Electrolytic capacitor question

[flashthunder]

Hi, looking for input to a capacitor question. I am in the process of recapping a scope power supply. The scope is a Tektronix 2245B and built @1990. Some of the electrolytic capacitors that I purchased are physically smaller than what's currently installed in the power supply. For instance, new 10uf, 100v, 105degee electrolytic capacitor is on the order of perhaps 6 or 8 times physically smaller than original. Has 30+ years of improved technology allowed for same performance in a smaller package...or should I search for a physically larger packaged cap? Thanks

[techneut]

Yes the elco's are now a lot smaller. People who are restoring old equipment open up the big can capacitors, empty them and place new capacitors in them to keep the original look.

[axantas]

I did restore restore/recap video recording "machines" from the late seventies, replaced about 90 capacitors and ended up in front of some quite empty pcb's - size did not matter, so I just left it the way it was. One interesting example here:

[KungFuJosh]

Yeah, the size change is totally normal. One thing that annoys me is that some companies make larger electrolytics and charge extra for them to look like old style caps. Pop them open, and they're just newer / smaller caps in a larger shell. 🙄

[Zero999]

Why are you recapping? It's often unnecessary to replace all of the electrolytic capacitors.

Smaller isn't always better. Smaller capacitors might have a higher ESR and also won't be able to dissipate as much heat generated by the  ripple current, as well as larger ones.

If you must replace a capacitor. Choose one to match the PCB footprint, with a low ESR, high ripple current and temperature rating.

[flashthunder]

Thanks all...exact info I was hoping for! Zero999, I will not be replacing all caps...only those on the rails that show above ripple specs. Also, interesting about cutting the can open and installing another capacitor inside to retain original look...very clever.

[Zero999]

I always aim for voltage first, i.e. it needs to be equal to or greater than the part being replaced, then pitch, but note the can must fit and check it's not too tall, and finally value. For supply decoupling, the value isn't critical. It's okay to round down a tiny bit, so use 390µF, rather than 400µF, but not much, I wouldn't normally go under 95% of the original but it's usually fine to go much larger, but I wouldn't recommend going more than double.

[David Hess]

Aluminum electrolytic capacitors have gotten better so the same ratings are found in smaller cases, however the smaller cases have less surface area which lowers their ripple current rating.  To preserve the ripple current rating, I replace the original capacitors with higher voltage ones so that the case is the same size, or even larger if the placement allows.

[floobydust]

My philosophy is to shotgun and replace all the PSU electrolytic capacitors based on age. At 34 years old, some Spragues and Philips caps dry out. Not all brands last the same.

Unless you want to use time in your life troubleshooting and measuring ripple, taking the gear apart repeatedly. Some caps failing can cause damage, like in the SMPS. I find it cheaper to replace all in the  PSU and be done with drama.

I find the old +80/-20% tolerance spec for electrolytics now means manufacturers have it figured out. They deliver -15% consistently. I am talking Nichicon and Chemi-con are lower (rated) value than the old caps that are still good. So I upsize ~20% accordingly to match the original values.

The physical size is much smaller due to technology improving. If size is critical then higher voltage parts or long-life parts are larger.

[BrokenYugo]

Definitely do up the voltage rating to match case size and increase ripple rating. I've replaced 50 year old caps with smaller modern ones that ran warmer than the iffy looking original, not unacceptably warm for 105c parts, but enough to feel it and colder is always better.

I also concur with Floobydust's observation about new high grade Japanese caps always reading on the low end of tolerance. Rarely hurts to upsize a little since they were traditionally allowed +80% anyway. That will also get the case size up some.

[factory]

I've seen too many capacitors from the 1980s & early 1990s with failed seals, that leak corrosive & conductive electrolyte over boards, for me to leave them in place. Sometimes it's not obvious there is capacitor pee until the cap is removed.
Some PSU failures caused by this, then go on to cause major damage elsewhere.

Brands that I've seen leak include; Philips, Panasonic, Nichicon, Rubycon, Chemicon, and Sprague.
Last week we had a failure of critical equipment at work, with strange fault codes & random shutdown, a Matsushita/Panasonic cap was responsible, the capacitor pee had even found it's way through a filled via to the other side of the PCB, the equipment in question dates from 1994.

Going too small in physical size for replacements can lead to SMPS caps, with high ripple running hotter than the old ones, in extreme cases they vent, yes we had this happen at work when bean-counters choose the replacements.
Of course there are different cap technologies, that can be smaller & have lower ESR, but they come with their own failure modes.

David

[Xena E]

Just a Cautionary note.

Smaller cased electrolytic caps for any given ratings can just mean that further replacement intervals may be reduced as they dry out quicker; manufacturing has gotten so good that life specs are as close to reality as posted values, and very often I have found new 10uF alelcaps claiming 10% tolerance and measuring  just north of 9uF! Sneaky barstards.

*Lower ESR is not always a desirable feature*

Shotgun replacement of capacitors en masse, is a real good way of turning the job into scrap if you're not very careful.

Xena.

[Messtechniker]

Smaller isn't always better. Smaller capacitors might have a higher ESR and also won't be able to dissipate as much heat generated by the  ripple current, as well as larger ones.

To get around this, I sometimes replace one old large (mostly dead) capacitor by two smaller capacitors in parallel. Same total capacitance of course. This helps with ESR and heat dissipation.

[factory]

Just a Cautionary note.

Smaller cased electrolytic caps for any given ratings can just mean that further replacement intervals may be reduced as they dry out quicker; manufacturing has gotten so good that life specs are as close to reality as posted values, and very often I have found new 10uF alelcaps claiming 10% tolerance and measuring  just north of 9uF! Sneaky barstards.

*Lower ESR is not always a desirable feature*

Shotgun replacement of capacitors en masse, is a real good way of turning the job into scrap if you're not very careful.

Xena.

Yes every type of capacitor has different pros & cons.

And there is risk in leaving the original caps in place too, of course they might be of a type that is OK too.

If the OP has no experience of repairing/replacing parts, then it would be wise to practice on something of low/no value, with similar double sided or multilayer boards, before working on this scope.

David