|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
204.185.81.23
In Reply to: Man, I gots my lowmu approved PS parts, now must rectify posted by bobbyj on March 10, 2007 at 10:27:13:
Hi Bobbij,Lowmu here. I had the same problem of no center tap on Greg's 245 amp's Signal DU 1/2 PT. We also eliminated his existing TV damper diode bridge and employed a Graetz ( sp. ) bridge, two 1000 PIV diodes to get the negative, or ground, and simply let the 5U4GB do the B+. That works well enough, sounded fine, used two uF 4007s. To make use of the 6 VAC winding, you could add some power Rs of equal but low ohmic value, into each leg of the winding, to drop it down to 5 VAC under load. Loaf the dissipation - rating on the Rs please.
Now a days, I am not even using the C-47U anymore, as I have gone to a Zen - simpler two-stage filter system with a pair of Triad C-40X Ls ( 10 ohms DCR, 0.32 Hy, 600 mA., $10.79 at Allied ) for L1 & L2, and some 40 uF ASC Percy oils for C1 & C2.
If its only a low draw stage you are powering, consider use of a bleeder R at the output of the filter, B+ to ground, using Mills MRA-12s.
Just see my most recent posts from this week on this page, below. Let us KNOW what you hear, measure and fully document your key voltages before and after - and get them right please !!
Follow Ups:
It seems there is a refinement to the theory of Low L, Low C, Low DCR. Some parts that possess these qualities reportedly do not work.Look at this post: http://www.audioasylum.com/forums/tubediy/messages/118417.html
Unless I am just being gullible, I think it is saying 10 ohms DCR is better than 2 ohms, because the 10 ohm choke has more henries or, if not, a 2 ohm choke with .5 henries has too much wire and iron in it, or so I gather.
Now, I was poking around in some of the old low DCR posts from Jeff and found a statement that Dennis's proprietary amp design, which presumably is ultra-fi, uses some chokes that are only 1 or 2 ohms DCR. Mileage varies, I guess.
Yesterday, I was all set to build a low DCR power supply (L1/C1/L2/C2 per the current recipe) out of good parts I have on hand. A pretty high current transformer (260-0-260, unknown milliamps but it is big, 45 ohms DCR but DCR is of secondary (pun) importance), a 5U4GB rectum-fryer, 40uf poly in oil caps, and a pair of .3H 3 ohm 1 amp Stancor 2690 chokes. Now I am afraid that my chokes are too low DCR or too big, so I would need to buy something different. Oh well, before a few days ago I thought the power transformer was not even mid-fi, so easy come easy go.
Stancor 2690 chokes: http://www.alliedelec.com/Search/ProductDetail.asp?SKU=928-0075&R=928-0075&sid=45F346803041E17F
(I only paid $10 for the chokes, used. Bought them for a DC filament supply, but I now like AC on my filaments).
So, are these Stancor chokes in the category of the C47-Us, as no good for the Low L, Low C, Low DCR power supply?
BTW, when I model the power supply I end up showing 800mV of ripple. Is this a failing of PSUD II, with real world results coming out much lower in the ripple department?
http://www.alliedelec.com/Images/Products/Datasheets/BM/STANCOR/Stancor_Actives-and-Passives_9280070.pdfAbove link is to the Stancor data sheet to all their chokes. In the Allied catalog I got real excited because they list the C-2686 as 250mH with a DCR of 0.43 ohms. Well, come to find out, it is really 25mH. Too bad it was not the other way around; I'd like to have seen a 250mH choke with a DCR of 4.3 ohms. Darn the bad luck. :0(
Why don't you build the supply and listen to it on an experimental basis. For over a year I have made reference to a Triad C-40X, and I would sub in the Triads after you get some experience. The UTC vintage chokes were heavy ( lotsa core and too much wire ) and that is why the more diminuative Triad ATE the UTC L's.Do not worry about ripple,if you are using it on an output stage, this supply is about DYNAMIC capabilities, and it far overshadows any concept of ripple, which is more static of a parameter in my way of thinking.
I know for a fact that the main power supply Ls in Dennis' Serious Stereo 2A3 are NOT as low in DCR as what one reader has posted. Its more like the Triad C-40X, ten ohms. That will do swell. There is a BALANCE in the audio design that is necessary, and whereas ole Lowmu used to answer everything with " lowest DCR ", I have come to find out that adjustments have to be made, to incorporate Low C and Low HY ALSO. The net result of this triad of power supply design factors far exceeds low DCR alone, after HEARING this in Greg's amp last month.
Heavy ( and good ) wiring plays a big role in this supply giving you the most possible.
Do not be discouraged, dig into it and give it a listen. sub in C-40Xs later on. Report back your listening experience please. Thanks. Don't be overly concerned with PSUD. Cheers !!
Build it and see sounds fine. Time is difficult to find to get anything to closure, but we all have that problem.By the way, I also have some relatively small JVC "reactors" that seem to be .3 or so H and 9 ohms. Not as big as the C-40X. I could try them too.
Finally, I based my comment on Dennis's amp having sub-2 ohm DCR chokes on a post under your name but in which I thought either you or Dennis was speaking, saying:
"FYI MikeyB, ....... ALL the Serious Stereo 2A3 amp's inductors are what Jeff loves to call " ultra hi fi " i.e.:
Ten ohms or less DCR,................ and several are WELL under one ohm!!"
The post was at:http://db.audioasylum.com/cgi/m.mpl?forum=tubediy&n=110681&highlight=
I did not understand that that was somebody else just speculating.
Yes, didn't mean to hang you up !!!The Ls under one ohm are just some RF chokes Dennis uses here and there, but the main supply chokes are just what you were thinking..... ten ohms or less, very close to 10 !!! Listen to those 0.3 HY "reactors" too.
I like the " triple paralleled clip lead " method to test things quickly and substitute parts or configurations. ONE clip lead, in this type of supply, doesn't cut it at all - loses it in the wire !! You will LOVE how this sounds is my guess.
Recall, our "final" configuration at Greg's was JUST the two stage ( better step response ?? ) filter, and with 40 uF C's, we had .48 VAC on a 270 VDC supply to the two 245'. On a separate / similarly ironed supply, except we used 50 uF C's, we had 20 mVAC to the two 76s plus an extra shunt draw, about 26 mA total draw, on a 170 VDC supply. Ripple was NOT a factor at Greg's house in our listening experience ( RCA field coils and Altec 515 ) . Sounded delicious, it put a smile on my face.
Depends what you want to use it for... If you use low-L, low_C, you may need more stages to get enough filtering. "Good" parts for one circuit may be marginal or useless for another.I have found PSUD quite accurate, so perhaps your simulation isn't matching your actual components or load current.
Tom:I have generally found PSUD accurate too. So, I was asking how it has come to be accepted by some that .3 H / 40UF plus .3 H / 40UF is enough filtering of a power supply, if (as I believe to be the case) the ripple will be between a half and one volt. Drlowmu says build it and listen, it may not be the problem I think.
The amp itself will be a test bed for low power DHTs or DHPs, 45, 46 or 47 or similar. Mainly, I want to test some ideas, starting with these power supply ideas.
Steve Root
0.5V is too high for a single-ended triode - about 20 mV p-p to an 8 ohm speaker - but it might be OK for a pentode or push pull. Try it and see - but be sure you have room for more filtering if necessary.
Hi Tom,Well, thats only your opinion and not mine, now a days Tom, on an output stage.
You have not experienced this type of supply, have you ? You should !! We use 105 dB RCA FC's and about 98 dB Altec 515s at Greg's house, and have no problem.
My favorite filter ( so far ), has been two-stage, not three, because of the blinding instantaneous speed and timing the two stage gives. Filter it some more, put in five stages if you want, and you will wonder where the music went as far as both timing and dynamics !! A higher number of filter sections will be, by comparison, all askew and out of kilter and out of time. What happened to the music ???
Front end supply, with the same basic two-stage filter, and less draw, is 20 mVAC.
Funny we should have this discussion Tom, over the past, for 25 years, I would actively tube regulate, TWO regulators in series with a draw down power R between them !!, to get low ripple, usually 1 mVAC in my triode amps.
I really like this new-to-me Low C, HY and DCR unregulated approach, as no real comparison to anything else in audio that I know of exists today - IMHO.
I'm workin on mine! I need to build a separate front end supply though.
I'm putting together the main B+ now, for the output stage (2A3). I thought I would need "less" ripple on the driver? Can I use a similar two stage low everything supply and not have noise on my front end?
I was going to use a low mu triode: 6P5, 6J5 or better (IMHO) triode wired C3o which has a Rp of 2K while mu of 17, (the 6J5 we should have had). I have some 1660 IT's, also plate chokes, assuming only these two coupling methods would work with this PS? I guess loading the driver with a resistor "kills" the point of the PS? Jeff, please tell more about your driver and the problem solving you went through thanks.
Hi bobbij,Read my prior post, only 20 mVAC to the 76 stage, with a two stage filter, so YES, do it the same for your front end. Use a Mills MRA-12 ( one or more ) shunt bleed across the B+. No problem. Use a real triode !!
No, loading the plate with a good resistor is fine, and is wider band than any choke or IT.
Thanks Jeff, OK, I thought the BW was "wider" with a resistor, but wasn't sure the dcr was contrary to the rest of the effort. I too, like the "idea" of "pure" triodes. I like 6p5, 6j5, etc...But the C3o is a surprising devise, a little on the bright side, perhaps due to the low Rp. Now, to load a 6P5 "properly" I'm assuming a 50K plate resistor, so I'll have a 250V drop at 5ma idle current, so I'd need at least 450-500 VDC to start with?
Hi,
Please kindly re-read Dennis' comments in my post of his most recent two emails, where he clarifies and refines this subject, second email, under point ( H ).Design with low DCR except for the audio circuit, there, High DCR to allow it to swing !!
Got it ??!! URL follows.
Jeff Medwin
This post is made possible by the generous support of people like you and our sponsors: