Scope keeps blowing fuses

The video shows testing of two Y caps. These were originally factory tested at 1500VAC/2700VDC, and have no reputation for embarrassing field failures after 20+yrs.

X2 caps are desiged for 'impedance-limited' applications and are tested at around 1KVAC/1500VDC.

Case degradation is a significant factor in aging of early production. Prior to single-board assembly mfring philosophy, these parts were shipped with long leads, loose, in bulk containers, knocking heads all the way. They were then subjected to short and long term lead stress in manual lead dressing and final physical attachment.

Rifa was first to use that distinctive clear lacquer. I'm not sure how many times that formula was 'improved', but parts would arrive with visible cracks and would crumble around lead wires during lead-forming.

That being said, there were sufficient 'incidents' in the first

20yrs of use to ensure that nobody in their right minds would use them in new product, even in a list of alternates, in the last 25 years.

TEK scopes that employ these parts are antiques, maintained by enthusiasts, who will tell you all about the effects of heat, humidity and just plain age on plastics, epoxies and other components/materials used in them.

RL

Phil,

You will not be able to purchase the 'exact' replacement, as the series number no longer has the same operating voltage markings, safety approval listing, pert numbering or material construction.

After all the merging and take-overs are done, it's unlikely to be manufactured at the same physical plant, or using similar equipment or materials, though safety docs used to be pretty inflexible in that regard - one of the main trade effects of VDE/IEC safety regime.

Costs reflect a low volume mfring environment due to buyer reluctance to bother with 'improved' parts.

RL

Yandex viewer is blurry, even when viewed as "original size".

Try downloading the PNG file and viewing it with a better viewer at

100% magnification. The details are quite readable in Irfanview (Windoze) and Nomacs (Linux). In Irfanview, the image can be "Sharpened" by hitting "S" on the keyboard.

Ditto.....I was going to suggest this but some ppl want to find *the* problem. For *vintage* stuff, it makes no sense to me, Replace all the caps, check all resistors. Having a solid power supply is critical for the rest of the unit to work properly, so why screw around with replacing only one bad component? As long as you got it apart.....Do you really want to revisit this later when another cap dies?

Here we go, gentlemen, the finished "repair" on the PSU board:

Before (the faulty RIFAs are arrowed since they're a bit hard to spot)

After (the new replacements are not arrowed, since they're blindingly obvious)

It's just awful, isn't it?

I've done worse. I don't like the exposed hot wires, but since you're the only person likely to be working on the scope, it's probably an acceptable risk.

I suggest you tack the capacitors to the PCB with a little REMOVABLE glue, such as hot melt glue or non-acetic-acid RTV.

Probably won't fit back in the scope cos of the IEC filter. Tons of info out there on this power supply and how to fix:

It does actually fit back in again - *just* about.

Many thanks for that indeed.

Do you know why these caps usually have their self-resonant frequency marked on the side of them? That's always puzzled me.

Fail short...really? Why would you want a cap that is across the supply line to fail short? Just to make sure the fuse blows? It is my understanding that X and Y capacitors experience line spikes, and will endure multiple small breakdowns over the years as a result of this stress. They fail open because local heating caused by a microscopic dielectric breakdown results in polypropylene melting, which seals off the "edge" of the breakdown. Is this not correct??

No idea if that's the failure mechanism or not. However, if you read my orginal remark, I did say they are *supposed* to fail short, which makes pretty clear that was the intention of the manufacturer even if they were not successful in many instances.

I got a handful of brand new PME271 Y caps a few years ago, for reasons that I'm sure can't possibly have been good ones. All of them have developed a couple of small cracks in the casing, just from sitting unused in a parts box.

So if anyone was wondering if they'd fixed the defects in the design at any point in the last 40+ years, apparently not.

Yep, they're a menace all right. These days I pre-emptively check any mains-powered electronics from the 80s and 90s for them, which thankfully meant I got them out of my Tek 2445A before they popped. Wish I'd been more alert with the 2235A; I spent ages cleaning sticky brown crud out of the crevices of that one.

And then there was the HP spectrum analyser where I thought I'd saved myself a heap of trouble by extracting all eight(!) Rifa capacitors from its power supply. Plugged it in, and half an hour later a thick cloud of truly foul phenolic smoke poured out. Opened it up to find that it had blown part of the case off its power inlet. Turns out that Schaffner filtered power inlets can *also* contain Rifa (or very similar) caps, concealed inside a metal casing and potting compound just to add extra force and noxious smells to the eventual and inevitable explosion.

I'm pretty sure anyone who works on older electronics eventually develops their very own rant about these accursed objects.

Rayner

You could probably have leadformed the new parts to suit.

My guess is you just prefered to solder from the comp side.

RL

It's really not my thing. When people say, "Just recap anything suspect" and there might be fewer than half a dozen to do, the prospect still fills me with horror. If they only knew how ironic the "Just" bit strikes me in suggestions like that.

The problem with vintage equipment, however well designed and well built, is that parts within it age. Typically, items that are approaching, or exceeding 30 years old will have some marginal parts within. Further, if those parts are 'common' as in "many of them", if one fails the others will not be far behind. Some truisms:

a) As in the Holland America commercials - Time is a precious commodity. b) Most (not all) hobbyists are not limited by cost constraints for a few parts. So, a US$20-or-equivalent investment to save a valuable item is not outrageous. c) Most of the effort in repairs is 'getting to it' - the taking-apart, the testing, documenting, and then reassembly. d) The actual repairs take very little time.

So, when whatever the device might be, when it is taken apart, *THEN* is the time to refurbish the entirety to the greatest extent possible - to avoid having to repeat the process if nothing else. Analogy: the engine in your vehicle spins one (1) bearing. You would replace _ALL_ of the bearings, as the incremental cost of the additional bearings against the cost of the tear-down is tiny, and the rewards significant. There is a school of thought that suggests that repairs are made only to the minimum necessary. Subscribers to this theory are the ones that one sees stranded on the side of the road in a blinding snowstorm.

Your power-supply went through a traumatic event - and it is supplying a complex and expensive piece of equipment. Consider it in that light.

As to soldering and technique - time and repetition will give you more confidence.

Peter Wieck Melrose Park, PA

That would be Viking....

"The actual repairs take very little time."

If you only knew the irony of that statement when the task is completeed by someone with my eyesight and fine motor skills.

No it won't. I've been doing this for more than 50 years and I'm still every bit as shit at it as I was half a century ago. I have not improved one iota and only a damn fool would believe the future's going to be any better than the past with a skill set like mine.

The key to faster resolution of fusing faults is test method for fault isolation.

Possible methods include:

1. Input voltage attenuation below fault threshold and probe test points.

1. isolate secondary sections by disconnect

2. Check for thermal hotspots safely.