| falcott |
Hi everyone (and experts), this is my first post.
[Intro] It's been a few decades since I was a kid building crystal sets with my grandad (my guru in all things life and electronic), but since he has gone to the big amplifier in the sky, I have to direct my questions to you good people.
[Down to business] I am starting a valve buffered IGC (a la Joe Rasmussen), my first electronic project for some years. My first questions concern the power supply. :scratch: I would like to put caps over/parallel with the bridge diodes (MUR860).
1. Is this worth it (for sound/whatever reason)?
2. Should they be ceramic or film?
2. What values would be best (a) for the amp, and (b) for the valve? (I assume that the filament supply would not need caps over the diodes.)
Thanks!!!:happy1: |
|
|
| Peter Clements |
Welcome aboard.
Think it is great idea to snubb out the spikes at the source. The bridge will be going off like a machine gun at the peak of the wave on each cycle charging the first cap.A choke input is highly desirable to help control this.A valve rectifier even better. IMHO.
You can make a bridge with "better" diodes such as Shottkeys or Hexfreds to help here.
A .01 1KV cap and a 470K 1W across each juntion is a good start. Others have used both components in series to make a snubber. Lynn Olsen and Budda Camile have done some reserch on this subject and have written papers.Lynn found the quietest rectifer to be damper diodes. These are cheap rugged and not only sound wonderful but take 30 seconds to slowly ramp up the voltage , protecting and extending the life of youur components. What more could you ask for. |
|
|
| falcott |
| quote: | Originally posted by Peter Clements
Think it is great idea to snubb out the spikes at the source. The bridge will be going off like a machine gun at the peak of the wave on each cycle charging the first cap.A choke input is highly desirable to help control this.A valve rectifier even better. IMHO.
You can make a bridge with "better" diodes such as Shottkeys or Hexfreds to help here. |
Yes, really an amp is nothing more than a modified power supply, so why not clean it up right at the beginning. (And now seems a good idea to ask another dumb question: I suppose “IMHO” isn’t short for “I Mean, Hold On”? So many people on the net use it and I have no idea as to what it means.)
| quote: | | A .01 1KV cap and a 470K 1W across each juntion is a good start. Others have used both components in series to make a snubber. Lynn Olsen and Budda Camile have done some reserch on this subject and have written papers.Lynn found the quietest rectifer to be damper diodes. These are cheap rugged and not only sound wonderful but take 30 seconds to slowly ramp up the voltage , protecting and extending the life of youur components. What more could you ask for. [/B] |
Your reply (wonderful, thank you) has really inspired further questions, and a different angle of attack, so I would like to treat them as separate entities:
Extension to original question: Are you saying that I should use a cap and a resistor across each diode?? Don’t resistors reduce the voltage?? And to repeat a purely practical question: Is it better to use ceramic or film caps here?
Next topic: As a challenge to conventional wisdom, damper diodes seem to be a retrogressive step, although I would not doubt their effect on sound quality. (Look out; more dumb questions ahead.) Surely fast diodes regulate the current better for the dynamics in music? How, then, do damper diodes affect the dynamics? Apart from that the concept seems a sound (no pun intended) one. Can you suggest type/part nos? Also, does the slow ramp up with these diodes smooth out the power up thumps?
Be gentle with me; it’s been a long time... |
|
|
| Peter Clements |
Hi again falcott,
IMHO=In my humble opinion.
In audio now there as many people as ideas to what is best.Some would say that a valve buffer is a retrograde step.Audio DIY is a process that changes your ideas and opinions through the journey. I am a junior. Joe Rasmussen along with fellow Kiwi Allen Wright make some of the best audio equipment in the world. Allen was brought up in the sand world (transistor)and has stated publicly that his valves would be out the door real fast if he could make as good a sounding gear with with only sand in it.In fact he also mixes FETs and and sand regulators with his valves in his creations.What is a retrograde step? Some common 1920's triodes have distortion so low that it is hard to measure today.These are the most linear amplifying devices known to man Their distortion characterististics match the ears distortion curve almost exactly. The brain filters out the distortion of the ear to hear a pure tone, just as you cannot hear your own heart beat.This is why in tests in the real world show that most people cannot hear 2nd harmonic distortion until it is quite high, however odd order distortion is very hearable.
So IMHO if you stick with a bridge, put some ceramic caps across the terminals. This would be the minimum. With the caps you could add some resistors, yes this will soften the bridge, but that is what we are doing. Or series the components across the terminals to act as a snubber, there won't be much DC leakage here.A definate upgrate would be damper diodes IMHO. However remember there is more voltage drop, 10-20v, heat, and a a lot of power to their heaters. They will stop any on thump . A 6DQ3 is rated at max 1200ma max piv 5500v.The heaters are hungry at 1.8v each. No free lunch. |
|
|
| falcott |
Peter, thanks for the leg-up. Don’t get me wrong, retro is what I like. To begin this venture back in to electronics, I was surfing around the net looking for a valve amp to build where I came across numerous references, raves and claims about Gainclones. The idea sounded simple enough, and remarkably effective, and the idea of a hybrid excited me far more when I read of Joe’s VBIGC. I think hybrids are cool, like a blown hemi in a ’32 coupe. (That’s hot rod, guys.)
At the end of the day, it is the music we are listening to, not to the specifications.
Anyhoo, on the subject of hybrids, I searched the Tubes forum and found reference to hybrid bridges using 2x solid state diodes and 2x vacuum diodes, and also to Schottkys with a single TV damper diode. For now I will build my project with MUR860 diodes and ceramic caps, and will experiment with a damper diode set-up when I get the amps up and running. Any leads on a schematic for (single tube?) hybrid bridge rectifiers??
Let’s put some sand in a bottle and shake it up a bit...
Cheers, Peter!
:up: |
|
|
| falcott |
Apparently Schottkys when used in a bridge rectifier are totally silent (when within specified rating) and have no need for snubber caps. I wondered why everyone are not using Schottkys by choice, for they obviously make the bridge so simple; perhaps it is because they sound different to fast diodes with cap snubbers? I think I will give Schottky bridges a try with this project. Are these higher specification bridges an advantage even for the valve buffer heater/filament? I imagine that the type of bridge used for the filament will have no effect on the sound of the amp...
Thanks again for the leg up. :) |
|
|
| Peter Clements |
Hi falcott,
Try it all. Have your mains trans wound with the correct multitap voltages , make it large and put your projects on a big chassis.I for years had a huge amp with two power supplies that were mulitaped and used the large cheap valve sockets that are easily interchangable, and tried all sorts. I ended up with boxes of caps and valves. And now I know what I like.For example with coupling caps I have a lot of the "wonders" and prefer the cheap RS film and foil caps.
A lot has to do with what speakers that you are using. Many very good amps are wasted with average speakers. When I was using Lowther speakers the local Gainclone agent brought around the, Japanese original, amp. It was OK. But still flat and anemic sounding.A DHT 45 amp would be miles ahead.You can build such a 45 or 46 amp with as few as 5 components in the direct signal path. However put this amp on some plastic cones with cross overs and the tables may well be turned.
You may be suprized at how the heater power supply afects the sound. To my ears AC and current sorced sound best.
If Joes amp apeals to you go for it and enjoy.
Cheers,
Peter.
By the way what area do you live in. |
|
|
| falcott |
| quote: | Originally posted by Peter Clements
By the way what area do you live in. |
I'm in St Heliers (Auckland). |
|
|
| Peter Clements |
Hi falcot,
I am in Birkenhead. Small world.
If you would like to compare an SE amp or what ever mail me.We also have a group of grey hairs that get togeather to chew the fat on audio stuff , amazing collection of knowledge to tap into.
Cheers,
Peter. |
|
|
| falcott |
| quote: | Originally posted by Peter Clements
A .01 1KV cap and a 470K 1W across each juntion is a good start. Others have used both components in series to make a snubber. |
Yes, maybe we can get together, Peter. It will be a little while before we can try out my amp: it is still a box of assorted components, with some yet to arrive.
In regards to the above recommendation, isn't the 470K resistor a bit high? Perhaps a 100 or 200R would be better? |
|
|
| Peter Clements |
Hi falcot,
Diode noise is generated by the turn off characteristic. This is a reverse pulse that excites the LC resonant circuit of the secondary winding of the transformer. This LC circuit produces a burst of RF energy centered on the resonant frequency of your transformer. The fundermental frequency can be between about 5 to150K. This RF transmitter in your amp is the thing that I have against diodes.These pulses are radiated to other parts of the amp and amplified with your signal and become a real sonic killer. This RF is coupled back into the AC mains and can affect your other unprotected low level stages.Schottkys have few minority carriers and dont affect the LC circuit as much, however they are harder to tame.
The value of the resistor as you can see is set experimentally. It has been said that a good starting point is the DCR of your transformer.A .5 watt carbon and not a wire wound.You will need a good scope to see these bust effects.
"Better" transformers ie C core or toroids are worse than old EIs with a high secondary R. These "better" transformers will need more than a simple R+C to fix.I some times wonder how much of that grey flat sound of sand amps is due to the woefully inadequate or down right rubbishy power supplies that they all use and copy again and again.As you can see it is not any easy thing to fix and this is properbly why people ignore it.They pretend that is does not exist.Each design will take quite a bit of work to totally fix. This is what makes tubes easy. Plug and play.
Cheers,
Peter. |
|
|
| falcott |
Hi Peter, one more question, if I may. Regarding Joe's VBIGC power supply:
http://members.ozemail.com.au/~lisaras/ps.htm
Joe uses Schottky diodes in the amp bridges, but for the valve bridge he has 1A 1000V fast recovery (UF4007?) diodes. Can you suggest why the voltage rating for the valve's diodes should be so high, since only 35v are going through them, and why 1A 100V Schottkys are not used here (or if, indeed, they can be)? :confused:
Thanks for being able to walk me through my sandbox! |
|
|
| Peter Clements |
Hi falcott,
Did Joe mean FREDs for the valve bridge? Another flavour of super diode. There is nothing wrong with over rating diodes. But UF4007 s are real over kill here. I do not know why Joe does as he does with his diodes. I would just follow his recomendations.
I did remember that Joe used a 6922 but I did not realise that he used as low a voltage as he does. FREDs would be great here.
Cheers,
Peter. |
|
|
| falcott |
| quote: | Originally posted by Peter Clements
FREDs would be great here. |
Except that the ratings for FREDs are *HUGE*. Schottkys are less harsh, therefore seem to me to be a better bet, but they are not made in such high V rates that Joe specifies.
I dunno... |
|
|
| Peter Clements |
Hi falcott,
Well use Schottkys then. As I say it does not matter to over rate, but I cannot get into Joe's head to know why he has asked for Freds. A lot of designers do over rate when they are spending other peoples money.
I have noticed that RS in Mt Wellington have only a small stock of Schottkys now, so sourcing from there is not as good as it used to be. Your choice may be limited if you use them.
Cheers,
Peter. |
|
|
| falcott |
Hi Peter,
I have sourced most of my components through Farnell, and they seem to have a reasonable selection. I'm almost ready to place another order with them. They've offered me an account, with 20% discount on my first order, so I am waiting more on them than on deciding what I want from them
Cheers. :) |
|
|
| carlmart |
| quote: | Originally posted by falcott
Except that the ratings for FREDs are *HUGE*. Schottkys are less harsh, therefore seem to me to be a better bet, but they are not made in such high V rates that Joe specifies.
I dunno... |
Why don't you try the Motorola MUR types everyone is having good results with?
When this discussion on which diode was best for audio started (I read it on The Audio Amateur, but I think Stereophile was raising it) the options seemed between standard diodes, Schottky diodes, FRED diodes and fast/slow diodes.
The latter seemed to be the best sounding. Other people tests with them that came later, many on this Chips Amps Forum, confirmed those first findings by listening. I should put some credibility on that first.
Doesn't Farnell strock some fast/slow diode? Voltage shouldn't be a problem, as on Schottkies. Try them and see.
Carlos |
|
|
| carlosfm |
I tried MBR16100 (Schottkys) an I like them.
It's a 3 leg TO220 diode.
I use 2 legs in parallel in each diode, 16A.;) |
|
|
| falcott |
I'm wading through a lengthy thread on the solid state forum on high speed diodes vs fast recovery vs schottky vs anything else...
http://www.diyaudio.com/forums/show...&threadid=12539
Good stuff, and worth a read. It has answered most of my questions that I couldn't find whilst searching the chip amp forum. Check it out! :D |
|
|
| mikelm |
| quote: | Originally posted by Peter Clements
Diode noise is generated by the turn off characteristic. This is a reverse pulse that excites the LC resonant circuit of the secondary winding of the transformer. This LC circuit produces a burst of RF energy centered on the resonant frequency of your transformer. The fundermental frequency can be between about 5 to150K. This RF transmitter in your amp is the thing that I have against diodes.These pulses are radiated to other parts of the amp and amplified with your signal and become a real sonic killer. This RF is coupled back into the AC mains and can affect your other unprotected low level stages.Schottkys have few minority carriers and dont affect the LC circuit as much, however they are harder to tame.
|
As you say, this RF is coupled back in to the mains along with all the other rubbish from all the other noisy bridge rectifiers connected to your local substation and probably a good dose of every other type of RF interference known to man.
In light of this I am at a loss to understand why people then put caps across the diodes and couple all this garbage straight into the amp.
Try it and see what you think...
But I think you will find a 1uF cap directly across each secondary will produce a much better effect. This will reduce the resonant frequency to a much lower value and filter out some of the mains garbage. This mod has HUGELY improved the sound of any Power supply that I have tried it on.
If you want to reduce the frequency of ringing AND damp it to about 1/2 a cycle try 0.1uF across the secondary and 1uF and 470 ohms across the 0.1uF. In theory this may be a better solution but I have not tried it yet. |
|
|
| Peter Clements |
As you say, this RF is coupled back in to the mains along with all the other rubbish from all the other noisy bridge rectifiers connected to your local substation and probably a good dose of every other type of RF interference known to man.
Correct. This is why a faraday sheild in the transformer is absolutley vital.
In light of this I am at a loss to understand why people then put caps across the diodes and couple all this garbage straight into the amp.
As I pointed out work by "Budda" Camille as taught us the diode **** can be totally wiped out if you know how to do it. Standard diodes being easier than Schottkys to fix. |
|
|
| mikelm |
Sounds interesting...
do you know where this can be found ? .... google didn't find it.
thanks
mike |
|
|
| Peter Clements |
| quote: | Sounds interesting...
do you know where this can be found ? .... google didn't find it.
|
Page 39 Fall 1994 Sound Practices.
John does say " Done properly, silicon diode supplies can be built that are quieter than untreated Vacum diode designs."
I have not tried any of this myself. His methods and cures look complicated to me. They are not simple snubbers. |
|
|
| falcott |
| quote: | Originally posted by mikelm
If you want to reduce the frequency of ringing AND damp it to about 1/2 a cycle try 0.1uF across the secondary and 1uF and 470 ohms across the 0.1uF. In theory this may be a better solution but I have not tried it yet. |
Do you mean, therefore, that the 0.1uF, the 1uF and the 470 ohms are in parallel, between the secondary -ve/+ve?
Thanks for your input, by the way. This is a steep (re)learning curve for me, but I'm getting a good handhold on the way up. |
|
|
| falcott |
Another question, if I may:
I intend to use snubbers between the Tx secondaries and the diodes in the bridge. I am using a single 300va 2x25v. For the resistors in series with the secondaries, what wattage rating aught to be used? (And should they be wirewound?)
:scratch2: |
|
|
| mikelm |
Do you mean, therefore, that the 0.1uF, the 1uF and the 470 ohms are in parallel, between the secondary -ve/+ve?
yes
Thanks for your input, by the way. This is a steep (re)learning curve for me, but I'm getting a good.
Well, trust your own ears, I hope this idea helps
mike |
|
|
| mikelm |
| quote: | Originally posted by falcott
Another question, if I may:
I intend to use snubbers between the Tx secondaries and the diodes in the bridge. I am using a single 300va 2x25v. For the resistors in series with the secondaries, what wattage rating aught to be used? (And should they be wirewound?)
:scratch2: |
Any 1 watt metal film resistor should be fine. Obviously the cap should have voltage rating higher than 36V but there again, they normally do. |
|
|
| mikelm |
| quote: | Originally posted by Peter Clements
Page 39 Fall 1994 Sound Practices. |
How do find this ?
After 30 mins of looking around I'm none the wiser.
mike |
|
|
| Peter Clements |
How do find this ?
After 30 mins of looking around I'm none the wiser.
mike[QUOTE]
Sound Practice magazine seems to have gone. The back issues on www.southernelectric.com has diappeared. This is a great shame as there was so much information shared in this journal. Maybe someone with a working scanner can send you a copy. |
|
|
|
