• Updated 2023-07-12: Hello, Guest! Welcome back, and be sure to check out this follow-up post about our outage a week or so ago.

Capacitor Recommendations for SE/30

smrieck511

Well-known member
I'd like to make sure I'm getting the best ones. I know I need (10x) 47uf 16v and (1x) 1uf 50v. Which are considered high quality and reliable? Are tantalums the best option? Not currently planning to do the axials. Thanks for any help.
 

3lectr1cPPC

Well-known member
If you get tantalum caps, get 25v caps for the 47uf ones instead of 16v. There’s a complicated reason why that I don’t fully understand, but yes.
 

croissantking

Well-known member
Why not go for polymer tin can caps, they look authentic and won’t leak.

I prefer to buy from the premium manufacturers - Vishay, Wurth, Panasonic, Rubycon, etc.
 

cheesestraws

Well-known member
Macs aren't particularly picky. Tantalums you have to be a bit careful about because they can be a bit incendiary and, ethically, tantalum is a conflict mineral so it's best to get them from a reputable manufacturer who can tell you where they got the tantalum from.

I've been using tantalums recently, purely because that's what I have in stock. I know other people have used polymer tin can caps like @croissantking suggested, and I've heard of people using MLCCs too. I can't see any reason why you couldn't use something like niobium-based capacitors if you felt hipsterish, either.
 

smrieck511

Well-known member
Here's what I'm thinking about.
Are these a genuine fire hazard?
Does the 1uf tantalum cap also need to be a higher voltage than normal (50v) like the 47uf's?
 

Attachments

  • Screen Shot 2023-03-09 at 3.56.19 PM.png
    Screen Shot 2023-03-09 at 3.56.19 PM.png
    246.8 KB · Views: 26

3lectr1cPPC

Well-known member
There's no fire hazard about tantalum caps. In 30 years they might short out and pop, but the tiny bit of fire for a couple seconds won't do anything inside a metal or plastic computer case. The 1uf cap is fine.
 

mitchW

Well-known member
There's no fire hazard about tantalum caps. In 30 years they might short out and pop, but the tiny bit of fire for a couple seconds won't do anything inside a metal or plastic computer case. The 1uf cap is fine.
There is an issue. Tantalum caps have much greater voltage derating. In simple words, tantalum capacitor that "sees" 12V needs to be more than 16V rated, that one should be 25V. Other types of cap technology don't have this problem. Apple used same rated capacitors (16V 47uF) on multiple voltage rails (like 5V, 3.3V and 12V). So if you knew which capacitors are on which voltage rails, the you could put 16V tantalums everywhere, except on 12V rail.

For the fire hazard, I got a recapped Mac II back once that experienced smoke and fire at power on. It worked fine before, the customer got smoke the second it turned it on. One new tantalum cap shorted. Was able to repair it afterwards, of course.

Myself, I have used 16V 47uF Tantalum caps on anywhere with no ill effects, except on this one. .
For the new capacitors, I will probably order Polymer solid state ones, which proved to be better than Tantalum anyways.


And yeah:
 

smrieck511

Well-known member
Thanks everyone. Yes JDW's video is comprehensive to say the least. I learned alot after i posted the cart screenshot above.
 

3lectr1cPPC

Well-known member
There is an issue. Tantalum caps have much greater voltage derating. In simple words, tantalum capacitor that "sees" 12V needs to be more than 16V rated, that one should be 25V. Other types of cap technology don't have this problem. Apple used same rated capacitors (16V 47uF) on multiple voltage rails (like 5V, 3.3V and 12V). So if you knew which capacitors are on which voltage rails, the you could put 16V tantalums everywhere, except on 12V rail.
I did mention this above in an earlier reply about the 47uf caps, so yes, no harm in using 25V caps for all of them. That’s a great explanation of why as well, didn’t type one myself as I didn’t want to get anything wrong. Not too confident in my knowledge of that sort of thing.
 

smrieck511

Well-known member
Where I'm confused is it seems JDW was saying solid tantalums can have as much as 50% tolerance. The solid 25v tantulums I posted in the screenshot above say 20% tolerance? So they would seem ideal if they're not too big?
 

avadondragon

Well-known member
The tolerance value in the screenshot is the extent to which the actual capacitance is allowed to vary from its nominal value. Thus a 100µF capacitor with a ±20% tolerance could legitimately vary from 80μF to 120μF and still remain within tolerance. It is unrelated to the voltage derating which is 50% for normal tantalums. If you look specifically for Polymer Tantalums they should have a 20% voltage derailing. This voltage derating value isn't something that's generally listed in the part description.
 

smrieck511

Well-known member
These look ok? 16v is ok with polymer just not solid tantalums since there is only 20% voltage derailing?

80% of 16v = 12.8v...so it seems like its cutting it close.
Any reason not to go with 25v polymer...slightly more expensive but are they also larger?
 

Attachments

  • Screen Shot 2023-03-09 at 8.52.48 PM.png
    Screen Shot 2023-03-09 at 8.52.48 PM.png
    287 KB · Views: 21

avadondragon

Well-known member
Yep. That's correct.

They could be larger. That looks like Mouser in the screenshot - You'll have to check the "Case Code - mm" section of the datasheet for the part you're looking at to see what size it is. I think 7343 is the right size.

As a side note: I recently bought a bunch a parts myself and I had originally planned to do a Digikey or Mouser order but after loading everything in my cart I went and checked Console5 and they were shockingly cheaper for the Mac related stuff I was ordering. They also specifically made note that they were including 25v polymer capacitors for C3, C4, C9, C10 in their SE/30 kit which made me feel pretty good about them knowing what to provide.
 

smrieck511

Well-known member
Yep. That's correct.

They could be larger. That looks like Mouser in the screenshot - You'll have to check the "Case Code - mm" section of the datasheet for the part you're looking at to see what size it is. I think 7343 is the right size.

As a side note: I recently bought a bunch a parts myself and I had originally planned to do a Digikey or Mouser order but after loading everything in my cart I went and checked Console5 and they were shockingly cheaper for the Mac related stuff I was ordering. They also specifically made note that they were including 25v polymer capacitors for C3, C4, C9, C10 in their SE/30 kit which made me feel pretty good about them knowing what to provide.
 

mitchW

Well-known member
My hands on experience with Apple laptops:
For the polymer caps, on the MacBooks that run 12.6V PPBUS_G3H (main power rail), they used all 16V caps. And they hold very well, no instances of being shorted or anyhow degraded over years. And those machines also run very hot, something that isn't a case with vintage Macs.

But some MacBook Airs also used tantalum caps on their PPBUS_G3H, well those are different story, and they sometimes short, albeit rarely. I think they used 12V rated caps (or 16V?) on 8.6V PPBUS_G3H. And 15-inch 2010 MBPs had tantalums in the GPU power supply that failed (got high ESR) and that caused black screens, etc. The solution was to replace it with Polymer cap.
 

smrieck511

Well-known member
Think I've got this right now but wanted to run it past you guys first. I appreciate the help.
------

I'm gonna order Tantalum Polymers.

Physical Size:
The 47uf's are Case Code D (7343)
The 1uf is Case Code B

The big question ....
If I just order 25v for all the 47uf caps, will that be any detriment at C1, C5, C7, C8, C12, C13 (the 47uf's that only see 5v) ?

Also...
C3, C4, C9 and C10 can see 12v so they definitely should have 25v.

C6 is the 1uf 50v (Case Code B)...(no need to up the voltage rating on this one?)

C2 and C11 are the axials. I may replace those keeping the old leads intact. Gonna do the surface mounts first and test the board before I do that.
 

Attachments

  • Screen Shot 2023-03-10 at 11.32.48 AM.png
    Screen Shot 2023-03-10 at 11.32.48 AM.png
    273.6 KB · Views: 19
Last edited:

s_pupp

Well-known member
As an experiment, I once tried using 22uF MLCCs and a 0.47uF MLCC. I put in a 50MHz Daystar 68030 accelerator. The SE/30 ran without problems. I am NOT advocating for this approach, but it is evidence that choice of capacitor type is not a critical issue. I would strongly advise paying attention to the voltage derating advice for tantalums as given by others above.

I plan to recap with aluminum polymer caps, in order to maintain an authentic appearance.
 

chiptripper

Well-known member
Re: the derating thing, in JDW's thread @Michael_b breaks down what the failure risk looks like with a 16v tantalum seeing 12v looks like, at least in real world usage in an SE/30 circuit. Very low.

That being said, MrFahrenheit has had 16v tantalums (Apple factory no less) go boom on IIfx boards, and @mitchW has run into the issue on a Mac II board. So, unless there's a compelling reason not to 25v is a good default.
 
Top