XPost: sci.electronics.design
From: bill.sloman@ieee.org
On 5/03/2024 8:17 pm, Cursitor Doom wrote:
> On Mon, 04 Mar 2024 16:41:40 -0800, john larkin wrote:
>
>> On Mon, 04 Mar 2024 23:03:05 +0000, Cursitor Doom
>> wrote:
>>
>>> On Mon, 04 Mar 2024 09:16:53 -0800, John Larkin
>>> wrote:
>>>
>>>> On Mon, 04 Mar 2024 15:55:53 +0000, Cursitor Doom
>>>> wrote:
>>>>
>>>>> On Sun, 03 Mar 2024 18:28:46 -0800, John Larkin
>>>>> wrote:
>>>>>
>>>>>> On Mon, 04 Mar 2024 00:31:27 +0000, Cursitor Doom
>>>>>> wrote:
>>>>>>
>>>>>>> On Sun, 03 Mar 2024 16:22:12 -0800, John Larkin
>>>>>>> wrote:
>>>>>>>
>>>>>>>> On Sun, 03 Mar 2024 23:50:08 +0000, Cursitor Doom
>>>>>>>> wrote:
>>>>>>>>
>>>>>>>>> On Sun, 03 Mar 2024 15:16:12 -0800, John Larkin
>>>>>>>>> wrote:
>>>>>>>>>
>>>>>>>>>> On Sun, 03 Mar 2024 22:41:09 +0000, Cursitor Doom
>>>>>>>>>> wrote:
>>>>>>>>>>
>>>>>>>>>>> On Sun, 03 Mar 2024 13:28:05 -0800, John Larkin
>>>>>>>>>>> wrote:
>>>>>>>>>>>
>>>>>>>>>>>> On Sun, 03 Mar 2024 18:03:51 +0000, Cursitor Doom <
d@notformail.com>
>>>>>>>>>>>> wrote:
>>>>>>>>>>>>
>>>>>>>>>>>>> Hi all,
>>>>>>>>>>>>>
>>>>>>>>>>>>> Using my Peak ESR/Capacitance meter, I was carrying out
in-circuit
>>>>>>>>>>>>> checks on large electros in the linear PSU I've previously
mentioned
>>>>>>>>>>>>> here. I was getting a lot of "in-circuit/leaky" warnings for two
of
>>>>>>>>>>>>> them, so I pulled them out completely and am getting the same
warning
>>>>>>>>>>>>> when they're checked out of circuit, which surprised me as it's
>>>>>>>>>>>>> unusual IME. Anyway, the leak would have to be very bad to
result in
>>>>>>>>>>>>> ripple, would it not? AFAIK, the leading culprit for ripple is
caps
>>>>>>>>>>>>> which have lost a significant amount of capacitance or else
developed
>>>>>>>>>>>>> a very large ESR. Do I have that right?
>>>>>>>>>>>>> I'd like a "second opinion" as it were on the leakiness of these
caps.
>>>>>>>>>>>>> What's the best old-school method for testing for this? I just
want to
>>>>>>>>>>>>> ensure the ESR meter isn't faulty (highly unlikely but the
possibility
>>>>>>>>>>>>> must be eliminated to be sure).
>>>>>>>>>>>>>
>>>>>>>>>>>>> CD.
>>>>>>>>>>>>>
>>>>>>>>>>>>> PS: the aforementioned caps are 47000uF 16V Vishay ones - and I
have a
>>>>>>>>>>>>> brand new spare that's also testing as "leaky"!
>>>>>>>>>>>>
>>>>>>>>>>>> How leaky? You could apply 16 volts and measure current for a
while.
>>>>>>>>>>>> It will typically taper off over some minutes or hours.
>>>>>>>>>>>
>>>>>>>>>>> Well, the 'perfect cap' in series with an ammeter will cause an
>>>>>>>>>>> initial surge of current which will taper off over time and
eventually
>>>>>>>>>>> settle at zero. But a leaky cap will continue to pass a small
amount
>>>>>>>>>>> of current, I would assume, indefinitely? Again, I would guess
that a
>>>>>>>>>>> real-world cap in good condition would continue to pass a tiny
amount
>>>>>>>>>>> of current - a negligible amount?
>>>>>>>>>>
>>>>>>>>>> An electrolytic is hardly a perfect cap. After the theoretical cap
>>>>>>>>>> charge, you might see some mA of leakage, tapering off to uA's after
>>>>>>>>>> some minutes or hours.
>>>>>>>>>>
>>>>>>>>>> If the current ever increases, as it will at some voltage above
rated,
>>>>>>>>>> it's probably on its way to destruction.
>>>>>>>>>>
>>>>>>>>>> Get a power supply and an ammeter and experiment. Your capmeter is
>>>>>>>>>> obviously not telling you much.
>>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> Just spotted the meter only covers up to 22,000uF!
>>>>>>>>> we
>>>>>>>>
>>>>>>>> I don't trust C or L meters, especially for large C or L values, or
>>>>>>>> cheap meters.
>>>>>>>
>>>>>>> How would you test for leakage, then?
>>>>>>
>>>>>> Power supply and DVM.
>>>>>
>>>>> But then how do you determine - given that electrolytics come in all
>>>>> sorts of votlage and temperature ratings, capacitance values etc - how
>>>>> much leakage current in each case is "too much" leakage current
>>>>> rendering the cap unsuitable for use?
>>>>
>>>> That's for you to decide. No instrument is going to have red and green
>>>> LEDs to tell you if a cap is suitable for your circuit.
>>>
>>> That's not how cap testers work, though. They take virtually none of
>>> the following into account:-
>>>
>>>> Gross capacitance
>>>> C vs voltage
>>>> C vs temperature
>>>> Leakage vs temperature and voltage, both polarities
>>>> Dielectric absorption
>>>> Failure voltage or current
>>>> ESR vs temperature
>>>> ESL
>>>> Lifetime
>>>> Power dissipation/cooling
>>>> Solderability/washability
>>>>
>>>> One should measure or calculate whichever of those might matter in
>>>> your circuit.
>>>>
>>>> Post your circuit and we can talk about it.
>>>
>>> No circuit involved. The question relates to generic smoothing caps of
>>> linear power supplies.
>>
>> Where in a circuit is a "smoothing" cap? Oops, sorry, there is no
>> circuit.
>
> Yes, there's no circuit. Marconi/Aeroflex never published one and and
> a lot of their designs are still under wraps. It's a real PITA.
If you are repairing the units, you should be able to trace that bit of
the circuit. It isn't going to be complicated.
--
Bill Sloman, Sydney
--- SoupGate-Win32 v1.05
* Origin: you cannot sedate... all the things you hate (1:229/2)
|
