke and burn.

.......Phil

RIFA caps smoke, especially the old ones. This is typical for vintage Tektronix scopes and RIFA caps:

"Old RIFA capacitors and a disaster story"

YouTube videos featuring smoking RIFA caps:

I sometimes test the HV leakage in capacitors using an ancient Megger:

Unsolder the capacitor for testing. You really don't want a high voltage power source connected to your oscilloscope.

Don't know your line voltage, but if it's 240, check the input voltage setting on the scope.

If RFI caps are replaced and bridge is ok, check big electros (C1021, C1022) and main switch (Q1050) for shorts.

At 240V, a leaky bulk cap can blow the gas tubes.

Once a gas tube fires, it's breakover voltage will reduce, so must be replaced, too.

RL

The smoking caps were metalized paper construction. These were good enough in their day, with a demonstrated mtbf and safe self-healing, while seals lasted. The outer sealing body tended to shrink and crack with age, allowing ingress of air and moisture, compromising performance. They are no longer made.

Metalized film parts are usually less dramatic in self-clearing of internal faults, but in extreme cases, you still get thermal runaway, as the faulting area expands and eats away at succesive layers of the wrapped structure.

Film parts will generally fail at the end schoopage joint, where metalization and film stresses accumulate in mfring.

RL

Hi, everyone (including the original poster).

When the power supply unit fails, there can be a basic myriad of possible problems and/or failed components.

Here are some things to look for:

Current limiting resistors, rectifier diodes, transformers, transistors, caps., filters, coils, etc... Shorts and opens are common problems. Look for bad connections, cold solders, etc... Lots to check for to "pin in down". Especially with older devices. Since it is an analog device (as was said), it is assumed the equipment is old, so we have to go to the old school way of checking everything, looking for key things, in order to come up with some accurate diagnosis.

Good luck and have a great day.

Charles Lucas

I know about the RIFA caps exploding - thanks for the interesting video tho. What is the recommended replacement type? (assuming they can physically fit into the same space on the board) I have a few vintage Tek scopes that I should do some preventative maintenance work on.... J

When I test RIFA capacitors with the Megger, I sometimes see capacitors the show some leakage for maybe 250 msec and then recover. I've been assuming that this demonstrates that self healing is function. Known good capacitors don't do that. Since testing required removing the capacitor from the PCB, I just replace any capacitor that shows temporary leakage. (i.e. better safe than sorry).

This video shows what might be two self healing RIFA capacitors in action. Watch the current on the display: "40 Year Old Rifa Capacitors High Voltage Leakage Testing"

(3:09) The effect is more visible with a panel meter than with a digital readout (except when the meter is moving because I'm turning the crank on the Megger).

I haven't bothered to perform a post mortem on a RIFA capacitor. I have several suitable microscopes and would expect to find holed in the dielectric (paper or plastic) and the metalization. The RIVA caps that failed have warped, cracked, melted, or crumbled outer plastic cases. For meaningful microscopy, I would need an old RIFA capacitor in an intact case.

"RIFA - Replacement Choices"

Kemet bought RIFA. My older Tek scopes survive 30 to 40 years and usually blow electroltyics before RIFA caps. Rather than experiment, I simply bought the same value/voltage X2 "safety" caps from Kemet (via Mouser or Digikey):

I wasn't very careful when I ordered replacement caps many years ago. I think I ordered "metalized paper", which in retrospect, was not the best choice. Going down the list of dielectrics, only "metalized paper" and "metalized polypropylene" are available for "safety" caps. Kemet is recommending their R53 X2 caps with "metalized polypropylene film" dielectric.

X2 is for "line to line" filtering while Y2 is for "line to ground" filtering.

OK. What is known:

a) Scope is blowing mains fuses. b) RIFA -branded caps blew up. c) The power-supply board was affected.

Suggestion:

1) Remove the power-supply board. 2) Replace all, repeat: ALL the capacitors on that board of any nature, type or style with the correct (but modern) replacement type. 3) Check each other component on the board. Diodes, regulators, transistors, whatever. 4) Replace any that are either bad or the slightest bit questionable. 5) Replace board - carefully (now might be an opportunity for a DBT to 'do its thing' - but a metered Variac would be better).

I kinda-sorta think that will do it. KISS and all that.

Peter Wieck Melrose Park, PA

On Sun, 23 Oct 2022 11:16:19 -0700, Jeff Liebermann snipped-for-privacy@cruzio.com wrote: (chomp...)

When in doubt, replace everything. I was wondering if someone sells a recap kit for the Tek 2465. Yep:

It includes some RIFA -> Kemet replacement caps.

Jeff, remind me to post a picture of the cap replacement I've carried out. You'll be both horrified and astounded. If I had more time I'd post it now, but it'll have to wait for the time being..

I do have a metered variac, so plan to use that in series with DBT current limiting, as the bulb can react to sudden over-current much more quickly than I can. I'll slowly bring the variac up whilst checking for anything getting warm on the board. That's the best I can manage as I'm no tech repairer; haven't got the necessary fine motor skills, I'm afraid.

Not a problem. Let a hand soldering robot do the precision work. Something like this soldering robot:

More: and even more:

------------------

==================================================

** Rifa PME271 X2 caps are still made and widely available. Strangely they are also the most expensive X2 caps on offer for no good reason. Described as being of "metalised , impregnated paper " construction. Rifa also claim to be the only supplier . Other makers, dozens of them, use metalised polyester or polypropylene films as used in DC rated caps - but with one big difference. X2 types are double wound = a technique that creates two caps in series.

Such caps are immune from *corona discharge* occurring in tiny air pockets trapped inside the cap as it it being wound. They are still vulnerable to spike voltages of a few kV which eat away at metalisation and reduce cap value over time.

IME Rifa X2s are notorious for catching fire and filling the room with acrid smoke. Happened twice, right in front of me, first case was a room heater and the second with a portable TV.

When used for suppression in triac dimming equipment, they invariably fail early too. Bad news.

..... Phil

Yikes. When a fuse blows, you replace it *once*. If it blows again, it ain't the fuse or a line transient.

Start with a DMM and check for shorted diodes in the input bridge and then for a shorted filter cap at its output. (The schematic is more or less completely illegible, or I'd give the ref des.)

Next most likely is a switch FET.

Cheers

Phil Hobbs (proud owner of a 2467 that he almost never uses)

Yeah, well twice in this case. Replacing the RIFAs *usually* does the trick (I have to say, in a half-arsed defence of my appalling conduct).

Not sure what you're viewing it with, Phil. Others don't seem to have that problem and have been able to see the part numbers okay.

Yes, the chopper is always a prime suspect with SMPSs