• Hello, Guest! Welcome back, and be sure to check out this post for more info about the recent service interruption and migration.

Performa 5400 capacitance test results – how do these look to you?

CircuitBored

Well-known member
Today I finished phase one of my Performa 5400 refurbishment project. This machine works well but has started to present a geometry issue with the display, so I decided to remove all the electrolytic capacitors and test the actual capacitance of each one versus its supposed values. Below are the results for the analogue board and the CRT yoke board. I have not gotten around to testing the PSU yet. All three of the aforementioned boards are LG branded. 

Picture 3.png

As you can see, there is a great deal of variance in how far each part has strayed from its intended value over the years. Absolutely none of these caps have leaked, much to my delight!

All the caps that are significantly below spec are Samwha brand. My gut is telling me to replace these first. The 1000uF caps are especially suspect as a cursory trace-tracing suggested that they are part of the horizontal and vertical sweep circuits. The display issue I have had involves an otherwise normal picture that is "wavy" along the Y-axis, so this discovery sort of connects to that in my mind. The "wave" is not static and worsens the longer the machine is powered up so I am certain that this is not a simple geometry calibration issue. 

My previous recap projects have all involved Macs with obviously failed caps so I've never had to do this level of troubleshooting before. I would greatly appreciate some input from someone more experienced than I. Which of these test results look worst to you? 

I am under the impression that electrolytic caps can actually gain capacitance as they age, especially if they are run too far below their intended voltage – is this essentially correct? Would it be prudent to replace the caps that are more than 5% above their intended values? Even with this criterion I would still be looking at a near-total overhaul of the electrolytics in this system. 

I may yet just bite the bullet and replace all the electrolytic and RIFA caps in this machine as a precaution but I'd love to hear some input from the folks here first. 

One final question: is it safe to power up the PSU of this machine without anything connected to it? I'd like to test its output but don't want to risk destroying it by assuming that it works the same as the PSU in my other Performa.

Thanks in advance to anyone who chips in!

 

CircuitBored

Well-known member
Here are some photos of the boards. These are really nicely made and have component labels on both sides - what a treat! 

From my cursory research it seems that the analogue and yoke boards share a great deal of components with the G3 iMacs/AIO. 

PSU:

IMG_2970.jpeg

Analogue/yoke:

IMG_2969.jpeg

It seems that this PSU has RF-shields/heatsinks obstructing a lot of the capacitors, much like the G3 iMacs. Grrr...

IMG_2972.jpeg

One oddball discovery: a random diode, which was hidden under a glob of hot glue on the backside of the board.

IMG_2973.jpeg

 

68krazy

Well-known member
What an incredibly thorough deep dive into the current health of your machine's capacitors — this is fascinating.  With all the debate about recapping in this hobby, actual data like this is a big step in the right direction.

I don't know how to interpret your findings, but, on my iMac G3 teardown and recap, five out of the seven capacitors that were leaking in the machine were all Samwha.  Mounting evidence that 90s Samwha caps were not very good?

 

68krazy

Well-known member
More thinking out loud — I wonder what is considered an acceptable tolerance for capacitor values?  And I wonder how your capacitors would read when tested for ESR and current leakage (not physical electrolyte leakage).

Badcaps.net had a pretty active forum about capacitors years ago.  I'm not sure if they are still around, or as active as they used to be, but they could be a good resource for us if we want to deepen our understanding of capacitors.

 

cheesestraws

Well-known member
I wonder what is considered an acceptable tolerance for capacitor values?


The usual tolerance for electrolytics, IIRC, is around ±20%.  They are not a precise component, generally.  So it's not unlikely that all the ones measured above are, technically, within spec, although the 118% one is cutting it very, very close...

 

CircuitBored

Well-known member
I wonder how your capacitors would read when tested for ESR and current leakage
I would love to know too - sadly my meter can't measure ESR! Even with an ESR tester it'd be hard to make use of any readings I made without the spec sheets for the capacitors unless I came across some crazily high values. 

It's a bit of a pickle because my issue is clearly made worse by the caps warming up, so they might all behave themselves while cold. Perhaps I should try warming them up outside of the machine and doing my measurements again. 

Mounting evidence that 90s Samwha caps were not very good?


This is a hunch that I might just go ahead and put to the test. It'll be a good first step to replace the caps that are both off-spec and from a vendor known to have caused issues in other machines from the same era. I think my plan for now is to replace the Samwhas (and the caps that are more than 10% over spec) and see what happens, before moving on to more thorough testing. It is rather frustrating that none of these caps have their tolerances written on them. 

 

CircuitBored

Well-known member
The usual tolerance for electrolytics, IIRC, is around ±20%.  They are not a precise component, generally.  So it's not unlikely that all the ones measured above are, technically, within spec, although the 118% one is cutting it very, very close...


It's hard to decide whether it's a good or bad sign that both the 22uF 16V caps are identically inaccurate. Are they well-made caps that have held their slightly-wrong value, or less well-made caps that have slipped up in sync? The former strikes me as more likely but I'm far from being a capacitor expert. 

 

CircuitBored

Well-known member
@68krazy I have done a little more research and it seems that Samwha is not a very well-regarded brand at all. The company doesn't even exist any more!

I am definitely going to be replacing those parts with nice, high-quality caps. 

It strikes me that it would be very useful to compile a list of which Macs have bad caps in them that won't catastrophically fail but will still cause issues. By now we all know which Macs contain caps that are a threat to the entire machine but it's far less well-known which machines have components that are simply... iffy? 

 
Top