Question on capacitor arrangement for aleph 2 amp

[BrianGT]

I am building a set of Aleph 2 amplifiers, and I have two possible setups for capacitors, and I am not sure what would work better. This is the setup for each amp, and I would have the same on both amplifier.

setup 1: 4 x 30,000 uF 63v capacitors (macintosh ones from brigarelectonics.com $6 each)

setup 2: 20 x 5,000 uF 50v capacitors (mallory, I got them from a local electronics surplus in Atlanta, GA for $0.75 each. I bought one box, which contains 42 of them, and they have about 8 more boxes there.) Would more then 100,000 uF of capacitance sound better? I can easily buy more of these capacitors and make a bigger array.

What is the better setup? Is there anything special to do when hooking up the capacitors, besides hooking them up all in parallel? I would probably make a thick copper strip to run down the capacitor bank.

Thanks for any input,

--
Brian
gte619j@prism.gatech.edu

 

[grataku]

There is no evidence that 0.1 F would 'sound better' than 0.05 F. Beyond 50000uF you will find that turn-on short circuit starts to be a serious problem . You can surely test that by adding progressively more caps until you blow your main AC breaker.
For these types of constant current draw amps a pi filter works alot better than an exorbitant amount of caps.
Anyways, no matter what you do, if you haven't done it already, make sure to stop by the Leach lab (gatech, right?) and test out the ESR of those surplus caps.

 

[BrianGT]

Sounds good.

The caps are actually brand new, in a sealed mallory box of 42. They are the standard CG502U050U3C ones that are 5000mfd 50vdc

I will run by prof. leach's lab next week and check them out.

--
Brian

 

[BrianGT]

How would you setup a pi filter with more then two capacitors? Would you divide the number by two and stick a inductor in there?

--
Brian

 

[Nelson Pass]

Yes, but you can also cascade such filters for even
wilder filtering factors.

 

[BrianGT]

Do you know of any online reference for this from people who have done it before? I tried to search for pi filter on the forum, but pi is less then 3 characters long and cannot be searched for, and I could not find any threads relative to this.

I guess you would make a cascaded array of capacitors and inductors, but I am not sure of what values to use, and I would like to know if anyone has any existing designs.

In general, is 20 smaller caps better then 4 larger ones?

Thanks,

--
Brian

 

[Dustin Haug]

Nelson or whomever may know, with regards to cascading PI filters. Could you use several smaller inductors, arranged like this C-L-C-L-C. Where C is say a 30,000 uF cap and L is a 500uH to 1mH inductor. Reason for using the smaller inductors being the expensive of larger (2+mH) high current inductors.

 

[GRollins]

Inductors in series are like resistors--the value is additive. A C-L-C-L... chain will give you excellent results.
I disagree about large cap banks being audible. They are--primarily in the bass. However, it is a diminishing returns curve and you'll need to keep an eye on inrush current.
After having been let loose on a large pile of computer gear that was going to get junked, I ended up with a decent number of electrolytics. I promptly went home and wired a mess of them to the Threshold S-500 that was running my subs.
Whoa!
But the S-500 had (I think, I haven't been in one recently) about 120,000 uF already. I added roughly 150,000 uF. The difference was there, unquestionably, but I wouldn't have put up enough money to actually buy the caps, as it's not that noticible above, say, 100 Hz. Simply not cost effective, given the price of big computer grade cans. However, if you get the caps for free...
Note that I used a soft start circuit, and that the amp took <i>quite</i> some time to turn off after flipping the switch.

Grey

 

[macka]

Question on capacitor arrangement for Aleph 2

Hi Brian,


Leading on the recents posts I suggest you have a look at Duncan amp page, there is an easy to use download PS simulator and you can design your own supply from this.

For example with the Alpeh 2,

Using a classic Pi network of T + BR + C +I + C+( 15R= load)

Where T = 36 volt, C = 25000 uf, I= 2mh @ 0.10 ohms

The ripple will be only 44 millivolts! and V out = 44.79 volts

If we remove the inductor and add the capacitors to make 50,000 uf the ripple increases to 496 millivolts which is 10 fold higher!

It depends on whether the 496 millivolts is audible and the value of capacitors you have available, and cost.

The inductors must be high current with low dcr (ie 0.2 ohm or less,) and suggest you look at Madisound, Sonic craft or Zalytron who have nice high current chokes for audio.

I recall Mr Pass has used these chokes with success in a previous thread.

I hope this helps.

regards

macka

 

[BrianGT]

Thanks, I will look into this. I appreciate your response. It does seem like a great idea to use chokes. I am wondering if it would be better to use a cascaded arrangement of smaller ones with an array of capacitors.

--
Brian
gte619j@prism.gatech.edu

 

[FEThead]

source for chokes

Here is a 10 amp 10 mH choke for $25 (plus shipping):

http://www.surplussales.com/transformers/powerchklist.html


scroll down to cpw-500006001201

I have no connection with the company but I have bought from them before and they have good stuff.

Happy listening!

 

[hifi]

what about this rigg:


1,1mH 1,1mH 1,1mH
Transform - - - L1 - - - L2 - - - L3 - - - Aleph
I I I I
68µF 31mF 31mF 31mF
10x I I I
film caps V V V
I
V

The coils are 10A 0,1R types the caps are 10pcs of 6,8µF 160V polyester filmtypes the other three stages are RIFA PEH169 105C 31mF 50v types (!!!) The transformer is a 900VA specialy made for audio (somewhat different core material i believe)

should i do something different i got all the caps dirt cheap att surplus brand new made in last november the rifas have a shelf life of 10years !!

/micke