Please help me understand - opamp power supply question
Sep 7, 2008 at 9:41 PM Thread Starter Post #1 of 14

AudioPhewl

500+ Head-Fier
Joined
Jun 20, 2008
Posts
641
Likes
12
Hello all,

Can someone please explain the following to me. Burson Audio offer their discrete-component-opamp modules here, though I suspect the same answer would apply to any opamp installation:-

Burson Discrete Opamp

On their "Discrete opamp 101", they list a mod of solder a good quality foil capacitor (MKP caps) across the power intake (V+ and V-) of the Discrete Op-Amp. You can try different value capacitor between 0.1uf to 1uf. Different value and brand will yield slightly different result however; the rating of the capacitor should not lower than 50V:-
Burson discrete Opamp 101

To my mind, using an good electrolytic in this instance will make a lot more sense. Firstly, it will provide additional filtering of the power supply feeding the opamp replacement. Secondly, it will provide a far bigger power store for the opamp to tap into on demand.

Am I missing something? What would the indicated installation have over one utilising a basic, low-ESR electrolytic?

I've been talking to others about this, and feel I must be missing something rather obvious - but for the life of me, I can't see the logic in using a small foil cap in this instance...

Thanks... I'd love to lift the veil of confusion over this!
smily_headphones1.gif


~Phewl.
 
Sep 7, 2008 at 9:45 PM Post #2 of 14
A 1uF capacitor is a 1uF capacitor is a 1uF capacitor...as far as power storage goes, anyway.

The advantage of electrolytics over foils in this application (if I'm keeping my types straight!) is that electrolytics are comparatively much smaller for a given capacitance.
 
Sep 7, 2008 at 9:48 PM Post #3 of 14
Cheers, ashmedai.
smily_headphones1.gif


So would the smaller foil capacitor bring any advantages to the table?

~Phewl.
 
Sep 7, 2008 at 10:18 PM Post #4 of 14
Foil caps are substantially larger if we're talking the 1uF/50V variety, and head towards soda-can sized if you want higher ratings. For 1uF/50V, we're talking roughly a 60% increase in diameter and an 80% increase in length, plus foil caps are often found in axial packages which are not convenient for direct substitution using an existing PCB.

Also, the foil cap I looked up to find that size ratio costs about $50. Each. If you buy 25 of them. Not that one sampling makes for great statistics, but I gather that electrolytics are also cheaper, and there's no point spending a ridiculous amount on one capacitor when you could spread the same price increase across the entire circuit and get a much better result that way. Foil is nice - if your project has the money and space for it.
 
Sep 7, 2008 at 10:26 PM Post #5 of 14
Electrolytics are not intended for local high frequency bypass, which is all this capacitor is for. You need to consider more than the bulk capacitance, although with electrolytics that varies +-20% generally anyway - the ESR, leakage current, ESL all contribute to the high frequency response here. Ideally, a 100nF film, something like a Wima MKS2, will do the job. Electrolytics have their place in packing large capacitance in small spaces for power supply smoothing, but local decoupling should always be film or ceramic.
 
Sep 7, 2008 at 10:30 PM Post #6 of 14
Hah, I need some coffee or something. I had him trying to substitute foil for electrolytic (which I guess comes up more often?).

General rule is electrolytics for power, and as Guzzler says, film/ceramic for filter caps.
 
Sep 7, 2008 at 10:33 PM Post #7 of 14
Yes - foil is relatively rare these days. They tend to be much smaller capacitances with tight tolerances. Electrolytics are much more common. Short answer here - yes to ceramic/film, no to electrolytics and foil! That's not to say an either wouldn't work, but they're not ideal.
 
Sep 7, 2008 at 10:36 PM Post #8 of 14
^^So the smaller foil cap will be doing something that a larger electrolytic couldn't?

Would a combination of the two therefore be best? The larger store of an electrolytic, along with the filtration of a smaller foil/poly capacitor?

~Phewl.
 
Sep 7, 2008 at 10:39 PM Post #9 of 14
Each in their place, really. The very things that make one type good for the place it belongs can make it unsuitable for other applications.

Electrolytics are decent at buffering power and let you get away with a larger capacitance in the same place, which can be more important. But they make cruddy filters.
 
Sep 7, 2008 at 10:40 PM Post #10 of 14
A combination of the two is the worst of any solution, despite what you may see in some designs - possible oscillation, funny gaps in the impedance. As I said - go with a good quality FILM (not foil) and be done with it. You can always experiment later on.

Incidentally, the MKP they recommend is actually a film...
 
Sep 7, 2008 at 10:55 PM Post #13 of 14
Cheers, I'll have a look over the Maplin website seeing as they opened a store here a little while back.
smily_headphones1.gif


~Phewl.
 
Sep 8, 2008 at 12:56 AM Post #14 of 14
AP....the reason I chose the 1uf K42Y-2 PIO to begin with was for it's outstanding sonic qualities first, but more importantly the 160V rating kept the size within reason and the cost made it a real bargain without equal. At around 1.25 US per cap....anyone can afford it.

I had spares sitting around from another project so when I read about the cap tweak it was an easy choice to grab one of those caps and try it out. I'm glad I did.

Great thread BTW.

Peete.
 

Users who are viewing this thread

Back
Top