The heatsinks I supplied in the RevA kits were 4" sections of the stuff sold by barrredboss on ebay. I think you can find a link to his stuff somewhere in the last 50-100 posts somewhere, shoudl be able to search by name.
I have purchased a bunch of 4" and 3" pieces from him, which we will be supplying, in case you don't want to buy a 12" or 24" piece. I'm waiting for them to arrive to determine how much we need to charge for them, and how it will affect shipping charges (especially international).
More to come on that soon. If you want more then 1 or 2 though, I would go right to him. Great guy, and he'll cut it up for you for free.
[edit] http://search.ebay.com/_W0QQsassZbarrredbossQQhtZ-1QQfrppZ50QQfsopZ1QQfsooZ1QQrdZ0?
I have purchased a bunch of 4" and 3" pieces from him, which we will be supplying, in case you don't want to buy a 12" or 24" piece. I'm waiting for them to arrive to determine how much we need to charge for them, and how it will affect shipping charges (especially international).
More to come on that soon. If you want more then 1 or 2 though, I would go right to him. Great guy, and he'll cut it up for you for free.
[edit] http://search.ebay.com/_W0QQsassZbarrredbossQQhtZ-1QQfrppZ50QQfsopZ1QQfsooZ1QQrdZ0?
maf_au said:Carlos, you're talking chinese
Can you explain the what it is about the schematics that is yours to ask people this? I'm not critisizing or judging what you say, I just don't understand where you are drawing the line as to what is 'yours'
Deam it, is it so hard to understand?
You can freely use the snubber, I have absolutely nothing against that, it's not mine.
But I did design these PSUs, much work was done to get to these exact values and final results, believe it or not.
Even if these are quite simple designs, they are MUCH better than a single pair of caps.
Are these mine?
Yours?
Respect, please.
I showed my Beta setup of the rev C a while back, using separate heat sinks that were rated 1.8 C/W, just as recommended by the National design guide for the parameters used in this amp.
I have since switched to both channels mounted to the heat sink Brian mentions that he used in the rev A kits. It runs about 4 deg C hotter, not really much to worry about.
My rev A uses a slightly smaller heat sink and it runs maybe a degree or two hotter, so I would say things are in the right ball park with what Brian recommends.
And yes, barredboss is great to work with.
I have since switched to both channels mounted to the heat sink Brian mentions that he used in the rev A kits. It runs about 4 deg C hotter, not really much to worry about.
My rev A uses a slightly smaller heat sink and it runs maybe a degree or two hotter, so I would say things are in the right ball park with what Brian recommends.
And yes, barredboss is great to work with.
We all want to do better
For a snubber to be optimized, should consideration be given to the transformer's characteristics of resistance, capacitance, inductance, etc.? If there be truth in that, then would failure to consider those attributes leave one open to the failure of the power supply system in meeting requirements for ripple, impedance or % regulation?
An EI transformer has different characteristics than a torroid. I submit that a power supply optimised for the torroid would not perform the same with an EI, and vice versa.
I think this leaves one in a situation when they choose to use snubbers, the final "alignment" should be done to suit (match) the transformer.
For a snubber to be optimized, should consideration be given to the transformer's characteristics of resistance, capacitance, inductance, etc.? If there be truth in that, then would failure to consider those attributes leave one open to the failure of the power supply system in meeting requirements for ripple, impedance or % regulation?
An EI transformer has different characteristics than a torroid. I submit that a power supply optimised for the torroid would not perform the same with an EI, and vice versa.
I think this leaves one in a situation when they choose to use snubbers, the final "alignment" should be done to suit (match) the transformer.
carlosfm said:
Deam it, is it so hard to understand?
You can freely use the snubber, I have absolutely nothing against that, it's not mine.
But I did design these PSUs, much work was done to get to these exact values and final results, believe it or not.
Even if these are quite simple designs, they are MUCH better than a single pair of caps.
Are these mine?
Yours?
Respect, please.
Thanks for clearing that up Carlos. I think I understand you now.
I don't know if you intended it, but you sound annoyed at my question. That was not intended, please don't be angry. I am not trying to take your designs, or disrespect you, just understand you.
Michael
float said:
Sorry Float,
Meant to answer that before.
You can find some info here which should help:
http://www.diyaudio.com/forums/showthread.php?postid=683346#post683346
And here is even better stuff.
http://www.diyaudio.com/forums/showthread.php?postid=702645#post702645
Cheers!
Russ
Russ White said:
And here is even better stuff.
http://www.diyaudio.com/forums/showthread.php?postid=702645#post702645
Cheers!
Russ
Thanks again Russ; completely missed that thread.
Snubber Theory
Ed,
Do you remember where that paper on snubbers was listed? I also remember that the critical elements were transformer and diode parameters used to calculate snubber values. It would be worth digging it out and seeing what values come from that paper based on the typical trannies we use.
I did notice a difference on the rev C beta between the cheapo Parts Express trannie and an Avel 330 VA I am now using, but have not tried to measure the underlying difference.
For what it is worth, I am still burning in a BrianGT 3886 with Carlos snubber PS that I can directly compare with the rev C. Right now, I prefer the rev C, but both are being fed directly from the CD player through an ALPS 50K pot and it might be more reasaonable to have a buffer in the BGT 3886.
Ed,
Do you remember where that paper on snubbers was listed? I also remember that the critical elements were transformer and diode parameters used to calculate snubber values. It would be worth digging it out and seeing what values come from that paper based on the typical trannies we use.
I did notice a difference on the rev C beta between the cheapo Parts Express trannie and an Avel 330 VA I am now using, but have not tried to measure the underlying difference.
For what it is worth, I am still burning in a BrianGT 3886 with Carlos snubber PS that I can directly compare with the rev C. Right now, I prefer the rev C, but both are being fed directly from the CD player through an ALPS 50K pot and it might be more reasaonable to have a buffer in the BGT 3886.
Re: Snubber Theory
Hi Craig,
The BGT amp is a NIGC, and has a high inputZ so no buffer is necessary unless you have a very highZ source.
Cheers!
Russ
P.S. I think Eds post was more about the absurdity of calculating one single set of values for snubbers which would work for any chipamp. And Eds observation is spot on, even if subtle.
hayenc said:
Hi Craig,
The BGT amp is a NIGC, and has a high inputZ so no buffer is necessary unless you have a very highZ source.
Cheers!
Russ
P.S. I think Eds post was more about the absurdity of calculating one single set of values for snubbers which would work for any chipamp. And Eds observation is spot on, even if subtle.
Russ,
Yes, Ed and I are in complete agreement. Theory says the snubber is calculated on the basis of transformer and diode characteristics and the Mauro and BGT's use different diodes. So even with the same trannie, need to check the value derived from the diodes. The other question, of course, that you allude to is the sensitivity to choice of values. Not clear how well tuned the snubber values need to be to these parameters to be effective. Might be worth examining, but for now, just enjoying the amps as they are.
Once I get as many hours on the BGT as on the rev C, I will see how they really compare, but for now, it is hard to meet the image and detail on the C.
Of course, will be interesting to see how both work with Kookabura.
Yes, Ed and I are in complete agreement. Theory says the snubber is calculated on the basis of transformer and diode characteristics and the Mauro and BGT's use different diodes. So even with the same trannie, need to check the value derived from the diodes. The other question, of course, that you allude to is the sensitivity to choice of values. Not clear how well tuned the snubber values need to be to these parameters to be effective. Might be worth examining, but for now, just enjoying the amps as they are.
Once I get as many hours on the BGT as on the rev C, I will see how they really compare, but for now, it is hard to meet the image and detail on the C.
Of course, will be interesting to see how both work with Kookabura.
Russ,
I just hope we can make it a more technical and scientific investigation and not tick off anyone again. As a scientist, I know the process, including intuition and things not yet demonstrated with measurements and theory, and the need to make measurements to better understand and parameterize those characteristics.
But then, as I a scientist, I am also used to sensitive personalities, so know that challenge also. Humility, humor and a genuine desire to understand.
Let's work some theory and make some measurements.
Craig
I just hope we can make it a more technical and scientific investigation and not tick off anyone again. As a scientist, I know the process, including intuition and things not yet demonstrated with measurements and theory, and the need to make measurements to better understand and parameterize those characteristics.
But then, as I a scientist, I am also used to sensitive personalities, so know that challenge also. Humility, humor and a genuine desire to understand.
Let's work some theory and make some measurements.
Craig
hayenc said:
Right exactly!! Excellent points. I am sure you would agree then that the best way to make samples and perform experiments is not to poison/taint the sample, so as to predispose it to a designed end. You must always have a control. Single ended tests are practically worthless.
Cheers!
Russ
Yep, and I wish we scientists never fell into that trap, but we are human also. That is the reason we stress science being a community effort, with results that can be communicated to others, repeated by them, and that they then generate the same reproducible results.
Before I get slammed too much for that, I also very much respect the individual who goes against the stream, as my own research took years to get repeated and accepted. And I very much respect those things we have not figured out how to measure and may never figure out how to measure. We still need to have our subjective experiences and value them, but the trick is to know when to do science and when we can leave it behind and just enjoy the beauty.
Cheers, and tonight is an nice new Irish Whiskey I just got.
Before I get slammed too much for that, I also very much respect the individual who goes against the stream, as my own research took years to get repeated and accepted. And I very much respect those things we have not figured out how to measure and may never figure out how to measure. We still need to have our subjective experiences and value them, but the trick is to know when to do science and when we can leave it behind and just enjoy the beauty.
Cheers, and tonight is an nice new Irish Whiskey I just got.
but she sent me a trashy e-mail...
Guys, its way past my bedtime but when I heard the cackling from the computer room I had to investigate.
Jim Hagerman's "Calculating Optimum Snubbers" is a 13 page pdf found here:
http://www.hagtech.com/pdf/snubber.pdf
and I'm heartened with the interest you bring to the topic.
Guys, its way past my bedtime but when I heard the cackling from the computer room I had to investigate.
Jim Hagerman's "Calculating Optimum Snubbers" is a 13 page pdf found here:
http://www.hagtech.com/pdf/snubber.pdf
and I'm heartened with the interest you bring to the topic.