The current through BJTs will be higher - to keep it at the same level you have to increase the value of the current supplying resistor (look at the sch. in post #1) or trim the CCSs.
It's a matter of simple calculus. You have, say, 9mA JFETs and you degenerate each of them with Rs of about 10R (20R pot) so they will run at Id of about 7mA. Also, we want about 7mA through BJTs. So, we need about 14mA through each current supplying resistor (or CCS, whatever you deside to use). Now we need to decide how much voltage swing we want at the output. Nice value is 10V_peak. Let's allow some overhead so our Vce=Vds=15V. It means that we will make the voltage divider at the bases of BJTs to give us about 16V at those bases and we have 17V difference to rails (+/-33V). Now we use the Ohm's Law to calculate the value of current supplying resistor: 17V divided by 14mA gives us the 1k2 value. Play with it to find what you like - just mind the Pd_max. If you come to result less than 1k, you better lower the PS voltage and use the CCSs.
Last edited:
PS Q&A
Juma,
to use the phase splitter as you tried it out with a good result (post #271): can I still use the same regulator configuration (maybe recalculate a few values) if I start with a lower voltage on trafo secondaries to get down to those 15Vdc (what Vdrop across the reg should I count with?).
p.s. I do not need the other PS half for +-24V since i will not be using the buffers, but I am stuck with that regulator since it is already on the pcb, so it makes sense for me to use it. else I am not sure of the end result re: Zoutwise etc. if i stay with +-33V rails. care to drop a hint re: rescaling the PS?
Thx!
Juma,
to use the phase splitter as you tried it out with a good result (post #271): can I still use the same regulator configuration (maybe recalculate a few values) if I start with a lower voltage on trafo secondaries to get down to those 15Vdc (what Vdrop across the reg should I count with?).
p.s. I do not need the other PS half for +-24V since i will not be using the buffers, but I am stuck with that regulator since it is already on the pcb, so it makes sense for me to use it. else I am not sure of the end result re: Zoutwise etc. if i stay with +-33V rails. care to drop a hint re: rescaling the PS?
Thx!
Making a +/-15V out of +/-33V would be impractical (too much excess voltage/power to burn) and needlessly complex. I made the PSU from post #66 that way not because it's the best way to do it but because I had such transformer at hand. You should use small (10VA) 2x15V AC transformer to get needed voltage in a simple way (straightforward rectifier + cap multiplier will do great) and save the PCB you have for the time when you need the high drive preamp.
For calculating V_drop in regulated PSU use the Ohm's law - it will never let you down (I wrote all important current and voltage values in previous posts).
If you insist on using that PCB, use the 2x9V_AC transformer and omit the parts you don't need (+/-24V).
For calculating V_drop in regulated PSU use the Ohm's law - it will never let you down (I wrote all important current and voltage values in previous posts).
If you insist on using that PCB, use the 2x9V_AC transformer and omit the parts you don't need (+/-24V).
Or maybe we can use on board +/-24V instead of +/-33V (that will require just two jumpers)? I hope that existing PS could safely drive phase splitter stage together with buffer stage.
In this case R10, R36 for positive and R13, R38 for negative supply (see schematics here) can be used for additional drop from 24 to 15V what is not so demanding as from 33 to 15V. Make sense?
In this case R10, R36 for positive and R13, R38 for negative supply (see schematics here) can be used for additional drop from 24 to 15V what is not so demanding as from 33 to 15V. Make sense?
Thanks guys for brainstorming. It really helps out. I set out for the purchased board trying to save myself time on this project (currently scheduled for Fall; cannot do every project on my own from scratch or it takes forever!), so I thought for the phase splitters for future bass/sub bridged amps I would go with Prasimix pcbs.
however, first: I have matched 170/74s in BL grade only (and even if I buy from the diyaudio store LSK/Js in C grade there is no guarantee they will not be lower than 15mA), second: I do not need the high pre gain and the hard driving PS. So now it came down to either using a workaround with the existing board, or saving those and doing my own board (phase splitters plus a Salas regulator I guess).
p.s. I feel like
since I just started routing my own pcb for 4 channels of Aleph Pv1.7 to drive F5/ACA combo in a biamped system (a separate project). I really wish I had waited for these Juma preamps since I spent a fortune on those Wima caps, and now I got a 4x10 XO where I can adjust individual gains in the digital domain. The only thing AP still has working for it is balanced/unbalanced combos for its versatility. I would have dumped the project and sell the caps but I already have two transformers at 60Vac; 
)
>anyways, thanks a lot, I will figure out something. probably a workaround with Prasimix board.
however, first: I have matched 170/74s in BL grade only (and even if I buy from the diyaudio store LSK/Js in C grade there is no guarantee they will not be lower than 15mA), second: I do not need the high pre gain and the hard driving PS. So now it came down to either using a workaround with the existing board, or saving those and doing my own board (phase splitters plus a Salas regulator I guess).
p.s. I feel like
since I just started routing my own pcb for 4 channels of Aleph Pv1.7 to drive F5/ACA combo in a biamped system (a separate project). I really wish I had waited for these Juma preamps since I spent a fortune on those Wima caps, and now I got a 4x10 XO where I can adjust individual gains in the digital domain. The only thing AP still has working for it is balanced/unbalanced combos for its versatility. I would have dumped the project and sell the caps but I already have two transformers at 60Vac; )>anyways, thanks a lot, I will figure out something. probably a workaround with Prasimix board.
Juma,
I am afraid I have one more:
(still looking at different ways to fully utilize the board): if I use the buffers only for the output with the higher output impedance (i.e. Out- output on each phase splitter), will my phase relationship to Out+ be maintained, i.e. the buffer is non-inverting, right? Thanks a lot/. S.
I am afraid I have one more:
(still looking at different ways to fully utilize the board): if I use the buffers only for the output with the higher output impedance (i.e. Out- output on each phase splitter), will my phase relationship to Out+ be maintained, i.e. the buffer is non-inverting, right? Thanks a lot/. S.
Indra1. The title of your circuit may easily be LSK Preamp meets diyF4. But your diyF4 is truly bare bones; stripped off the design circuits which manage its bias,and coupling. The generalized picture of your schematic is like that of a classical power amp; meaning [diff amp followed by LVA or last voltage amplifier, followed by Class A driver,and lastlyt output stage:
- The folded cascodes [like LVA] develop bias for the output Mosfets via R23 and R24; or by using instead the classical shunt regulator which is normally affixed to the heat sink.
- May use a Class A made of complementary bjts which have a low Zo to drive the capacitance of the Mosfets.
- And thats it
http://www.moxtone.com/Uskok pretpojacalo.html
This circuit looks like LSK preamp with servo .
Sent from my iPhone using Tapatalk
This circuit looks like LSK preamp with servo .
Sent from my iPhone using Tapatalk
The original link was working but the site owner appears to have closed it to DIYaudio linking! Remove all after .com and retry if you need to see it.
I must say that it is most likely coincidence but I have had a lot of trouble with my mac since opening the original link shortly after it was posted. It seems that the site owner has a very bad opinion of this forum - whether he is justified or not is not for me to judge.
I must say that it is most likely coincidence but I have had a lot of trouble with my mac since opening the original link shortly after it was posted. It seems that the site owner has a very bad opinion of this forum - whether he is justified or not is not for me to judge.
I believe there is a bit of contention about the nature of who was first to develop the circuit. I just go with the old saying, "there is nothing new under the sun". Turth is, I wonder if Juma's version of Nelson's version of Curl's version of Uskok's version doesn't sound exactly the same...if not better