F5 power amplifier

[rdk845]

Originally posted by rdk845
I'm having problem with my F5 having bias currents. Now I have taken out the channels from heat sink and slowsly increased the P1 and P2. But I don't see the voltage more than ~100mA. The voltage across the P2 never goes higher than ~3.2V even when I go the way up in N-channel part, while across P1 I can get higher than 4.2V. In both channels I can't get more than 0.1V across R11 and R12. .....

Originally posted by Nelson Pass Sounds like you need to increase the values of R3 and R4 to develop enough voltage. Try 3.3K or so.

I changed R3 and R4 to 4.4K, but same thing. The V across P2 won't go higher than 3.2V. As the voltage go above 3V, V over P2 goes up very slow, and V over P1 goes up even though I don’t change P1! With V across P1 above 5V, V across R11 and 12 goes up fast, but DC over Out and Ground stays at ~ 1.4V. Any idea what I can do to find what’s wrong? It appears the orientations of transistors are identical as those in Juno’s photograph…

[Nelson Pass]

Take the bipolar current limiters out and see if you get a difference,
if not then take R3 and R4 out altogether and report back.

Watch your current!

[danny_66]

Thanks Choky for the choke clarification.
So I need some real chokes in other words: big and heavy.
I've read some choke threads and maybe I'll roll them my own, with the help of Praxis(nice software) I can measure the inductance.

Now for the PS I have eight 22000uF caps.
Should I go for dual mono PS with two transformers and a CRC with 4x22000uF per channel
or a single transformer and a CRC or CRCRC with 8x22000uF for both channels ?

I did some simulations in Duncan amps PSU designer:
dual mono CRC 2x22000uf per rail: 360mV ripple
single PS CRC 4x22000uf per rail: 120mV ripple
single PS CRCRC 4x22000uf per rail: 70mV ripple

Better regulation or better channel separation ?
Or is the channel separation with a single PS already good enough?

Regards,

Danny

[AndrewT]

what about
rC -[RC ch1
. . -[RC ch2

this way you save on transformer and rectifier and share the first rC between the two channels.
Then some isolation effect using separate RC for each channel.
The final Cs use your 8pieces of 22mF to make +-44mF per channel.
The first C requires very high ripple capability. This is easier to achieve cheaply by paralleling caps maybe +-4*10mF.

Would a diode in each channel feed to hold up the voltage if one rail is very heavily loaded?

[Zen Mod]

Originally posted by danny_66
Thanks Choky for the choke clarification.
So I need some real chokes in other words: big and heavy.
I've read some choke threads and maybe I'll roll them my own, with the help of Praxis(nice software) I can measure the inductance.

Now for the PS I have eight 22000uF caps.
Should I go for dual mono PS with two transformers and a CRC with 4x22000uF per channel
or a single transformer and a CRC or CRCRC with 8x22000uF for both channels ?

I did some simulations in Duncan amps PSU designer:
dual mono CRC 2x22000uf per rail: 360mV ripple
single PS CRC 4x22000uf per rail: 120mV ripple
single PS CRCRC 4x22000uf per rail: 70mV ripple

Better regulation or better channel separation ?
Or is the channel separation with a single PS already good enough?

Regards,

Danny

try to find in Babelfish thread my recipe for 5A chokes , ~8mH at EI cores ;

when you find that , you can decide are you in mood for just 2 or 4 of these , then you'll know more about preferred supply configuration

[Phil Townsend]

Hey Danny... When you ran the sim with Duncan Amp, what inductance and dcr did you use.
I have often wondered why people never use big iron chokes in Solid state amps. For my hollow stuff I always use chokes in the PS. Big iron small caps... seems to sound better.

Phil
Snowing big fluffy flakes in Santa Fe!

[danny_66]

Hello Andrew,
That's a good idea, I did that in my Aikido preamp but with chokes:
I used one transfo, rectifier and CLC for both channels and after that I split into two LC, one for each channel.

Hello Choky,
found your recipe :
EI72 package
height of package ~30mm
entire window filled with 1,25 enameled wire
air gap ~ 0,5 mm
and also a very useful guide on:
http://geek.scorpiorising.ca/windingchokes.html
Now I need to find some EI cores...

So putting the two recipes together I get for my F5:
CLC - RC (channel 1)
. . . - RC (channel 2)
With 2mH chokes and 8x22000uf caps I get 22mV ripple

This way I have very good regulation and better separation without using too much iron, nice

Thanks,

Danny

[danny_66]

Hello Phil,
I simulated a choke of 2mH and 0.1ohm resistance, in the real world not too big in size and weight.
Grtz,
Danny

[Phil Townsend]

I built the low DCR PS that John S. came up with...
I am currently using it in my 45 parafed amp. But because the 5AR4 rectifier tube has a hi resistance I swapped it out with SS diodes. It worked well but the ss didn't make me happy. so back to the 5AR4/GX37. Mo better! The amp sounded deeper, faster and make music like gangbusters! I'll see if I can find the schematic... I lose stuff all the time. The long and the short of it was... very small inductors. LCLCLC
Hammond 159ZC 60mH @.7 Ohms
50 mike cap (solen) No stinkin electrolytics
Triad C47U 300mH @3 0hms
50 mike solen
Hammond 159ZA 300mH @6 ohms
50 mike solen
Hammond 159ZA 300mH @6 ohms
50 mike Solen
But this was for 350 volts at 120ma with 40mV of ripple
I bypassed the last solen with a 5 mike Mundorf
I'm not sure how to do something like this the very high current on the F5 nor do I have a clue if it is even a good idea?

[Zen Mod]

Originally posted by Phil Townsend
I built the low DCR......

play with Duncan's PSUD ;

use as much L you can in CLC ( or more ) filter ;

but - for much current you really need plenty mF