Improving the power supply of the TFM series.

Custom Equipment Mods.
Post Reply
User avatar
brit01
TFM-45
Posts: 580
Joined: Sat Jul 12, 2008 3:29 pm
Location: Montevideo, Uruguay

Improving the power supply of the TFM series.

Post by brit01 » Mon Aug 30, 2010 1:24 pm

I've spent a couple of years working with valve amps now and discovered tweaking the power supplies can make a big difference in the sonic capabilities.

I'm planning to look at the rectifiers and caps in my TFM-55.

Probably will experiment with changing to ultra fast rectifiers (maybe Hexfreds), and bypass the lytics with film/PIO etc.

Also have upgraded some resistors.

But the psu will be my next project.

Has anyone else modded the power supplies?
TFM-55, CT-3, Klipsch F-1, Rotel RCD 930AX

User avatar
BillD
R.I.P. Friend
R.I.P. Friend
Posts: 7126
Joined: Wed Jun 06, 2007 8:29 pm
Location: The west's most mid-western town, Scottsdale, Arizona

Re: Improving the power supply of the TFM series.

Post by BillD » Mon Aug 30, 2010 2:12 pm

Remember that these power supplies aren't anything like you've encountered in a tube amp. The input to the thing that looks like a transformer but is really a second cousin that will burn out if exposed to the full duty cycle of A/C (50Hz or 60Hz). So, what Carver does is use a triac to control the angle (tangent in the sin wave) and duration of the input to the PRIMARY of the transformer (mag field coil), resulting in an incomplete sin wave that looks mostly like this:
triac output.jpg
triac output.jpg (32.16 KiB) Viewed 1239 times
where the length of the wave dictates how much output the secondaries produce. This duration is controlled by feedback from the amplifier. When the voltage starts to sag, the amplifier tells the power supply to lengthen the conduction time on the primary. But, as far as rectification goes, you can see that you aren't going to get more than 60Hz to play with, so ultra-fast rectifiers, necessary in digital power supplies, aren't necessary here. What could help are ultra-fast switching diodes and maybe snubbers. The last upgrade I made to an M-500t, I changed the switching diodes from 200ns devices to 50ns. As far as snubbers go, HERE is a pretty good guide published by Cornell Dubilier.
It should sound like it isn't there!
There is a difference between hearing and listening...
Making life enjoyable through expensive electronics.
_________________
Carver: C-4000 & C-1 preamps, PSC-60 preamp/tuner, TX-11a tuner, M-400 (2), C-500, M-500, M-500t, M-500t Mk.II, A-500x, AL-III loudspeakers (2 pr.)
Sunfire:Theater Grand III processor, Ultimate Receiver, Cinema Grand Signature 400 ~ seven, True Subwoofer Mk. II, D-10 Subwoofer

User avatar
kingman
(pair of) SILVER-7 TUBE AMPs
Posts: 6406
Joined: Sun Nov 30, 2008 2:10 pm
Location: Upstate South Carolina

Re: Improving the power supply of the TFM series.

Post by kingman » Mon Aug 30, 2010 3:56 pm

BillD wrote:Remember that these power supplies aren't anything like you've encountered in a tube amp. The input to the thing that looks like a transformer but is really a second cousin that will burn out if exposed to the full duty cycle of A/C (50Hz or 60Hz). So, what Carver does is use a triac to control the angle (tangent in the sin wave) and duration of the input to the PRIMARY of the transformer (mag field coil), resulting in an incomplete sin wave that looks mostly like this:
triac output.jpg
where the length of the wave dictates how much output the secondaries produce. This duration is controlled by feedback from the amplifier. When the voltage starts to sag, the amplifier tells the power supply to lengthen the conduction time on the primary. But, as far as rectification goes, you can see that you aren't going to get more than 60Hz to play with, so ultra-fast rectifiers, necessary in digital power supplies, aren't necessary here. What could help are ultra-fast switching diodes and maybe snubbers. The last upgrade I made to an M-500t, I changed the switching diodes from 200ns devices to 50ns. As far as snubbers go, HERE is a pretty good guide published by Cornell Dubilier.
Yea...what he said. :-k
In reality, it only matters what it sounds like to YOU!!!

User avatar
BillD
R.I.P. Friend
R.I.P. Friend
Posts: 7126
Joined: Wed Jun 06, 2007 8:29 pm
Location: The west's most mid-western town, Scottsdale, Arizona

Re: Improving the power supply of the TFM series.

Post by BillD » Mon Aug 30, 2010 4:07 pm

Sorry, forgot where I was. :D
It should sound like it isn't there!
There is a difference between hearing and listening...
Making life enjoyable through expensive electronics.
_________________
Carver: C-4000 & C-1 preamps, PSC-60 preamp/tuner, TX-11a tuner, M-400 (2), C-500, M-500, M-500t, M-500t Mk.II, A-500x, AL-III loudspeakers (2 pr.)
Sunfire:Theater Grand III processor, Ultimate Receiver, Cinema Grand Signature 400 ~ seven, True Subwoofer Mk. II, D-10 Subwoofer

User avatar
kingman
(pair of) SILVER-7 TUBE AMPs
Posts: 6406
Joined: Sun Nov 30, 2008 2:10 pm
Location: Upstate South Carolina

Re: Improving the power supply of the TFM series.

Post by kingman » Mon Aug 30, 2010 4:09 pm

BillD wrote:Sorry, forgot where I was. :D
Just wish I had that kind of knowledge Bill!!! =D> =D> =D>
In reality, it only matters what it sounds like to YOU!!!

User avatar
brit01
TFM-45
Posts: 580
Joined: Sat Jul 12, 2008 3:29 pm
Location: Montevideo, Uruguay

Re: Improving the power supply of the TFM series.

Post by brit01 » Mon Aug 30, 2010 6:01 pm

so ultra-fast rectifiers, necessary in digital power supplies, aren't necessary here. What could help are ultra-fast switching diodes and maybe snubbers
Thanks for the info guys.
Yes I was aware of the triac system but not too much.

When I refer to the rectifiers I did mean the diodes that convert the 56v and 98 v AC rails to DC.
Those are the diodes you meant?

HEXfreds are very good in these situations and yes snubbers also but Freds are excellent just on they're own.

Also lytics are pretty horrible in any system. The large ones (I renewed a couple of years ago here, remember?), maybe could be bypassed with some small Russion PIO, 0.1 or 1 uf. Or maybe film but the PIO may soften those lytics up a bit.
TFM-55, CT-3, Klipsch F-1, Rotel RCD 930AX

User avatar
BillD
R.I.P. Friend
R.I.P. Friend
Posts: 7126
Joined: Wed Jun 06, 2007 8:29 pm
Location: The west's most mid-western town, Scottsdale, Arizona

Re: Improving the power supply of the TFM series.

Post by BillD » Mon Aug 30, 2010 7:27 pm

No, I was referring to the diodes that control rail switching. Snubbers are to control ringing.
It should sound like it isn't there!
There is a difference between hearing and listening...
Making life enjoyable through expensive electronics.
_________________
Carver: C-4000 & C-1 preamps, PSC-60 preamp/tuner, TX-11a tuner, M-400 (2), C-500, M-500, M-500t, M-500t Mk.II, A-500x, AL-III loudspeakers (2 pr.)
Sunfire:Theater Grand III processor, Ultimate Receiver, Cinema Grand Signature 400 ~ seven, True Subwoofer Mk. II, D-10 Subwoofer

angelod307
SILVER-7 TUBE AMP
Posts: 2654
Joined: Sat Jun 06, 2009 10:56 pm
Location: manning, sc

Re: Improving the power supply of the TFM series.

Post by angelod307 » Mon Aug 30, 2010 10:19 pm

kingman wrote:
BillD wrote:Sorry, forgot where I was. :D
Just wish I had that kind of knowledge Bill!!! =D> =D> =D>

x2
CarverFest: I just go for the arts and crafts

Post Reply

Return to “!!! MODS !!!”