Hi!
I want to build a Borbely DC100 with a more sophisticated power supply by having 2 power transformers for each channel, one transformer handle positive rail and the other for negative, any thoughts?
Also, don't you think having Salas regulator for the DC-100 (both driver and output stage) for good result?
Thanks!
I want to build a Borbely DC100 with a more sophisticated power supply by having 2 power transformers for each channel, one transformer handle positive rail and the other for negative, any thoughts?
Also, don't you think having Salas regulator for the DC-100 (both driver and output stage) for good result?
Thanks!
I had DC 100s, 102, with 700,000 uf with LCL chokes and final 100,000 uf Blackgates . Now i changed to Milleniums.
Borbely DC 102 and Millenium have on board front end regulators.
Borbely DC 102 and Millenium have on board front end regulators.
DC 100
Blast from the past!! I built two of these for use with the SWAN FOUR D'apolito
design. Why not three trannies ..two with regulators for the respective front ends and one high capacitance setup for outputs. the frontenders from apex jr.
Are u using AUDIO AMATEUR PCB"s??
Blast from the past!! I built two of these for use with the SWAN FOUR D'apolito
design. Why not three trannies ..two with regulators for the respective front ends and one high capacitance setup for outputs. the frontenders from apex jr.
Are u using AUDIO AMATEUR PCB"s??
i think the sales regulators for such an output stage would need more heatsink than the amp is what i think. its a fantastic design and Erno's regs are some of if not THE BEST discrete regulators in the business, i wouldn't replace them, salas drew plenty of inspiration from him
As a dual mono amp - one transformer for each channel - you can use one xfrmr 4 for each rail...xfmrs sure, you can do that. Not much difference between that and paralleling two...
...max the filter caps. More ufd is good.
...consider hexfred type rectifiers.
Doing something to regulate the supply to the front end is a good idea.
Also improve the heatsink for the drivers. Make sure the driver heatsinks if made larger are also grounded to the PCB.
Do not try to "push" the rail voltages up.
You can add an extra pair of outputs per channel if you like... probably not a great idea to add two pair per channel.
Use adequate heatsinking - more than you think you need.
Regulating the output stage is impractical and may not sound any good anyhow.
Watch your ground scheme and keep the ground connections low impedance.
_-_-bear
...max the filter caps. More ufd is good.
...consider hexfred type rectifiers.
Doing something to regulate the supply to the front end is a good idea.
Also improve the heatsink for the drivers. Make sure the driver heatsinks if made larger are also grounded to the PCB.
Do not try to "push" the rail voltages up.
You can add an extra pair of outputs per channel if you like... probably not a great idea to add two pair per channel.
Use adequate heatsinking - more than you think you need.
Regulating the output stage is impractical and may not sound any good anyhow.
Watch your ground scheme and keep the ground connections low impedance.
_-_-bear
Actually how many transformer per channel? I confuse.
I have a hard time to locate a proper capacitor, have thought of Black Gate but they ceased production for most value, now I have Panasonic, Hitachi and DNM and ELNA RFS Silmic II left on the list........
Hey, what capacitance you suggest can push the amplifier to limit? Also, how many extra output devices you really suggest?
Don't bother with designer capacitors 😉 Good quality low ESR types are all you need. 1 transformer per channel if monoblocks is fine. 1 big one for both channels is also fine. Don't increase the amount of mosfets as it will have an effect on other areas such as compensation (unless you really know what you are doing... but you wouldn't be asking these questions if you knew that much 🙄) The law of diminishing returns applies to how many capacitors you should use.... plus the need for soft start if you go too far!
Make sure you apply the corrections to the original design re the zener supplies, also the caps after the zeners need to be bigger and there is a mod with some small caps across the mosfets IIRC. I built the servo 100 and found all these things to be necessary (I believe the relevant areas are the same on the DC100...?). Without them it sounded average at best.
For FILTER caps you just want the most ufd you can get cheap... you don't really need low ESR if you bypass the filter bank with something like a decent size polypropylene...
The changes are documented in Audio Amateur... iirc.
I suggest going too far and using a soft start...
WRT to the transformers, use the highest VA that is practical above the minimum required - that being what is suggested in the original article.
It matters not if you wire them in parallel or use separate xfmrs for each rail.
_-_-bear
The changes are documented in Audio Amateur... iirc.
I suggest going too far and using a soft start...
WRT to the transformers, use the highest VA that is practical above the minimum required - that being what is suggested in the original article.
It matters not if you wire them in parallel or use separate xfmrs for each rail.
_-_-bear
Low ESR caps are not of course strictly necessary but personally I prefer that route over adding large polyprops etc as the low ESR is maintained to lower frequencies. You can always add some polyprops as well of course 🙂
Hi Bear!
My friend and I decided to build the power supply differently. 2 x 2kVA toroid power transformer, one for positive and the other for negative "per channel" with Crees Schottky diode and CLC feeding regulator for the output stage.
Since we have regulator on the output stage and we can put much smaller capacitance, let's say 10,000mf DNM capacitor. Also, 10,000mf between rectifier and the filter choke, the choke will be low DCR toroid choke for both positive and negative individually. Therefore each channel has 40,000mf reservoir for both positive and negative before the regulator.
Of course, the driver stage will be regulated by Salas shunt regulator.
Any thoughts??
My friend and I decided to build the power supply differently. 2 x 2kVA toroid power transformer, one for positive and the other for negative "per channel" with Crees Schottky diode and CLC feeding regulator for the output stage.
Since we have regulator on the output stage and we can put much smaller capacitance, let's say 10,000mf DNM capacitor. Also, 10,000mf between rectifier and the filter choke, the choke will be low DCR toroid choke for both positive and negative individually. Therefore each channel has 40,000mf reservoir for both positive and negative before the regulator.
Of course, the driver stage will be regulated by Salas shunt regulator.
Any thoughts??
have fun... it will make sound.
might be great.
I happen to think it is very problematic to regulate the output stage - in effect you have
a reverse amplifier... it seems unnecessary if you have a good CLC filter. Imo.
I'm not quite clear on how you will use the xfmrs... if you use the same xfmr for two channels, you will get some crosstalk. Better to use one xfmr per channel, imo. But it may be ok anyhow.
_-_-bear
might be great.
I happen to think it is very problematic to regulate the output stage - in effect you have
a reverse amplifier... it seems unnecessary if you have a good CLC filter. Imo.
I'm not quite clear on how you will use the xfmrs... if you use the same xfmr for two channels, you will get some crosstalk. Better to use one xfmr per channel, imo. But it may be ok anyhow.
_-_-bear
The way we do just like Mark Levinson #20.6 monoblock. Each ML 20.6 monoblcok has 2 x 600VA power transformers, one for the positive and the other for the negative.
But we add a CLC filter network between the rectifier and the regulator.
But we add a CLC filter network between the rectifier and the regulator.
Btw, you have a DC-100 board and not a DC-100A board??
So you are building a monobloc pair?
If so, then fine to use the transformers as you say...
If you are building a stereo chassis, then I would not use two xfrmrs with separate rectifiers and filter caps for each of the two rails. It will work, of course.
Again, I see no benefit to the regulators, and certainly not after a robust CLC...
regulate or CLC filter the front end, ok.
_-_-bear
So you are building a monobloc pair?
If so, then fine to use the transformers as you say...
If you are building a stereo chassis, then I would not use two xfrmrs with separate rectifiers and filter caps for each of the two rails. It will work, of course.
Again, I see no benefit to the regulators, and certainly not after a robust CLC...
regulate or CLC filter the front end, ok.
_-_-bear
Last edited:
- Status
- Not open for further replies.