F5 Listening Impressions & Discussion

[bear]

Imho, 10ga is serious overkill.
I think 14ga is quite sufficient.
12ga would be nice.
10 ga is heavy duty if you need to minimize DCR for some reason...

All you need is wire and a way to make it go round and round on a bobbin or spool, etc...

There is no reason not to use an "iron core" to reduce the number of turns and maximize the inductance, imo. You just have to size the lams to meet the VA required for the job.

It is reasonable to use the secondary or primary of a suitable transformer for this...

_-_-bear

[woodturner-fran]

Well, I have a couple of lathes that can run at slow speed - ideal for winding I would say. However, its laying hands on the copper for almost no $$ is the problem!!

I do have a big microwave oven transformer though that might just do the trick.... further investigation required

Fran

[Zen Mod]

just search somewhere ( Babelfish J thread ? ) for my choke recipe

[bear]

Scramble wound doesn't look pretty, and isn't quite as many mh per turns as "perfect lay" but works fine and you can do it on a lathe with hand held feed. Watch your fingers, wear tight fitting clothes. The lathe has no brains and no mercy.

_-_-

[Bigun]

Originally Posted by bobodioulasso

Though i have not yet experimented it, i am quite shure that a capacitance multiplier could be a more elegant and a lot cheaper way to improve noise figure in a classA amplifier PSU.

That's what I think, it promises huge noise reductions and channel-to-channel isolation. But with FETs you get the 4V drop and with BJTs you need a driver device to handle the base current of the pass transistor. And you need a heatsink with mounting hardware, and the devices can fail. You do get the bonus option of zener regulation to the base, or a triac on the base to shut down the rails if there's a fault condition detected.

But the CLC offers a rather simple solution and need not be expensive if you accept the use of a common mode choke. Check out Digikey. I found a 4A choke with 6.9mH on tap for around $5 [PLK1332-ND].

[jackies]

That's a cool choke, and inexpensive.
However, it's supposed to be used in the AC line.

Maximum current is specified as 4A RMS.
Any ideas on how it will behave in a typical F5 CLC (CLCLC) PSU?

[Bigun]

I figure the choke doesn't know which side of the trafo you put it on, but I assume the current limit is mostly a thermal issue - if you don't overheat it then it doesn't matter what the waveform going through it looks like.


I haven't a clue how it would work for the F5 but there are many other nice looking chokes in the digikey catalogue if this one isn't suitable.



I'm still liking the look of the Cap Multiplier as a technically better option but with the choke there appears to be less to understand and so fewer possible complications. For some amplifiers, the turn-on thump is a deal-killer without a cap multiplier to slow things down.

[fishski13]

Originally Posted by a.wayne

It's been 4 weeks ......................

. my 2nd child should be popping out in the next couple of weeks. we made a decision to get rid of my dedicated hi-fi room soon than later, and i've been kicked down into the office in the man-cave. i'm in the middle of building a USB DAC - space is at a premium now and i've decided to pack away the CDs for a hard drive. although i did finish populating my F5 boards tonight.

[sangram]

@danny: thanks Maybe in the next build.

[danny_66]

I'm in the middle of assembling a F5 with a CLC PS.
I bought my chokes from Automatic Electric Europe
They are rated 10mH, 5A DC, 0.0695 Ohm, size: 80x70x60mm
The price was 26€ for a choke.
The nice thing about them is the low DC resistance, it's almost impossible to make an air choke with the same specs or it will be enormous.

DSCF2152s.jpg