Lambda 71 power supply restoration

[cos000]

Hello to everyone
I'm Giulio from Triest (italy).
I'm restoring an old Lambda 71 power supply and, after replacing few tubes that had low emission and the resistors that were way out of specs I came across capacitors
I have few oil capacitor cans (two of them are multi-section).
I tested them with a Heathkit IT-28 capacitor tester and the leakage indicator isn't stable but opens and closes. It's like flickering.
I can test them only at 600V while few of them are rated 1000 or 2000V.
I connected in series with the capacitors under test a multimeter and it showed few uA of leakage.
So my question is: should I replace them or I should't worry about?
If you need some other information just ask and I'll try to give them.
Thank you so much
In pictures there are the two multi-section caps.

[jonpaul]

Bonjourno, If there is no physical oil leak, and the capacitance is OK, the 600V leakage test will be OK, unlikely that 1000 V test is much different.

The Heath intermittent results are a puzzle.

What is the symptoms of the power supply problem? What is the DCV and ripple ACV  across the caps in operation?

These are quite reliable.


Ciao,

Jon

[cos000]

Hello
Thank you so much
They don't look physically leaking. Capacitance seems ok.
Power supply didn't have problems: I got it and I checked that everything was in spec.
I put the voltages into the attached pdf file.

Thank you so much

[coppercone2]

its probobly electostatic forces opening a piece of foil connection, there is probobly damage inside

Is the ESR low?

I would say the amount of force generated inside by the high voltage is very low, and if that is causing it to open, it has something that is barely making a good contact inside, and that they are damaged. Think about the internal construction, it seems obvious to me what is making it go open circuit inside! its made with stuff like sandwich foil 


Potentially very dangerous because this can happen: they get more damaged inside, the foil disconnects after its charged temproarily, you test it with a HVdmm and its reading 0 after 'discharge', so you think its OK, then it recontacts and your fucked if you happen to touch it.

TO keep a POS like that in service you would need to measure how much charge you put into it before service and then measure how much charge comes out of it when you drain it to make sure they are roughly the same to be sure its really gone. Personally I woulds stab a hole in it and throw it away or try to take it apart and see if it can be fixed.

 This is a normal source of really bad EMI also, loose contacts inside of HV stuff that arc or jitter during operation, say when there is a high ripple current, it turns into a oil submerged spark gap jitter switch with really low impedance ! This is what happens to a power company when they do a bad splice job in a 10kV line, it makes a crazy spark gap transmitter that operates on mechanical principles, often underground or in a inconvenient location, going open and close circuit with a very low source and load impedance, hopefully slowing opening itself without starting a fire or causing too much interference. It can really happen on any contact but it seems to manifest itself on higher voltages because it can arc.

Also, there might be soot building up inside of that capacitor, aka dirty oil.. maybe its becoming slowly conductive with each arc.

I would say: do not use until fixed or replaced. Going from 600V to 1kV and getting a open circuit because something is fluttering is NOT a reliable connection, the change in force is not that great. There is probobly a bad spot weld inside or something like that, not sure how they put them together, I imagine the very thin foil is a good candidate for spot welding .

[BrokenYugo]

Whatever you do keep in mind those are probably PCB oil filled, handle carefully and dispose of properly. Definitely don't dismantle them or disturb the seal.

I'd replace between that and the weird test results. I'd guess if they're doing funny stuff at 600 volts they're gonna do even funnier stuff at higher voltage.

[coppercone2]

oh yeah there is a hazard of PCB, that is probobly the worse hazard, totally slipped my mind but if its going open and closed circuit for no reason while its charged.. its a dangerous device. Don't make a PCB spill, its bad and might cause a circus of various things to appear, but do keep in mind intermittent behavior at higher voltages can easily be a sign of future failure. I forgot how old this thing is.

Wow, in my mind you stumbled on a giant pain in the ass. Spilling a bunch of PCB is one of the ways to get a hobbyist into trouble, IMO bigger problem then mercury because of the oil based nature, it can still leak or explode even if its fine..

You might not even want to keep a PCB oil device around if its fine, because I would say that's one of the toxins that does not have an overblown hazard label by any means.

I assume they have modern replacements that can fit in the same form factor? If not, and they are otherwise too big, you can always use sheet metal and rivets to expand the chassis, those old shell designs are pretty simple to stitch together, don't feel bad, they even do it to giant cargo ships to make them longer, its quite common. Since it has a curved bend radius its in your favor since you can practically fabricate the parts with a chair leg and a rolling pin.

And defiantly don't try to refill it or drain the oil if it does infact have PCBs if you think that might be a good idea, it is not a good idea, you will be left with a bunch of contaminated shit.

[cos000]

Thank you so much.
I'll keep them safe and find a proper place for disposal.
Anyway what replacement should I look for?
I find theese caps:
https://www.mouser.it/ProductDetail/80-F872DU254K480L
https://www.mouser.it/ProductDetail/647-PHC1604400KN
https://www.mouser.it/ProductDetail/647-PHC1604200KN
https://www.mouser.it/ProductDetail/80-C4AQQBU4660M12J
https://www.mouser.it/ProductDetail/505-MKP4-2.21KV10P37
https://www.mouser.it/ProductDetail/647-PPA2123100KJ
https://www.mouser.it/ProductDetail/80-F43KR3500ZA01K
https://www.mouser.it/ProductDetail/80-R76UR310050H3J
Should they be fine as replacement?
I choose the same DC voltage. What I'm afraid of is the AC one.
What do you think?

[coppercone2]

hmm, I am not an expert on these old things, but.. has technology with capacitors advanced that far or is there some kind of drawback with the newer ones  that are not oil filled.

rms current ? transient response?

[coppercone2]

https://www.antiqueradios.com/forums/viewtopic.php?f=6&t=215981

scroll down here and read what alan douglass said, it seems that you picked the replacements ok.

when they give you 480Vac you renember this is RMS
1357.6319999999998 Vpp = 480VAC RMS

but your not spending all that time in 1357V, only on the peaks, so they give you a RMS value that is normal so you can make a decision based on SOA

https://filmcapacitor-st.com/techconsiderations.php?id=Plastic-Capacitors-versus-Oil-Impregnated-Metal-Case-Capacitors

The only difference I can tell is that they are not shielded like the oil ones.

I think you are OK, I was just taken aback how much smaller the new ones are (not that I did not know but when I started lookin at em side by side it got me thinking that its a little weird).

I think you are OK, the region where you need oil filled capacitors I think nowadays moved to the 10kV range, so PP took over this segment in most cases

[cos000]

Thank you so much.
I thought the same thing about AC Voltage: It has also to be considered that theese caps are not used in AC circuit but they are after rectifier tubes so they take voltage from 0 to Vp of the rectified sine wave, right?

[BrokenYugo]

I think the AC rating has to do with using it in a phase shift application (motor run cap), not applicable here.

I'll hazard a guess that any film capacitors of proper capacity and DC voltage rating (you may even go up a bit) will do fine here.

[coppercone2]

Treat the AC voltage as 480Vrms, or 1340Vpp, or +- 700VDC

I don't know why its derated for AC, perhaps because they expect AC to be poorly regulated and spikey so they just give a general OK for 480VAC (standard industrial power kept within standard industrial regulations). Your results may vary  for a shady indian steel mill but I figure they mean like ... reasonable for a factory. It might be from a generator, spike, be connected to a dirty grid, so the 30% derate is probably a 'safe bet'.

i.e. got a reasonable 3 phase 480Vac machine? you can use these capacitors

1000Vpp would be 700VRMS, but these systems don't commonly exist, and when you have 480V ac industrial power you expect serious abberations and 'poor behavior', so if the loads were all scaled up from 480Vrms to 700Vrms, you might get these capacitors start to fail when subjected to industrial factory abuse.

The standards go


480V Voltage between phases in USA in commonly used 3 phase distribution

600V Three phase power voltage

690V Three phase power voltage used in industry for larger electrical motors (Europe)

when I read "larger electrical motors" when 480V is called "commonly used", I think of a different world.

The common definition is:
Voltages 600 V and below are referred to as “low voltage,” voltages from 600 V-69 kV are referred to as “medium voltage,” voltages from 69 kV-230 kV are referred to as “high voltage” and voltages 230 kV-1,100 kV are referred to as “extra high voltage,” with 1,100 kV also referred to as “ultra high voltage.”

So I would say.. they are being safe by saying its good for 'commonly used' voltages with common loads, and when you start getting into 600V+ territory you are entering 'medium voltage' which is kind of a different world, with different pay scales, different outlooks on safety, etc. They don't want this wimpy shit going into medium voltage world so they cap it off at the middle of the 'low voltage' range. When there is medium voltage around you got different procedures, more training, more pay, more serious accidents, etc.. so they will want to use special parts for this that are probobly engineered a little differently.

I would say I imagine when it gets past 480Vrms you start getting a little 'psycho' with the designs and parts that makes them much less economical. Your load impedance also change with these larger machines (they go to dangerous 600V for a VERY good reason, cost and ease of generation), so the parts probobly need better withstand voltages and pulse ratings and shit like that. Just think of it as a different world with phenomena like 'explosive arc flashes', insurance premiums, etc. It's kind of like CAT rating in a meter. You can have a high voltage rating but a low CAT rating (i.e. for testing a air ionizer or helium laser power supply). Capacitors don't have CAT ratings, so they just assume when its more then 3 phase 480Vac industrial, they want better parts, with more testing and better QC, so they just tell you 'dont go there' with a general derating to cover their ass.

[coppercone2]

I.e., look at this part. When they give it a 700VAC rating (for the 'hard core' industrial shit), you would expect the DC rating to be.. 2000V

But look at the part, when you put the 'heavy industrial fudge safety factor' into the equation, your 700VAC capacitor is rated for a tame 6000VDC.
https://www.mouser.com/ProductDetail/WIMA/FKP1Y023307G00KSSD?qs=RB4whv9F6rxg5UN5DdkzEA%3D%3D

So from that part if you go by 3000Vpp, you would expect to get roughly 2200V AC RMS, but the part is rated at 700VAC, so there is a difference of 1300VAC rms, that gives you an idea of what kind of abuse the '700VAC' power in a industrial area dishes out! at 480VAC you get 70% of DC, at 700VAC you get 30% of DC!

The correlation is nonlinear, as voltage of industrial generators increases, the other parameters don't track very well!

Alot of this comes down to seeing what short circuits look like and what kind of damage they can do... its fucking dangerous

the manufacturer is using the VAC rating as 'how tough is this fucking component'./