View Single Post
Old 10-23-2010, 02:33 AM   #7
Grumpy Old Fart
PCBONEZ's Avatar
Join Date: Aug 2005
City & State: Phoenix, AZ
My Country: USA
Line Voltage: 120V 60Hz
I'm a: Professional Tech
Posts: 10,630
Default Re: Suggested replacements

Physical size. - I think I was misunderstood.
If you look at any given series the physically larger caps will have lower ESR and higher Ripple ratings.
The uF and volts don't affect the ESR and Ripple ratings, only the can size does.
If you then look at a different series you will see the same pattern -but- it does not work going from one series to another. [Unless they happen to be the exact same grade.]
See the attached data sheets. - Look at the ESR & Ripple for all the 8x16mm, then 8x20mm, then 10x16mm, then 10x20mm.
See what I mean now?
ESR and Ripple change with the can size and the series. - Not uF or volts.

They are all important for various reasons.
The best thing you can do is know -why- each is important so you can 'juggle intelligently' when you have to choose alternates.

The Ripple current spec is like the amps spec on wire.
It tells you how much it can handle without over heating.
You DON'T want a smaller Ripple rating than the original cap.
- - A higher Ripple rating is fine.
Higher Ripple is effectively 'more heavy duty'.
There are no 'over-kill issues' with the Ripple rating.

The ESR tells you how easy it is for Ripple to get through the cap.
It is like a Resistance.
Since the caps job is to pass Ripple through itself to ground less 'resistance' is better.
- - A Lower ESR value is better.
[Lower without getting crazy... In some circuits reducing it too much can be bad. The capacitors work with the inductors and other capacitors in the circuit and changing one without changing the others can 'de-tune' the overall operating characteristics. Most circuits are not super sensitive to reduced ESR and reducing by 30% or so won't noticeably affect anything. But that is not -always- the case and as this is a general statement with no specific circuit in mind the possible exception/problem needs to be stated.]

Most of the time uF should be kept the same if at all possible.
- Because it's usually EASY to find something with the right uF...
That said: In general filter circuits are probably designed for something like +/-50% of ?some? uF, but since you don't know how close ?some? uF is to what actually got installed it's best to stick close to what you found there.
If you MUST use another value going up is preferable.
Going up no more than 20% is usually very safe. After that it's hard to say.
Reason 1 - uF goes down as caps age with moves you closer to the original value.
Reason 2 - Almost all Low ESR electrolytic caps are +/-20% for uF. As such, the circuit it is found in should support a range of +/-20% of the uF marked on the original cap. [Now read Reason 1 again.]
- There are specific situations where reducing uF is considered okay but for the most part that is getting into 'Modding' and is out of the realm of simple cap replacement.
- If you are -seriously- 'stuck' and the ooooonly option is to go to a lower uF then keep it within the +/-20% uF [of the original] and you have a fair chance things will be okay. Don't do this just to be lazy because as the cap ages the uF will drift even lower. In other words that choice may come back to bite you later on. Probably at the worst possible time.
- When/if you have to lower uF in a VRM, lowering the ESR drastically is okay and a good idea. [Only in a VRM.] That is basically what happens when Polymer are used in place of electrolytics in a VRM. [And it's a well tested and popular Mod.]

Volts. - The volts rating is a -limit-.
The ACTUAL volts in the circuit must be less than the voltage rating of the cap at all times.
- - Using higher voltage caps is okay.
[10x higher than the CIRCUIT voltage would be over-kill but should still work. Contrary to popular opinion more than that might not.]
- - Using lower voltage caps is ONLY okay IF you KNOW the voltage in the circuit NEVER goes higher than the caps voltage rating. Over-volting caps can kill them in short order.
[Note: If you use a cap on a particular voltage for a while it -becomes- that voltage cap. For example if you take a 25v rated cap and use it on 5v then after a few hours is -becomes- a 5v cap. If you then put it on 25v it will be over-volted and possibly damaged permanently. This is because the oxide layer on the Aluminum foil adjusts it's thickness to the applied voltage through the electrolysis type chemical reaction between the electrolyte and the Aluminum Foil. There is a procedure to return it to being a 25v cap called 'reforming' but that is beyond the scope here.]

Temp rating is just the max temp where all the caps specs will stay within the advertised specs.
- There is however this little detail.
For each 10C rise in cap temp it's life is reduced by 50%. [Works backwards too.]
Assuming a 105C cap rated for 1000hr and a 85C cap rated for 1000hr.
The 105C
At 105C - 1000hr
At 95C - 2000hr
At 85C - 4000hr
At 75C - 8000hr
At 65C - 16,000hr
At 55C - 32,000hr
The 85C
At 85C - 1000hr
At 75C - 2000hr
At 65C - 4000hr
At 55C - 8000hr
Now look at both for 55C, a possible real case temp. - Get the point?
1000hr at 105C is 4x more life than 1000hr at 85C.
Note 567:
The Lifetime shown in data sheets is -NOT- LifeSPAN.
The correct [full] term in the data sheet is 'Endurance Lifetime'.
[I may discuss 'Endurance' later or you can look it up here in an old thread. Been discussed several times. I don't feel like all the Math for that right now so please push the 'I believe' button that says Endurance Lifetime does NOT tell you anything useful about how long the cap will last.]
It was easier and valid to illustrate the temp-life dependence because the % change of Endurance with temp will be the same as the % change of Lifespan with temp.
Attached Files
File Type: pdf MBZ_.pdf (88.8 KB, 206 views)
File Type: pdf MCZ_.pdf (62.7 KB, 96 views)
Mann-Made Global Warming.
- We should be more concerned about the Intellectual Climate.

Be who you are and say what you feel, because those who mind don't matter and those who matter don't mind.

- Dr Seuss
You can teach a man to fish and feed him for life, but if he can't handle sushi you must also teach him to cook.

Last edited by PCBONEZ; 10-23-2010 at 02:50 AM..
PCBONEZ is offline   Reply With Quote