|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
125.209.166.211
I have a power amp with power transformer secondary windings of 375-0-375 VAC and B+ = 500 VDC, with 2 filter caps in series: Nippon (United) Chemi-con REW 450VDC, 1200 uf . They have been in service for 20 years now and perhaps time to consider replacement. So far I have only found Nichicon NT with similar specs (450 VDC, 1000 uF). Hence 2 questions:
1. Anyone with experience comparing the Nippon Chemi-con REW with the Nichicon NT?
2. Suggestions for other brands?
Follow Ups:
nt
Observe, before you think. Think before you open your yap. Act on the basis of experience.
Thanks for sharing this link. About time someone produced a replacement for the famed BGs. You can tell I like them. Here's a picture of my custom amp that has 160,000 uF all Black Gate power supply!
The high voltage Cerefines were outstanding too.
Retsel
Notwithstanding other more exotic solutions, the Nichicon high voltage capacitors are excellent. So too are Panasonic TS-ED and other high voltage Panasonic electrolytics. I know it's a bit heretical, but I do not find that "any" film cap sounds superior to "any" electrolytic. Some metallized film caps impart a coloration that I find objectionable. I don't know why, and it's just my opinion.
However, if you can stay below 100uF, high voltage "motor run" caps are excellent film types for a PS. If you can stay below 200V rating, ELNA Silmic II are my favorites.
I agree with you on both parts Lew.
I've got some ASC caps with quite a "glare". I've heard it with other brands too. I've never heard a huge improvement in sonics when changing from a good quality electrolytic to a film cap of same value. I wish I could say otherwise. I use poly caps anyhow and still avoid electrolytics.
Is the amp a Cary ??If so, and IF I got the name correct ( "if" is a BIG word ) you may want to speak to Mike Samra, who greatly reduced the amounts of uF in the filter, and got significantly better sonic results as I recall from his post.
Generally speaking "today" many would use 200 uF total, or less, in such a B+ filter !!!
Additionally, we have very reasonably priced and small sized " new technology " film caps available today, that would be a BIG step up sonically, from any electrolytic. These high voltage films are called DC LINK caps, made by WIMA, Vishay, Panasonic, etc. Ultra high in audio performance.
Mikey, you reading this post ??
Jeff Medwin
Edits: 12/17/14 12/17/14
"Mike Samra, who greatly reduced the amounts of uF in the filter, and got significantly better sonic results"
Maybe, but here's the thing. If there's no measurable audio voltage across the caps in normal operation, they aren't influencing the sound. If that's the case, you don't need to change values or cap types. Neither will have any sonic effect.
--------------------------
Buy Chinese. Bury freedom.
You are right on the money. CARY!!
I've been thinking about reducing the uF but not sure by how much. These caps are expensive to experiment with. Besides it also depends on the rectifier design - my Cary is full wave center-tapped. Based on 375-0-375 VAC transformer secondary and a B+ of 495 VDC, that's a 6.6% ripple. Reducing the filter capacitance from 600 to 400 uF would increase the ripple to 10%. Good choice or not?
I've seen people reducing the capacitance by 30% in a Cary's 300B amp and claimed a much more musical (relaxed) sound. But that certainly is a personal preference. I like that kind of sound too. Besides, physical dimensions (limited to 50-51 mm cans) is another constraint.
I'd be happy to PSUD2 the supply, if you want me to. Have the schematic's URL ?
That ripple you quote is "off" in my opinion, without checking, simulating. Waht did Mike Samra post about HIS cary on the AA ?? !!!
jeff medwin
The ripple estimate is based on the following measured values:
Transformer secondary: 375-0-375 VAC; B+=495 VDC
Ripple = 1 - 495/(375 *1.414)=6.6%
Are you trying to say there is 32.67 VAC on the 495 VDC B+ line ??
.066 times 495 VDC ???
If so...no way !!!!! Put a multimeter on that spot and measure VAC !!
In any case, you do not need or want 600 uF effective there, way too large.
LOOK at the new technology WIMA DC LINK caps in the 800 to 1000 VDC range, maybe 100 uF or 150 uF !!!
Jeff Medwin
Perhaps my non-engineer perception of ripple is causing mis-communication. The "ripple" I mentioned was actually the voltage drop from peak to that at start of next charge cycle.
I have looked up datasheets of DC Link caps. They are too large to be used as drop-in replacement, being 85 mm in diameter & the existing caps are of 51 mm.
I agree that 600 uF maybe too large, but what would be the appropriate replacement value?
If you could tell me what the current draw is on the supply, ( current of tubes in each stage ) I could PSUD2 model it, and we could jointly figure out what uF values will be good to chose from. Maybe leave it up to others on Forum to comment, with YOU having the final choice.My rough "guess" would be to try 100 to 200 uF, but I prefer to MODEL/ simulate it. In my own tube amps, I never use over 50 uF, BUT, I also use multiple series chokes as in L1/C1/L2/C2. Let me see If I can look-up Mike Samra's post on what he did with a Cary to reduce it, it was a few years ago. We have better cap choices in late 2014.
Also, there are many makers of this new technology DC LINK cap, to name a few, WIMA, VISHAY, EPCOS, PANASONIC. The caps you use do NOT have to be round !!! You can " gut" the 1200 uF cans, and place the FAR FAR FAR better sounding DC LINKS inside, using GE Silicone 2 to hold them in place. These DC LINK caps are designed for pulsed response, which is exactly what an audio amp does, reproduces PULSED transients of the music.
I got a report today, of a St. Louis audiophile who replaced his Solen film caps banks with a 50 uF 800 VDC DC LINKS, and he said there was much more definition with the DC LINKS !! ( all DHT Type 45 amp ). No comparison. Its a bold new world in audio ...today !!
Jeff Medwin
Edits: 12/18/14 12/18/14
...the Cary amp he was referencing used a tube rectifier. 1200uF or 600uF immediately following the rectifier tube is WAY too much...overall, a very poor design. The primary purpose of reducing the value of those 1200uF caps was to eliminate rectifier tube arcing and excessively short life. With solid state rectifiers the input cap max safe value is much greater than with a tube rectifier.
I don't see a tube rectifier, do you see one ?
"Let me see If I can look-up Mike Samra's post on what he did with a Cary to reduce it, it was a few years ago."
Then you provide the following link: Mikes post about the Cary SLP90 .
If you read the entire thread, it will be clear Mike is working with a preamp utilizing a high-vacuum rectifier, NOT a SS rectifier. IOW, you're referencing a post where the primary goal is to reduce PS filter input capacitance to extend rectifier life. His report of improved sound is a nice side benefit. The PS you're working with uses SS rectifiers where the influence of the size of PS input filter capacitance on rectifier life is not the issue. Therefore, Mike's solution to his specific problem (arced rectifiers) may not be the optimum solution to your situation since you don't have vacuum rectifiers that are arcing.
I thought my post was quite clear about this but apparently it wasn't clear enough.
AOK. Good enough !! Happy Holidays to you and your family.
Jeff Medwin
Cary beat the living guano out of vacuum rectifiers. It's no small miracle that they found specimens that could survive the ABUSE , for even a short period of time. Those designs are hideous! :> ((
RTFM applies, "in spades", when dealing with vacuum rectifiers. Read the data sheets and take the info. provided to heart, seriously.
Eli D.
Thanks for the good advice.
First each KT88 is drawing about 35 mA (idle) at the moment. I would not go above 45 mA. Hope that the info you want.
As for gutting the 1200uF, I don't think I am skilled enough to attempt out of safety concern.
Looks like Mikey Samra used 140 uF total, which is about what I said...100 uF to 200 uF .Instead of "gutting" the big electrolytics ( Dremel tool, a messy job ), you can leave them in place, so the amp "looks" stock, and simply disconnect them. Place one to three 800 VDC rated DC LINK caps under the chassis, in parallel, to equal 140 uF, give or take 20 uF. Use good wire to connect them to the amp.
The REASON why I am pushing you on DC LINKS is because they will sound great, versus, say, an older style motor run cap. This is a new cap technology - much wider band, lower in ESR, and they are designed for PULSED operation, which IS our SE tube amp playing back music.
Jeff Medwin
Edits: 12/19/14
I'm very tempted to try out the DC Link caps too, but only if space permits. Not much room in the chaasis. If it comes to that, I suspect removing the electrolytics will give a better sound.
Allied handles Wimas and the less expensive Vishays, and has all the dimension specs in metric, for each uF / voltage. Just use 800 VDC rated for bullet-proof safety and reliability !!! I bet there is room, use GE Silicon 2 to attach caps, in a non-heavy field place.
We beat this to death. Any more, do through email, glad to try to help !!
Jeff Medwin
Owing to high B+ (~500VDC), my best option is to have 2 x Wima 700VDC, 190 uF (=380 uF) replacing the existing 600 uF. My main concern is about increased hum. Cary groundings are at convenient points on the steel chassis - not very nice. As suggested elsewhere on AA, I will connect all ground points to the -ve terminal of the filter cap and see if the amp will be more quiet. Even now, the hum is quite acceptable, but need extra margin for possible increase of hum.
Sonically speaking a slow recovery power supply using too much filtering is bad news.
"Sonically speaking a slow recovery power supply using too much filtering is bad news."
I guess an electric car with a smaller battery will go further?
This old tired Jeff Medwin theory is simply flawed. An amplifier or any electrical circuit will take what ever power it needs and that's all. It doesn't matter if that power or energy is stored in a capacitor or the current demand happens to come at the peak of the charge point. The VA rating of the power transformer must satisfy the needs of the amp circuit period. And if it does, "too much" filter capacitance makes no difference in terms of current delivery. Too little capacitance and the assocaited "charge pumping" is however a problem and is exactly what makes the LSES sound "dynamic".
Is it bad my amplifier is plugged into a major power grid across several states with comparably unlimited current reserve to the amplifier needs? Should I power my amplifier off a small Honda generator to balance the power supply to the current demand? Would that sound better?
Now if we want to talk about the effects of power supply impedance and how that effects the output stage, that's another matter and warrants some discussion of deeper issues. But that's not the same thing.
Do you build tube amps Gusser?
Do you build tube amps Gusser?Ahh, yeah, I do!
That's my former biamped system in my HT built in 2006. It has since been retired for DIY SS amps. I got tired of the re-tube costs.
I do run a DIY EL34 based amp in the music room.
Edits: 12/18/14
Fantastico. Love the passion.
What SS amp did you replace THAT with?
Funk ya!
Then the saying goes all power supplies sound the same. I am not a fan of ultra low power supplies with 10 volts of AC ripple either.
Hi,
I agree with you about 10 VAC of ripple. ANY idea what the ripple would be, to the Finals, to the Driver, on a modern LSES supply?
Answer : its about 800-900 mVAC to the Finals, no problem at all on high efficiency speakers, since the output transformer steps it down.
Second answer : its about one to two mVAC to the Driver, and the Driver's B+ is fully shunt regulated with a zero feedback, one part regulator, which gives fabulous stability and a wonderful low end.
The mu of 100 Driver "needs" that low mVAC, the mu of 4 Finals needs to be under 1 VAC if SE, and 2.5 VAC or less if Push Pull.
10 VAC would be bad. You weren't thinking 'ole Jeff is running 10 VAC as you wrote, were you ??
Regards,
Jeff Medwin
All power supplies do not sound the same.In fact a power supply with additional filter capacitance will better handle peaks.
And as I said, power supply impedance effects the so called "sound" especially in SET's.
Let's just be clear on what a LSES is and how it works. It is not a "properly designed" power supply for the intended load. It is current starved and that is exactly what produces the desired effect - artificial dynamics.
You could argue successfully that if turning an amplifier into a dynamics processor is the goal, then the power supply IS "properly designed" for that purpose.
What ever.
Again just be clear on how a LSES works and what in turn it does to the amplified audio signal. If that's what you want, then the LSES is in fact an elegant way to synthesize dynamics without resorting to analog VCA technologies, aka DBX, or digital processing techniques.
P.S. There is never a panacea in engineering either. The artificial dynamics boost comes at a price - more ripple induced hum and possibly more clipping.
Edits: 12/17/14
actually gusser, I'm starting to think "clipping" is part of the reason people prefer SE over push pull.
I especially think the 12ax7 DC into a 2a3 does soft clipping extremely well. (particularly with the bandwith limiting of small cap cathode bypass)
I don't agree with your assessment on the power supply, but there is some truth in the fact that the whole cap bundle, power supply, insufficient drive for miller etc all are a form of EQing.
Fine to disagree with my theory of how it sounds.
What is your technical analysis of the LSES power supply?
well... I think:
- due to class A and low current draw
- low mu finals allow for less filtering
- ac filaments means there is a noise floor regardless
- small chokes filter HF well
- ultra clean driver stage
that DFs supply works just fine for a 2a3 or 45. I don't think there is any "dynamics expansion."
I'm also working on my first amp with an l-critical choke input right now, so I'm not hard line. I do believe the supply works and is plenty quiet.
also in fairness, I'm actually running the power supply Jeff has now abandoned. So I'm working with a 2 henry choke.. (along with the flywheel input deal)
on a final note, I'll never cut up factory chokes or power transformer leadouts again either. :)
Nichicon makes good caps. You should have no problems using them. I would be very surprised if there was any audible difference between them in a PS filter application.
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: