Reply
 
Thread Tools Display Modes
  #1   Report Post  
Posted to rec.audio.tubes
Don[_4_] Don[_4_] is offline
external usenet poster
 
Posts: 20
Default parallel chokes in a power supply

I need opinions on the use of two chokes in parallel in an otherwise
standard Pi type filter. c - L//L - c
Assuming nearly identical chokes, ie, same part number and
manufacturer, will there be any problems?
I know the combined parallel inductance and resistance will be one half
that of a single choke, and the current capacity will be doubled.
Will one choke "take over" and burn out?

I have "space in the chassis problems", so I can't use one big choke.
-Don

  #2   Report Post  
Posted to rec.audio.tubes
Patrick Turner Patrick Turner is offline
external usenet poster
 
Posts: 3,964
Default parallel chokes in a power supply

On Mar 18, 2:17*am, Don wrote:
I need opinions on the use of two chokes in parallel in an otherwise
standard Pi type filter. *c - L//L - c
Assuming nearly identical chokes, ie, same part number and
manufacturer, will there be any problems?
I know the combined parallel inductance and resistance will be one half
that of a single choke, and the current capacity will be doubled.
Will one choke "take over" and burn out?

I have "space in the chassis problems", so I can't use one big choke.
-Don


Two parallel chokes in a CLC supply should have nearly equal winding
resistance so the heat generated in each is the same.

If they have equal inductance, the total L for 2 becomes half the
value.
Total L= 1 / ( 1/L1 + 1/L2 )

Would you still have enough L if you parallel 2 chokes?

The Vac across the chokes causes negligible winding heat.

Patrick Turner.
  #3   Report Post  
Posted to rec.audio.tubes
[email protected] arthrnyork@webtv.net is offline
external usenet poster
 
Posts: 81
Default parallel chokes in a power supply

On Mar 17, 11:17*am, Don wrote:
I need opinions on the use of two chokes in parallel in an otherwise
standard Pi type filter. *c - L//L - c
Assuming nearly identical chokes, ie, same part number and
manufacturer, will there be any problems?
I know the combined parallel inductance and resistance will be one half
that of a single choke, and the current capacity will be doubled.
Will one choke "take over" and burn out?

I have "space in the chassis problems", so I can't use one big choke.
-Don


Analogue vacuum tube projects are very tolerant (in general) to small
difference such as your project . I wouldn't worry about it . Unless
it will power a precision sensitive circuit , such as a very stable
RF source or a similar application .
  #4   Report Post  
Posted to rec.audio.tubes
Patrick Turner Patrick Turner is offline
external usenet poster
 
Posts: 3,964
Default parallel chokes in a power supply

On Mar 18, 8:23*pm, wrote:
On Mar 17, 11:17*am, Don wrote:

I need opinions on the use of two chokes in parallel in an otherwise
standard Pi type filter. *c - L//L - c
Assuming nearly identical chokes, ie, same part number and
manufacturer, will there be any problems?
I know the combined parallel inductance and resistance will be one half
that of a single choke, and the current capacity will be doubled.
Will one choke "take over" and burn out?


I have "space in the chassis problems", so I can't use one big choke.
-Don


Analogue vacuum tube projects are very tolerant (in general) to small
difference such as your project . I wouldn't worry about it . Unless
it will power a precision sensitive circuit , such as a very *stable
RF source or a similar application .


With a CLC filter the LC section after the first C has a low frequency
resonance
which is calculated at Fo = 5,035 / sq.rt ( L x C )
where Fo in Hz, C in uF, and L in mH.

This becomes critical in an SE amp audi amp which the OP may indeed
have.

Fo ideally should be less than 3Hz, but sometimes the required C2
after the L becomes too large a value in relation to C1 and L in order
to get Fo low enough.

Ideally, C1 should have reactance ( ZC1 ) at the rectifier ripple
frequency of 0.1 x RL where RL is the output Vdc / Idc amplifier
load. ZL should be 10 x ZC1 and C2 should be 0.1 x ZL.

If there is a tube rectifier then C1 can only be a rather low C value,
say 47uF.

Suppose F Ripple = 100Hz, then XC = 34 ohms.
Then XL = 340 ohms minimum so L = 0.5H.
Then ZC2 = 34 ohms so C2 = 47uF.
Ripple reduction factor = ZC2/ ZL = 34 / 340 = 0.1, and Fo between L
and C2 = 33 Hz
and the Q of the resonance curve is high. Usually the load of the
amplifier is an R value way above what is required to damp the
resonance. Suppose Vdc = 400V and Idc = 250mAdc, then RL = 1k6.
At Fo = 33Hz, ZC2 = ZL = 103 ohms.

Ripple reduction factor of only 0.1 is a poor quality outcome.

To damp the Fo to stop the PSU bouncing up and down at 33 Hz the RL
should be 102 ohms.
But one could have a a series R of 102 ohms between L and C which
would damp Fo OK but then there is an unwanted 25Vdc drop across the
102 ohms after the choke before C2.

But in practice with a tube rectifier one wants ZL = 100 x ZC1, or
3,400 ohms in the example case. This means L = 5H, a rather large
heavy item but its the right price for getting real Hi-Fi, OK. Then C2
still should be 0.1 x ZL which would give ZC2 = 340 ohms so C2
could be 4.7uF but this has a value of reactance at LF which fails to
anchor the OPT take off at the wanted LF pole of the amp.
ZC should be 0.01 x ZL so C2 = 47uF minimum. Then Fo L&C2 = 10.38Hz,
and this is better than 33Hz but the outcome is still low quality and
what is wanted is C2 = 470uF which gives Fo = 3.3Hz which is OK.

The peak charge currents of the tube rectifier are mainly determined
by C1 value and don't much increase because of following 5H and large
value C2.

At Fo = 3.3Hz, ZL = ZC = 102 ohms. To damp the Fo, 102 ohms is needed
for RL or in series with
C2 after L. A 5H choke might have winding R = 50 ohms which gives
3.125W heat at 250mAdc.
The Rw is effectively in series with L value, so additional R could be
50 ohms for a flat response.
Vdc drop across choke = 12.5Vdc and 12.5Vdc across the additional R.
In practice the resonance peaking at 3.3Hz is not going to be a
nuisance if the R is 50 ohms, mainly because 3.3 Hz is so far below
the wanted AF band. So in fact additional R isn't needed. But its Best
Practice to add the R.
And then we would make things almost perfect by adding a second 470uF
450v rated C to make the filter
C1 L C2 R C3, with the L = 5H and R = 50 ohms and make C2 = C3 =
470uF.

Ripple voltage becomes extremely low, but that's what is needed for an
SE amp. Its less important for a PP amp although the CLCRC filter is
unsurpassable for where a well smoothed screen supply has more
importance than a well smoothed B+ for the CT or OPT.

I don't share your your slap-dash approach to audio amp design. For
optimal outcomes we must worry about everything and spend a bit more
on L and C.

To avoid resonance effects at LF the alternative option is to dismiss
the tube rectifier with its very poor peak current handling ability
and use S1 diodes to feed whatever high C value we can afford.
Peak charge current is then limited by winding resistance or HT and HT
windings plus wall outlet resistance but to limit the high peak charge
currents with say C1 = 470uF, a series R between HT winding and bridge
should be used, say 10 x ZC at ripple F, so say 33 ohms. Instead of
any L, we then may use 2 sections of RC filters using 33 ohms and
470uF which gives a ripple reduction factor = 0.01, and because Vr is
so low at C1 the Vr at C3 is extremely low. The Vdrop across the total
100 ohms of added R is roughly about 25Vdc so the outcome is about as
good as the CLC with 47uF, 5H, 470uF.

But Vdc may be too high because the winding was designed for a tube
rectifier where the Vdc might be 1.1 x Vrms of HT winding. With an Si
rectifier we get Vdc = 1.35 Vrms, ( loaded ).
OK, then the three R values are made larger to get the Vdc down to
where we want it. There is increased heat production in the increased
R which would otherwise be liberated as heat in the rectifier tube.
The HT winding should have a number of taps to allow adjustments of B
+. Factory made power trannies seldom have enough taps, and expect the
user to work out carefully what he wants then buy accordingly.

Many DIYers just don't make it through the design maze without ending
up making a sub-optimal amplifier.

There is always some damn price to be paid unless design is done well,
and unless designers regard the exercise as critical.

There is much design given in pages of RDH4 or at at my website pages
listed at
http://www.turneraudio.com.au/education+diy.htm

Patrick Turner.

  #5   Report Post  
Posted to rec.audio.tubes
mike s mike s is offline
external usenet poster
 
Posts: 27
Default parallel chokes in a power supply


My gut feelings on this are -

Parallel L is not unlike series C and will likely have some of the
same problems. Yes inductance and resistance will be halved but peak
current may not, just as peak V in series C isn't, because real
components aren't perfect.

Series L would be better, two small chokes that can carry the current
but with half the inductance each.

How come two chokes are smaller than 1? It seems unlikely.



  #6   Report Post  
Posted to rec.audio.tubes
Big Bad Bob Big Bad Bob is offline
external usenet poster
 
Posts: 366
Default parallel chokes in a power supply

On 03/21/11 10:29, mike s so witilly quipped:

How come two chokes are smaller than 1? It seems unlikely.


the 2 chokes might be physically smaller due to lower current and
therefore thinner wire. It probably has something to do with the inside
diameter and wire thickness of the physically larger choke. I might
also guess that the larger choke is rated WAY more than twice the
current of the smaller ones. Or, it could have a potting can around it
making the whole thing 'that much bigger'.

Maybe it would work better to get the right choke for the job?

/me is a fan of LC or LCLC to take full advantage of the input
inductor's behavior of inducing voltage to maintain current flow.

  #7   Report Post  
Posted to rec.audio.tubes
Andre Jute[_2_] Andre Jute[_2_] is offline
external usenet poster
 
Posts: 631
Default parallel chokes in a power supply

On Mar 17, 3:17*pm, Don wrote:
I need opinions on the use of two chokes in parallel in an otherwise
standard Pi type filter. *c - L//L - c
Assuming nearly identical chokes, ie, same part number and
manufacturer, will there be any problems?
I know the combined parallel inductance and resistance will be one half
that of a single choke, and the current capacity will be doubled.
Will one choke "take over" and burn out?

I have "space in the chassis problems", so I can't use one big choke.
-Don


Some other matters to consider with various choke configurations:

http://www.audio-talk.co.uk/fiultra/...dre%20Jute.htm

Andre Jute
Visit Jute on Amps at
http://www.audio-talk.co.uk/fiultra/
"wonderfully well written and reasoned information for the tube audio
constructor"
John Broskie TubeCAD & GlassWare
"an unbelievably comprehensive web site containing vital gems of
wisdom"
Stuart Perry Hi-Fi News & Record Review
  #8   Report Post  
Posted to rec.audio.tubes
Don[_4_] Don[_4_] is offline
external usenet poster
 
Posts: 20
Default parallel chokes in a power supply

Thanks for all the help, everyone. I decided to use one big choke on the
top of the chassis, instead of two smaller chokes hidden away below deck.

Reply
Thread Tools
Display Modes

Posting Rules

Smilies are On
[IMG] code is On
HTML code is Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
WTB: PS AUDIO M-250 power supply, HCPS power supply TUBELOVER Marketplace 0 April 2nd 04 07:33 AM
WTB: PS AUDIO M-250 power supply, HCPS power supply TUBELOVER General 0 April 2nd 04 06:45 AM
Need Phantom Power Supply That is Not a Switching Supply Carlos42 Pro Audio 10 March 17th 04 06:24 PM
FS (Australia) Tube amp power transformers + chokes John Mackesy Vacuum Tubes 0 August 22nd 03 08:49 AM
FS (Australia) Tube amp power transformers + chokes John Mackesy Marketplace 0 August 22nd 03 03:58 AM


All times are GMT +1. The time now is 04:27 PM.

Powered by: vBulletin
Copyright ©2000 - 2024, Jelsoft Enterprises Ltd.
Copyright ©2004-2024 AudioBanter.com.
The comments are property of their posters.
 

About Us

"It's about Audio and hi-fi"