Replacing 10,000uf filter caps with 22,000uf?

Alan,

Noone is saying the inrush limiter is a bad idea. But not everyone is at the same skill level to start modifying all sorts of parts. You posted in another thread you think that it's bad advice to encourage others to even work on their gear because of the potential to damage the boards if soldering skills aren't super. Just run something til it fails and buy a new one was your basic premise rather than fix what you have or do preventative maintenance. So what is it?

The OP here is relatively new to working on gear (no offense meant z). He's learning rapidly, but the question posed in this thread shows that he could have done a bit more research into increasing capacitance, where, when, by how much, etc.

You're an EE and are comfortable designing your own amp, yet have still had some issues doing so. But, it sounds like you expect everyone to just follow along with your advice, and if they don't, or have a different way of doing things, you take offense. H/K hired engineers to design their amps and they chose a different way than you seem to be heading. No need to get offended, not everyone does everything the same way.
 
For detail, go back to my thread on protecting the switch, we are hijacking this thread.

Putting the inrush is just cutting the hot wire of the AC input and put the inrush in series. If you can change the cap, you can put in the inrush. But again, it's up to OP. I just put in my opinion. Please move the discussion in my thread and we can go a lot further.

I had issue, I solve the issue. This is engineering. You don't have all the solution, You ask, you listen, you think, you research and you learn and you solve the problem. Not just taking other's word on face value.
 
Last edited:
I would go with a 12000µF minimum which would approximately equal installed 10K from back then. New large capacitance caps seem to be made so well and accurately that the manufacturer can provide a ±20% spec and then build them at about 10-15% low and save money. This is what I asked Conrad about and there was a discussion to this point in the Capacitor Abyss. Any large caps, I now order a value that will provide at least the scat design value at -15% of the rated value. Of course, going up from that some won't hurt anything since the new caps are so close to the old caps and a small bump isn't going to do much. A big jump and the in-rush current and bridge rectifier might cause problems.
 
This is just slightly less un-specific than asking us if the stuff in the red tupperware in your refrigerator is OK to eat.
You just made me think of George Carlins take on leftovers in the fridge. Could be meat, could be cake, meat cake? GC Ask wife,is this good?, wife, smell it, GC, it doesn't have any, wife, then it's good. I wonder if those albums are on cd, off to check now.
 
There are those here that are quoting my post.

To be clear I gave my reasons, and they are not a criticism of what Alan has done with his amplifier. I mentioned my OCDness as a factor wanting me to make everything perfect.

To quote myself...

"Only you can tell if this will make an audible difference. If I had put as much time, effort and money into designing an amplifier as you have, my OCDness would struggle with accepting this. That does not mean that I am right in terms of the impact if any on sound quality."

Please do not change the meaning of what I have posted.

In simple terms, per the OP's original post and in general, although it can vary from one piece of gear to another, doubling the amount of power supply capacitance may hasten the failure of the power switch and have a negative impact on the power transformer and power supply diodes.

There are several methods to help ameliorate the negative impact of increasing power supply capacitance.

The math and science of power supply design is well know in the field of engineering. One thing we look at is the load step response of the power supply. We may do something to improve it on the secondary side, only to have do something on the primary side (something simple) that works against the improvement made on the secondary side, for instance helping to protect the switch and other components from an increase in turn on current surge. Or we may choose something more complex to help maintain the improvement on the secondary side of the power supply.

Anecdotal examples may relate what happened in one situation, but will not necessarily apply to other situations.

For example two amplifiers may both have switches that are rated for 15 amp of current, but one of them may be rated for only 35 amps of surge current and the other one may be rated for 100 amps of surge current. Doubling the power supply capacitance on one may hasten the failure of the switch and on the other not so much.

To paraphrase Sargent Joe Friday, it is all about the facts.
 
Inrush current in a capacitor can be calculated in the following way: I=dV/dt. Looks simple provided that one knows how to calculate dV.
For the simplest PS circuit shown in attachment with no load can one use the following simple formula:
I peak inrush=(Vac*1.414-Vfd) / Resr
Vac=AC voltage;
Vfd=forward Voltage drop across rectifying diode;
Resr=esr (equivalen series resistance) of the rectifying capacitor/capacitor bank.

Do not forget that any PS uses a transformer and that also determines Inrush currents which also has to be calculated.

Use of the NTC inrush current limiter is the simplest solution and anyone with sufficient soldering skills can do the job knowing which limiter to use.
Much better solution is to use a good soft start circuit.

cheers,
PS generally, one cap having twice the capacitance of a bank of two has higher Resr than two smaller caps of the same type paralleled together. To draw the right conclusion look at the formula.
 

Attachments

  • simple PS.jpg
    simple PS.jpg
    9.8 KB · Views: 11
I prefer to put it at the primary and use the inductance of the primary to soften the blow. I first started using a 15A, 5ohm inrush, I got a drop of 1VAC across the inrush. I was working with I LOVE MUSIC, he suggested the drop is a little high, now I ordered a 2.5ohm I think is 10A inrush from Mouser. This will give a little higher inrush current of 115V/2.5ohm=46A max. But the hot resistance is lower and drop less voltage in normal working temperature. I don't think it's too bad. I am still waiting for the new one to arrive. Ordering from Mouser is kind of frustrating, some times they come in 2 days, this has been over a week already. My project is in a stand still waiting for the parts!!!!

I think it is good idea to put a 10ohm in series with a 0.022uF RC across the contacts of the power switch also for arc protection regardless whether OP decide to increase the capacitance. But as I said, check the amp first, there might be inrush and RC arc protection built into the amp already. We don't know what amp or anything, we are guessing here.

Maybe I rush it too fast to suggest to increase the capacitance, but if the amp is valuable to the owner, I would put the protection regardless of increase the cap or not. YOu protect the switch and the rectifier diode.
 
The Citation 16 uses 10,000uf filter caps and I am in the middle of rebuilding one and I read about a guy on here using 22,000uf caps instead and that peaked my curiosity. An eBay seller that sells a rebuild service says this on the eBay ad...

http://www.ebay.com/itm/Harman-Kard...270648?hash=item35c9a48b38:g:-RIAAOxyOa9R4wvd

Our Citation 16 restoration service includes the following items and more:

United Chemi-Con 22,000uF low ESR main power supply filter capacitors for a total of 88,000uF. A common misconception is that the larger capacitance will tighten and increase bass response. Not necessarily true... but it will help prop up and sustain the unregulated power supply voltage rails from sagging momentarily under demanding high power output peaks. Based on the fact that the Citation 16 can produce up to 250-watts RMS per channel (and more) into 4-ohm loads before current limiting, these capacitors are quite appropriate. Current amplifier design recommends about 6800 to 8000 uF per 100-watt capability.
 
United Chemi-Con 22,000uF low ESR main power supply filter capacitors for a total of 88,000uF. A common misconception is that the larger capacitance will tighten and increase bass response. Not necessarily true... but it will help prop up and sustain the unregulated power supply voltage rails from sagging momentarily under demanding high power output peaks. Based on the fact that the Citation 16 can produce up to 250-watts RMS per channel (and more) into 4-ohm loads before current limiting, these capacitors are quite appropriate. Current amplifier design recommends about 6800 to 8000 uF per 100-watt capability.

Hmmm, interesting. Makes sense.
 
Just make sure you know what the "and more" side of the restoration includes besides the 22K caps is all we're saying. If they're also putting in beefier rectifiers, a triac across the power switch, etc, then that should also be done to yours if you go with the higher capacitance.
 
Wrong, In rush current limiter limit the peak current. That's what any good amp would do. I looked at my Acurus, they did exactly that.

In what world limiting the in rush current is a bad thing? You have to know the electronics, not just listen to other people. A 2.5ohm inrush limit the peak current to about 40A. You listen to people's opinion and think. This is common engineering sense.

If you want to, go back to my thread and we can debate instead of hijacking the thread here. I stand by my suggestion that inrush limiter WILL protect the switch, can handle more capacitance. I stand by more caps will make the amp sounds better. THAT if you are tinkering with changing caps, it's should be very easy adding an inrush limiter along the AC main input line.

If you worry about the rectifier, change to a bigger one!!! Again, if you dare to change the cap, why stop there...... If you want a better amp.

+ 1000 + and a .01 ceramic 1 Kv disc cap. across the power switch contacts will go along way in snubbing contact arcing like the more foresighted manufacture`s did/do :thumbsup:..
 
Last edited:
FWIW folks, I had to replace my main B+/B- caps in my Mac MC 2205 over two years ago..
They were originally 39,000 uF @ I forgot the voltage)..
But I after being in the repair business for over 40 years decided to up the capacitance to 100,000 uf per cap.@ Mac`s specified voltage or higher Chemcon replacement..
The only issue I had since then was I didn`t check and replace the in rush Thermistor.
And it probably was failing anyway, like my MC 225 was doing..

Anyway, after replacing the in rush Thermistor with a 12 amp one, the MC 2205, as was before the cap replacement, is a daily driver and is turned on when I wake up until I shut down for the evening(typically 16~18 hrs.) which is every day 7 days a week, 365 days a yr. with zero power switch/relay issues, and playing along happily as I type..

Just MY experienced opinion here doing this up capping of various SS/tube audio equipment over 40 years during the course of repair/reworking audio equipment..
No Dog to fight in this controversy but only my empirical experience.. Take care. Kind regards, OKB
 
my 2 cnts;
Bill Ferris makes a good points.. I've read if talking ps ss caps, member 'leesonic' chimed in /posted other day per some op's project and he doubled his on same model. I'm not familiar with this OP's ps layout nor wattage consumed, but if on the basic side could use just two instead of four if has a scope to check the w/form. I've also read many manuals where some ps's were woefully under uf, too as well as above 36k uf is unnecessary. Good ps layout would be my consideration if unit has a good sized transformer usually associated with large cap values.

I always suggest to newbies is to read lots of service manuals and I usually focus on the ps as where is all starts.. after a while you get a good comparative idea of quality designs and whats in the box. I believe I coined the phrase; "the capacitor abyss" :D
 
my 2 cnts;
Bill Ferris makes a good points.. I've read if talking ps ss caps, member 'leesonic' chimed in /posted other day per some op's project and he doubled his on same model. I'm not familiar with this OP's ps layout nor wattage consumed, but if on the basic side could use just two instead of four if has a scope to check the w/form. I've also read many manuals where some ps's were woefully under uf, too as well as above 36k uf is unnecessary. Good ps layout would be my consideration if unit has a good sized transformer usually associated with large cap values.

I always suggest to newbies is to read lots of service manuals and I usually focus on the ps as where is all starts.. after a while you get a good comparative idea of quality designs and whats in the box. I believe I coined the phrase; "the capacitor abyss" :D

Thank you Sir for the kind words. :thumbsup: And yes there seems to have been quite a few "Light Days Panty Shield" level PS`s :no: used in sound reproducing equipment over the years.. OKB
 
Good experience shared is what these forums are about.. I appreciate your knowledge as just not many w/ years of hands on experience vs. theory.
-----
"Light Days Panty Shield" level PS`s LOL... been a few amps I've been scared to approach... w/o considering wearing adult size 'pampers', too.
 
Good experience shared is what these forums are about.. I appreciate your knowledge as just not many w/ years of hands on experience vs. theory.
-----
"Light Days Panty Shield" level PS`s LOL... been a few amps I've been scared to approach... w/o considering wearing adult size 'pampers', too.
I hear ya. Thanks.
 
I've also read many manuals where some ps's were woefully under uf, too as well as above 36k uf is unnecessary. Good ps layout would be my consideration if unit has a good sized transformer usually associated with large cap values.

Well, it's like a lot of things. The range goes from what is necessary to accomplish the design goal to what one might do in massive overkill simply for peace of mind. I'm not aware there is any sort of hard limit to necessity/unnecessary in a general sense. It depends on many considerations. For example, just a couple here; a single ended Class A amp will need more capacitance than a similarly rated push pull because there is more constant heavy load on the PS driving the ripple higher and no benefit of any cancellation.
 
whoaru99; hell you've been here at ak for thousands of post replies!! (many probably lost when the new site format kicked in?)

Not blowing air up someones ego but I know who's who. Self admitted I'm no expert hands on for all the models you dedicated repair guys have but I do read (speed read) short of going brain dead. Just trying to focus on tech help.
 
whoaru99; hell you've been here at ak for thousands of post replies!! (many probably lost when the new site format kicked in?)

Not blowing air up someones ego but I know who's who. Self admitted I'm no expert hands on for all the models you dedicated repair guys have but I do read (speed read) short of going brain dead. Just trying to focus on tech help.

I'm not for sure what your post is supposed to mean, but it comes across bad to me.

If you took mine as some sort of "you are wrong" type of thing, you have misinterpreted the point of it.

Perhaps I have misinterpreted the point of yours?
 
Not at a Sir! I read/ read; any of your posts aside those gone under the pages. You come up on a lot of searches, too. :D Please hang in here!
bink :D
 
Back
Top Bottom