VOGONS


Capacitor replacement rules?

Topic actions

Reply 20 of 52, by fyy

User metadata
Rank Member
Rank
Member

OP here, thanks again for the responses guys. One more thing, why is it wrong (as a general rule) to replace a capacitor with one that has a higher capacitance? Wouldn't storing more charge be better generally (like having a larger battery)?

Reply 21 of 52, by Evert

User metadata
Rank Member
Rank
Member

From what I've read, it depends on the application, but 95% of the time it is better to stick to the original capacitance. I'm not well versed enough to explain why.

sigpic2689_1.gif

Reply 22 of 52, by smeezekitty

User metadata
Rank Oldbie
Rank
Oldbie
fyy wrote:

OP here, thanks again for the responses guys. One more thing, why is it wrong (as a general rule) to replace a capacitor with one that has a higher capacitance? Wouldn't storing more charge be better generally (like having a larger battery)?

As I said before, it depends but you can generally go a reasonable amount higher in filter applications.
Don't go too high though. If it is too big, it will have a large inrush which can damage upstream circuits.

Don't change capacitive values in resonate or coupling circuits unless you really know what you are doing.

Reply 23 of 52, by Jorpho

User metadata
Rank l33t++
Rank
l33t++
fyy wrote:

One more thing, why is it wrong (as a general rule) to replace a capacitor with one that has a higher capacitance? Wouldn't storing more charge be better generally (like having a larger battery)?

Storing more charge also effects the rate at which the capacitor discharges, which may have important implications for the functionality of the circuit. You can look at https://en.wikipedia.org/wiki/RC_circuit for some of the gory details.

Reply 24 of 52, by fyy

User metadata
Rank Member
Rank
Member
smeezekitty wrote:
As I said before, it depends but you can generally go a reasonable amount higher in filter applications. Don't go too high thoug […]
Show full quote
fyy wrote:

OP here, thanks again for the responses guys. One more thing, why is it wrong (as a general rule) to replace a capacitor with one that has a higher capacitance? Wouldn't storing more charge be better generally (like having a larger battery)?

As I said before, it depends but you can generally go a reasonable amount higher in filter applications.
Don't go too high though. If it is too big, it will have a large inrush which can damage upstream circuits.

Don't change capacitive values in resonate or coupling circuits unless you really know what you are doing.

Jorpho wrote:
fyy wrote:

One more thing, why is it wrong (as a general rule) to replace a capacitor with one that has a higher capacitance? Wouldn't storing more charge be better generally (like having a larger battery)?

Storing more charge also effects the rate at which the capacitor discharges, which may have important implications for the functionality of the circuit. You can look at https://en.wikipedia.org/wiki/RC_circuit for some of the gory details.

Thanks guys.

Reply 25 of 52, by Jepael

User metadata
Rank Oldbie
Rank
Oldbie
fyy wrote:

OP here, thanks again for the responses guys. One more thing, why is it wrong (as a general rule) to replace a capacitor with one that has a higher capacitance? Wouldn't storing more charge be better generally (like having a larger battery)?

I already explained it can throw off a control loop, like a switching power supply regulator feedback loop. Some switching power supply chips have a limit how large (or small) a capacitance you can put at the regulator output or the output voltage won´t be stable, it could start to oscillate or something, perhaps responding slower to quickly changing load currents, etc.

In basic linear power supplies, larger capacitance wastes space (money too) and decreases conduction angle by increasing the ratio of RMS current to average current. Because of higher currents flowing in shorter time, this increases transformer heating and stresses the rectifying diodes.

If the capacitor is not there to filter power but for something else, it can make time constant of a system longer, like slows down an oscillator or alters frequency response of filters or makes a voltage step response slower (for instance causing your speakers generate a huge "THUMP" instead of short snap when powering audio sources on/off).

So in general, if you have no idea what the capacitor does, keep the original value. If you do know what the capacitor is there for, then it is a different matter, and you are free to decide whether smaller or bigger capacitance will still work if you don't have an exact match.

Reply 26 of 52, by juglenaut

User metadata
Rank Newbie
Rank
Newbie

There is no I in team

Last edited by juglenaut on 2017-08-28, 22:53. Edited 1 time in total.

Reply 27 of 52, by juglenaut

User metadata
Rank Newbie
Rank
Newbie

There is no I in love...

Last edited by juglenaut on 2017-08-28, 22:53. Edited 1 time in total.

Reply 28 of 52, by juglenaut

User metadata
Rank Newbie
Rank
Newbie

It's been grit.

Last edited by juglenaut on 2017-08-28, 22:54. Edited 1 time in total.

Reply 29 of 52, by gdjacobs

User metadata
Rank l33t++
Rank
l33t++
Jepael wrote:

I already explained it can throw off a control loop, like a switching power supply regulator feedback loop. Some switching power supply chips have a limit how large (or small) a capacitance you can put at the regulator output or the output voltage won´t be stable, it could start to oscillate or something, perhaps responding slower to quickly changing load currents, etc.

Even substituting different ESR can result in a feedback circuit not behaving properly. The best bet is finding an approximate match for ESR and capacitance. Ripple, temperature, and lifetime can all be up specified freely.

All hail the Great Capacitor Brand Finder

Reply 30 of 52, by juglenaut

User metadata
Rank Newbie
Rank
Newbie

If esr is high at the frequency of operation desired, choose a slightly lower one.

A low esr is always desirable most times.

Reply 31 of 52, by Jepael

User metadata
Rank Oldbie
Rank
Oldbie
gdjacobs wrote:
Jepael wrote:

I already explained it can throw off a control loop, like a switching power supply regulator feedback loop. Some switching power supply chips have a limit how large (or small) a capacitance you can put at the regulator output or the output voltage won´t be stable, it could start to oscillate or something, perhaps responding slower to quickly changing load currents, etc.

Even substituting different ESR can result in a feedback circuit not behaving properly. The best bet is finding an approximate match for ESR and capacitance. Ripple, temperature, and lifetime can all be up specified freely.

My quote from two years ago? But you are correct and I think I've said in some other thread too, but for example LDO regulators require ESR to be within certain limits (not too high, not too low) for them to work properly.

Reply 32 of 52, by cyclone3d

User metadata
Rank l33t++
Rank
l33t++

I try to keep with the same capacitance at least, but it just depends on what I have on hand or what is available that I can get for a cheap price. I will not pay the outrageous prices from the online warehouses. I usually only buy batches of NOS high quality caps.

The thing about a lot of caps that need to be replaced is that the original manufacturer generally used really crappy caps in the first place which really were not what they say on them.

So even if you replace them with a better brand with the same rating, you are not going to get the same exact values of the original crappy caps that were on there.

For motherboards, I have often replaced with higher voltage and higher capacitance rating. I generally use low ESR capacitors.

Same goes for the caps in LCD monitors as well as computer power supplies.

I haven't had a problem yet.

Yamaha modified setupds and drivers
Yamaha XG repository
YMF7x4 Guide
Aopen AW744L II SB-LINK

Reply 33 of 52, by David_OSU

User metadata
Rank Newbie
Rank
Newbie
Kreshna Aryaguna Nurzaman wrote:

Well, my speakers and amplifiers are vintage stuff (JBL 120Ti and Sansui AU-7900, respectively). Last year I was going to buy replacement capacitors for the speakers' crossover, but the JBL 120Ti maintenance manual doesn't say anything about voltage rating and tolerance -only capacitance. Then I was told that using tighter tolerance makes better sound -less frequency overlap between drivers, so the sound becomes cleaner. Not sure if it's true or not.

Vintage JBLs already have high quality capacitors in the crossover. I wouldn't replace them unless you're sure that they have gone bad. If you do replace them, try to source replacements from an audio vendor that stocks parts specifically for those crossovers. You don't want to mess up the JBL magic.

Capacitor tolerance has nothing to do with sound quality. A deviation in capacitance will shift the filter cutoff frequency slightly, but I doubt you'd be able to hear a small deviation in crossover frequency. However, the type of capacitor (materials and construction) can have an effect on audio quality. There are specific capacitor types that are better for speaker crossovers, and I expect the JBL crossover already has them.

Reply 34 of 52, by juglenaut

User metadata
Rank Newbie
Rank
Newbie

Priceless

Last edited by juglenaut on 2017-08-28, 22:55. Edited 1 time in total.

Reply 35 of 52, by juglenaut

User metadata
Rank Newbie
Rank
Newbie

I don't like wasting my time.

Last edited by juglenaut on 2017-08-28, 22:56. Edited 1 time in total.

Reply 36 of 52, by gdjacobs

User metadata
Rank l33t++
Rank
l33t++
juglenaut wrote:
Rubycon, nichicon, kemet, cornell dubilier, panasonic, united chemi, Vishay, OHMITE, Toshiba TDK, yageo, Samsung ssl, mallory, a […]
Show full quote

Rubycon, nichicon, kemet, cornell dubilier, panasonic, united chemi, Vishay, OHMITE, Toshiba TDK, yageo, Samsung ssl, mallory, and last stackpole.

If it does not fall in this list I toss it and replace. I prefer rubycon and nichicon 105c

There are a lot of counterfeiters out there and going on ripbay is not on the list for me. Going cheap may open a huge risk, as there are people and individuals that manufacture knock offs, pay particular attention to the banding and base

I do not buy used or parts that are old stock, new or not that is older than 10 years. I do keep select recovered, and reformed components from appliances that have been in operation and usually only take what I need and must pass bench test before and after reforming, and have absolutely no corrosion

Sanyo/Suncon, Fujitsu, Elna (especially for audio) are generally between good/solid and great capacitors, application depending.

All hail the Great Capacitor Brand Finder

Reply 37 of 52, by sf78

User metadata
Rank Oldbie
Rank
Oldbie
cyclone3d wrote:

The thing about a lot of caps that need to be replaced is that the original manufacturer generally used really crappy caps in the first place which really were not what they say on them.

So even if you replace them with a better brand with the same rating, you are not going to get the same exact values of the original crappy caps that were on there.

Also the part in question might actually have design flaws that REQUIRE the use of a higher rated replacement parts. For example, original resistors might have lower rating than required and usually burned out after some time, so replacing them with the same value parts would only postpone the issue, not remove it. Repair manuals usually address these things, so if you can find one for the device you're repairing, use it.

Reply 38 of 52, by juglenaut

User metadata
Rank Newbie
Rank
Newbie

gdjacobs »

I do agree.

Sanyo is bar none for auto and home audio in both semiconductors and discrete components.

Sometimes it is difficult to find replacements in products as they seem to make components criteria specific to the manufacturers design and for a limited run or stream.

The manufacturers in your post are the ones I try to reclaim.

Reply 39 of 52, by Munx

User metadata
Rank Oldbie
Rank
Oldbie

I have a question about solid capacitors.

I'm replacing a couple of caps on this Radeon 8500 LE (it got quite a beating as the guy who sold it had it just rattling about in a box with a bunch of other cards) and one of them seems to have an inward dent in them:

The attachment P_20170826_104352.jpg is no longer available

Are these types of caps still able to work or should I replace it? If so, I assume its 47uF 6V? What does the 17 stand for?

My builds!
The FireStarter 2.0 - The wooden K5
The Underdog - The budget K6
The Voodoo powerhouse - The power-hungry K7
The troll PC - The Socket 423 Pentium 4