Capacitor replacement question Harman Kardon Citation 22

[AndrewL]

Hi, I have a Citation 22 I bought new in 1989, that I took apart to chase down a DC output issue. I'm thinking the DC output may be related to cracked solder joints in the pre-driver board with loose transistors, as when I moved that board or banged it, the DC output would increase or go away. I'll know when I put it all back together.

While taken apart I saw some capacitors that looked really bad and pulled a few to test. It turns out the bad looking ones were mainly damaged by a corrosive glue that was used to hold the larger ones in place. They looked terrible, but measured fine, for example the 470 mf / 100V ones measured pretty much spot on with ESR of 0.5 ohm, even after 30 plus years. I did however find that most if not all the 10mf/50v ones were bad or measured very poorly. I also found a few bad 100mf/16v caps, a few bad 330mf/16v caps, two failed 47mf/10v, and two totally failed 2,200 mf / 6.3V ones. Once I took them out, I replaced them with fresh ones as it made no sense to reinstall 35 year old parts. I used Nichicon PW for power supply related ones (330mf, 470mf), Muse ES for the 100 mf coupling input caps. I do however have a question about 5 capacitors on each side that I am not sure if they are coupling, power related or what else. Initially I thought they may be coupling caps and I was going to use Nichicon ES bipolar caps, but I also have some Nichicon FG audio ones I could use. I don't know enough about electronics to understand the schematic. I'm not sure what these caps do, and which would be more appropriate, maybe a more power oriented cap like the PW are called for here. I would really appreciate some help from those of you that actually understand schematics and what things do. I am enclosing a picture of the position of caps in question.

The question is; what do the 10mf/16v caps highlighted do, and should I use Muse ES bi-polar (supposedly very good as coupling caps), Nichicon FG (fine gold), or instead use a power related capacitor like the Nichicon PW???

Thanks,

Andrew

 

[fred soop]

Have not had that problem with my almost identical Citation 24, but the Citation 21 had a lot of capacitors with insufficient voltage ratings. I got tired of chasing intermittent problems and eventually recapped 2 units. Problem solved.

Those marked capacitors in your unit are "speed up" capacitors that allow transistors on one side that are turning on to pull excess charge from the other side to allow quicker turn off for those transistors. They are often unnecessary but these Citations are high power, EF3 output (3 transistors - output, driver, and predriver) and these units have extended high frequency response, so the "speed up" capacitors are a definite help.

 

[AndrewL]

Fred, thanks for the response. I never ever would have figured out what those caps are for! I was surprised at how many of those caps were bad. I have another citation 22 where the meter was acting up, and in that one at least the meter section, they were mostly in good condition.

Knowing what those caps do, would you use bi-polar ones or polar ones?

Andrew

 

[loudnoises]

Knowing what those caps do, would you use bi-polar ones or polar ones?

Andrew

What type is currently installed?

There is no electronic benefit to replacing a polarized cap with a non polar, only the ability to use a different type of you so choose. If it's not in the audio path, it's not worth the muse or a film IMHO. Stick to one of the other nichicon you have. I doubt there's a huge difference in performance between the two at this value.

 

[Blue Shadow]

I am enclosing a picture

Where did you get your copy of the schematic? It shows the voltages in red and the HFE service manual didn't when I downloaded it a number of years ago. Having the red voltages makes it much easier to find.

 

[AndrewL]

What type is currently installed?

There is no electronic benefit to replacing a polarized cap with a non polar, only the ability to use a different type of you so choose. If it's not in the audio path, it's not worth the muse or a film IMHO. Stick to one of the other nichicon you have. I doubt there's a huge difference in performance between the two at this value.

Thanks for the explanation. The installed ones are polarized. I’ll go ahead and use a standard type and save the fancy ones. Thanks again!

Where did you get your copy of the schematic? It shows the voltages in red and the HFE service manual didn't when I downloaded it a number of years ago. Having the red voltages makes it much easier to find.

I ordered the schematics from HK back in 1993 when a transistor blew. I agree, the red helps. I would be happy to share, I would just have to figure out how to fit the whole thing.

 

[Blue Shadow]

I would just have to figure out how to fit the whole thing.

The SM has all chopped up diagrams in the online versions since each page was scanned and they were not stitched back together. But it is out there.

Getting the raw scans would work. I stitched the pages of the Sony STR-V4 SM to get the three-page diagrams on one page. Easy enough to enlarge on a pdf for readability.

 

[onanysunday]

UES Muse sounds better than all those other electrolytics and are often times prohibitively large- preventing their use. Will the effect of every Muse installed be audible? No, but trying to grasp which ones may be audible by studying the schematic isn't always as straight forward as it seems to unequivocally predict. There have been plenty of times I didn't think a cap would matter in a given position, but did. Bi-polar caps sound better in general and the Muse have copper clad leads which makes them sound possibly twice as good as regular polarized. Also, they do everything a polarized cap can do -even blocking. I suppose it would be possible that using them would introduce some noise? compared to the quietest, tightest tolerance, best spec caps (rated for power supply for instance), but I've never taken out a Muse because it didn't sound as good in any cap position with regards to clean/ quiet/ black background when compared to UHE, FM, FC, UPW, UKL, etc.

 

[AndrewL]

I agree, I have never had to take out an UES because it did not sound good!

 

[AndrewL]

The SM has all chopped up diagrams in the online versions since each page was scanned and they were not stitched back together. But it is out there.

Getting the raw scans would work. I stitched the pages of the Sony STR-V4 SM to get the three-page diagrams on one page. Easy enough to enlarge on a pdf for readability.

I think I have some time to scan these. How would be the best way to share?

 

[onanysunday]

I agree, I have never had to take out an UES because it did not sound good!

I've done so much swapping and comparisons of different series of e'lytics it's sickening. At this point, I've come to the conclusion to just use UES as much as possible for all sizes and voltages. It's saves any back-tracking that I would otherwise be doing much of the time with many other choices.

Eventually though, they start physically maxing out your board space and you have to use another series of capacitor with more capacitance and higher voltage that is smaller. But up until then, I recommend them to save yourself some time.

 

[AndrewL]

Update. The amp is now fixed, no more DC at the output. The problem was related to cold solder joints on the pre-driver board. There are two groups of 4 transistors (Q422, Q424, Q426 and Q428) that are attached to an aluminum heatsink and hang horizontally from the Pre Driver board which is vertical. The truth is it's a lot of stress for these joints. The solder joints on these transistors cracked so there was intermittent contact, and this would throw off the DC balance. I'm assuming when contact was lost, the DC would go out of balance and spike. That is why when I banged on it, the DC would increase or go to zero. I had to redo the solder joints on all 8 transistors. I sucked up the old solder and used fresh solder. After that the amp turned on and measured perfectly. I was able to adjust DC to close to zero on both channels.

While I was at it, I replaced every single capacitor, except the four main filter capacitors (10,000 mf / 100v) as so many of them measured badly. Many of them had failed and measured as leaky with very high ESR. The funny thing is the amp still operated and sounded great with all those capacitors that were out of spec. I'll note that the the four big 470mf/100v caps on the pre driver board, looked bad, due to the corrosive glue used to hold them in place, but they measured perfectly.