Question about choke in the Neumann NG psu for U47 (schematic included)

[mad.ax]

Hello, and happy new year!

In the original NG psu, there's a dual choke of 2 X 8 Hy.
I'm wondering if it's completely equal to 2 single 8Hy chokes?
Or could there be some kind of "humbucking" arrangment in the dual choke which would further reduce the ripple?

Axel

 

[gyraf]

8H + 8H on the same core will be 32H across, as same-core-addition-of-turns raises inductance with turns-squared..

Jakob E.

 

[mad.ax]

Thanks for your answer Jakob, but I'm affraid it raises more questions!

Does it means that if measured individualy, each coil will still show 8Hy? But if both were measured in serie, , it would read 32?

Am I right assuming that if one wanted to achieve the same filtering while using two separate chokes, then on would have to use two 16Hy chokes?

So winding a twin coil choke would only be a trick to spare a second core?

Axel

 

[mad.ax]

For those interested in seeing what an original Neumann NG psu looks like, here is a link with lots of pics:
http://www.jogis-roehrenbude.de/Roehren-Geschichtliches/Neumann-Mikro/Netzteil_NG/NG.htm

Axel

 

[salomonander]

May i bump this dinosaur? Im building a psu for a u48 with zayances pcb and im wondering about the choke. On the bom its a 10H - but that seems to not be what neumann used. I saw some diy schematics with 2x15H as well. Does it not matter that much? Im not good with this kind of stuff....

I could fit two 15H i there. But is it beneficial? And what about magnetic interference?

Hammond makes a single 30H choke - but its only 40mA.

Thanks!

 

[gyraf]

40mA is what the u47 is rated at

I don't think the actual value of the choke is all that important - it just forms a very steep high-cut filter on the HT voltage to keep crap out of it. Actual cutoff frequency is not important, if you think of it..

Only, you will want to keep resulting possible resonance frequencies out of the audio band

Detailed walkthrough e.g. here: https://cds.cern.ch/record/2038629/files/265-289-Kunzi.pdf - caution, this gets hairy pretty quickly..


/Jakob E.

 

[salomonander]

Thanks Jakob,
I just looked at the paper - but im afraid its a bit over my head....
But telling from your post, any value between 10-32 would work?
And is a 40mA choke safe in this case, or would i want some headroom? Thanks and sorry for not being smarter

Attached a pic of a choke on ebay. 2x6.5H but im not sure about the power rating....

 

[RuudNL]

The later U47 power supplies even didn't have a choke.
(I just replaced the capacitors in an original one.)

 

[gyraf]

Yes, there are no chokes in the supply for my u47 iirc

/Jakob E.

 

[salomonander]

Sorry for being persistent but i assume that 2x6 choke should do? Or anything above 10 Henry?

 

[gyraf]

yes, agree..

 

[abbey road d enfer]

The thing that perplexes me is that those two inductors with heavy coupling and a cap in the middle create an all-pass filter, which has almost no attenuation, but changes the phase of the ripple.
Actually, removing the middle cap results in better filtering.
Indeed the value of the inductor is far from critical.

 

[salomonander]

Thanks Jakob!

 

[CJ]

looks like they were trying to save space, those selenium rectifiers hog a lot of cubic inches,

never seen that tapped choke before,  very interesting!

you could probably get away with one 8 Henry choke, 

 

[salomonander]

Thanks CJ. No 8H available. I can go 10H, 15H, 30H or 2x6H....
what would you choose? Cheers!

 

[CJ]

15 H would be cool,

120 Hz ripple would encounter 2 pi x 15 Henry Choke x 120 Hz Ripple Frequency = 11,300 ohms of resistance to the deadly 120 Hz.

 

[salomonander]

Thanks CJ - thats what i ordered today. Best