My intent was to provide an alternative to running the whole thing on a regulated or CM supply since it is really the VAS (or TIS for those who prefer) that is sensitive to ripple. I see what you are saying and I'm sure it wouldn't be too difficult to add the ability to choose wether or not to supply a boosted rail for the front end. The layout is not finalized so I'll see what I can do about that.
Thanks for the feedback, this is the kind of thing that will help me make the board better for all.
Hi Jason,
it is only suggestion.
with dual layer PCB we can use any input pairs
(any BC or SA-SC) maybe good to go
I've done it with single layer PeeCeeBee
12 holes for input traniewith dual layer maybe we can reduce the holes
Regards
Hi Jason,
it is only suggestion.
with dual layer PCB we can use any input pairs
(any BC or SA-SC) maybe good to go
I've done it with single layer PeeCeeBee
with dual layer maybe we can reduce the holes
Regards
Hmm, I like the idea if it doesn't make things too busy. I will have a play at that and see if it can be done without making a righteous mess of things.
I fear that making the board too adaptable will make for confusion and possibly errors during construction. I also don't want to add extra PCB real estate if I don't have to. Has me thinking of narrowing the CCSs down to just one type even.
Some builders may use a standard power supply rather than a SMPS. In that case you might need to know the current rating of an unmarked transformer --- like the one(s) in your junk box or an old amp transformer; Here is a formula to determine the VA of an E-I type transformer -->
You measure the core center area of the transformer which the wire is wound around.... the center part of the "E". Say you measure it as 2.5 x 1.5 inch. Find the area and multiply by 5. Then take the square of this result and you have the transformer VA. This example would make it a 351VA transformer.
THx-RNMarsh
You measure the core center area of the transformer which the wire is wound around.... the center part of the "E". Say you measure it as 2.5 x 1.5 inch. Find the area and multiply by 5. Then take the square of this result and you have the transformer VA. This example would make it a 351VA transformer.
THx-RNMarsh
LED Vf drops with increased temp. , offsets it's associated transistor's pos.
coefficient. This effect is increased with low LED I(f).
The LED CCS generally has slight negative Coefficient.
2Q CCS has a definite positive coefficient ... best used in a circuit that
has a second stage with another CCS (like the "blameless") .
In this case they cancel out.
Built a bootstrapped typical LIN amp (2Q CCS/LTP) ... in the cold
morning, VAS current was low. In the hot tennessee day ...
it increased by 2+ mA!
The other amp with 2 CCS's (LTP and VAS) was within .1mA going from
15 -42C.
PS - In a climate controlled scenario , the 2Q will work with a stable
ambient. LED CCS's are best for a "party amp" (indoor/outdoor).
The 2Q also will provide 6+db more PSRR (a better CCS) , it just has to be cancelled
out by some other stage or another similar CCS in the design.
OS
Thanks for popping in, OS. I do appreciate your input.
I originally went with the 2Q CCS because I was always taught it was one of the better 'simple' CCSs. Then Sheldon demonstrated the easy conversion to the 1Q+LED which seemed like a bonus feature not even thought of. Due to some comments having been made I was contemplating removing the 2Q CCS option from the next iteration and just leaving the 1Q+LED for the time being. I'm not a fan of the self-biased JFET option myself, but it seems popular with the VSSA crowd and so I might try to gracefully see if the options can be offered together. That or maybe try to better compensate the 2Q thermally...
As far as the PSRR goes, that's the primary motivation to integrate a capacitance multiplier to feed the front end. If there's virtually no power supply ripple, then the native PSRR of the amplifier need not be high. That said, there is certainly nothing wrong with striving for decent PSRR in the amplifier itself. We know this type of circuit is somewhat sensitive to ripple, so why not remove as much as possible. I'm almost certainly now going to add pads to allow the integrated CM to be fed from the main PSU via a jumper, or a boosted PSU for those who want to provide a few extra volts of drive for maximum efficiency.
I originally went with the 2Q CCS because I was always taught it was one of the better 'simple' CCSs. Then Sheldon demonstrated the easy conversion to the 1Q+LED which seemed like a bonus feature not even thought of. Due to some comments having been made I was contemplating removing the 2Q CCS option from the next iteration and just leaving the 1Q+LED for the time being. I'm not a fan of the self-biased JFET option myself, but it seems popular with the VSSA crowd and so I might try to gracefully see if the options can be offered together. That or maybe try to better compensate the 2Q thermally...
As far as the PSRR goes, that's the primary motivation to integrate a capacitance multiplier to feed the front end. If there's virtually no power supply ripple, then the native PSRR of the amplifier need not be high. That said, there is certainly nothing wrong with striving for decent PSRR in the amplifier itself. We know this type of circuit is somewhat sensitive to ripple, so why not remove as much as possible. I'm almost certainly now going to add pads to allow the integrated CM to be fed from the main PSU via a jumper, or a boosted PSU for those who want to provide a few extra volts of drive for maximum efficiency.
You can offset the 2Q's positive curve with a diode in series at the
emitter of the 2'nd transistor (the one without the Re).
The Vf of the diode will cancel the Vbe of the 2Q.
try 2 diodes for more cancellation .. (below D4/6).
you may have to double Re to maintain the same Ic for your CCS.
PS - If you are going to add the multiplier , it ... plus the led CCS will beat the 2Q.
You also could go all out "anal" and combine a temp. compensated 2Q (or 3 pin TL431 CCS - below 2) for the "ultimate".
OS
Attachments
Last edited:
The peeceebee is a "toy".
Drive a 500W amp with a CFA IPS and you might need a little
better PSRR (very high current = massive ripple).
In dual mono mode (separate PS's) poor PSRR might not be as crucial.
Single supply , massive amps .... it is.
OS
Last edited:
If you like them now , run them "CCS'ed" or with the front end (input pair/VAS)
running off a regulated supply.
They will "floor you" with their fidelity.
OS
Terry,
The CCS should help improve rejection of PSU flotsam and that would be the main motivation to use it. The VSSA and its variants like a reasonably clean power supply since the design's inherent ability to reject garbage on the PSU rails is naturally a little low. I've heard some say adding the CCSs make the design sound a little 'clinical', but I can't say that for myself. I like them all so far.
The magic of the TO-3 PeeCeeBee must be contained in those classic cans of goodness. Actually a little added distortion in the simplest of designs might make it sound more pleasing, not sure. I have been listening to my test mules from the first order and haven't swapped the PeeCeeBee back in to tell you.
The CCS should help improve rejection of PSU flotsam and that would be the main motivation to use it. The VSSA and its variants like a reasonably clean power supply since the design's inherent ability to reject garbage on the PSU rails is naturally a little low. I've heard some say adding the CCSs make the design sound a little 'clinical', but I can't say that for myself. I like them all so far.
The magic of the TO-3 PeeCeeBee must be contained in those classic cans of goodness
. Actually a little added distortion in the simplest of designs might make it sound more pleasing, not sure. I have been listening to my test mules from the first order and haven't swapped the PeeCeeBee back in to tell you.
Terry,
The CCS should help improve rejection of PSU flotsam and that would be the main motivation to use it. The VSSA and its variants like a reasonably clean power supply since the design's inherent ability to reject garbage on the PSU rails is naturally a little low. I've heard some say adding the CCSs make the design sound a little 'clinical', but I can't say that for myself. I like them all so far.
The magic of the TO-3 PeeCeeBee must be contained in those classic cans of goodness
A thought, could it be reason for still4given can't differentiate better sound with the CSS equipped VSSA PCBs that his PSU is a good CM PSU (clean makes CSS less needed) and therefor CSS don't shine in sound stage. But if he used classic PSU setup with just capasitor or C/R/C, then the VSSA's with CSS will hit and be listenable better sound stage.
Hi Brytt,
You are probably right. I didn't know what benifit the CCS gives. If it is to "clean up" the PSU it may very well be that my Mr Evil/PMI CM is so clean the CCS is not needed. I probably need to start using some conventional PSU for initial testing so I can hear the difference. I'm just about out of transformers now and have been using a big transformer and a variac for all my testing and running that through a CM.
As for the VSSA's, I have Jason's TO-3 Peeeebee sitting right on top of PMI's version and I A/B them often. There is almost no discernible difference in sound and the TO-3 version is dead quiet and Pete's version has a very slight hum. I don't have Jason's latest design in a case yet but when I had it on the bench and A/B'ed it against the others it was basically identical sounding. These amps are really quite amazing.
Blessings, Terry
You are probably right. I didn't know what benifit the CCS gives. If it is to "clean up" the PSU it may very well be that my Mr Evil/PMI CM is so clean the CCS is not needed. I probably need to start using some conventional PSU for initial testing so I can hear the difference. I'm just about out of transformers now and have been using a big transformer and a variac for all my testing and running that through a CM.
As for the VSSA's, I have Jason's TO-3 Peeeebee sitting right on top of PMI's version and I A/B them often. There is almost no discernible difference in sound and the TO-3 version is dead quiet and Pete's version has a very slight hum. I don't have Jason's latest design in a case yet but when I had it on the bench and A/B'ed it against the others it was basically identical sounding. These amps are really quite amazing.
Blessings, Terry
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.