Greg Stewart x2
ppap64 x 2 V2
Peterma x 3 V2
kenlaumm x 2 V2
j.burtt x 2
crowlie x3 V2
ed linssen x2 V2
mravinsky x2 V2
palmito x2 V2
xaled x2 V2
noizas x4 V2
wlowes 1x V2
pchw 2x V2
badrisuper X 1
tagheuer x2
Jaffrie x2 V2
Luke X2 V2
lll X2 V2
unixdeveloper x2 V2
ppap64 x 2 V2
Peterma x 3 V2
kenlaumm x 2 V2
j.burtt x 2
crowlie x3 V2
ed linssen x2 V2
mravinsky x2 V2
palmito x2 V2
xaled x2 V2
noizas x4 V2
wlowes 1x V2
pchw 2x V2
badrisuper X 1
tagheuer x2
Jaffrie x2 V2
Luke X2 V2
lll X2 V2
unixdeveloper x2 V2
Greg Stewart x2
ppap64 x 2 V2
Peterma x 3 V2
kenlaumm x 2 V2
j.burtt x 2
crowlie x3 V2
ed linssen x2 V2
mravinsky x2 V2
palmito x2 V2
xaled x2 V2
noizas x4 V2
wlowes 1x V2
pchw 2x V2
badrisuper X 1
tagheuer x2
Jaffrie x2 V2
Luke X2 V2
lll X2 V2
unixdeveloper x2 V2
BDL x1 V2
ppap64 x 2 V2
Peterma x 3 V2
kenlaumm x 2 V2
j.burtt x 2
crowlie x3 V2
ed linssen x2 V2
mravinsky x2 V2
palmito x2 V2
xaled x2 V2
noizas x4 V2
wlowes 1x V2
pchw 2x V2
badrisuper X 1
tagheuer x2
Jaffrie x2 V2
Luke X2 V2
lll X2 V2
unixdeveloper x2 V2
BDL x1 V2
Changed to V2
Greg Stewart x2
ppap64 x 2 V2
Peterma x 3 V2
kenlaumm x 2 V2
j.burtt x 2
crowlie x3 V2
ed linssen x2 V2
mravinsky x2 V2
palmito x2 V2
xaled x2 V2
noizas x4 V2
wlowes 1x V2
pchw 2x V2
badrisuper X 1
tagheuer x2 V2
Jaffrie x2 V2
Luke X2 V2
lll X2 V2
unixdeveloper x2 V2
BDL x1 V2
Greg Stewart x2
ppap64 x 2 V2
Peterma x 3 V2
kenlaumm x 2 V2
j.burtt x 2
crowlie x3 V2
ed linssen x2 V2
mravinsky x2 V2
palmito x2 V2
xaled x2 V2
noizas x4 V2
wlowes 1x V2
pchw 2x V2
badrisuper X 1
tagheuer x2 V2
Jaffrie x2 V2
Luke X2 V2
lll X2 V2
unixdeveloper x2 V2
BDL x1 V2
Greg Stewart x2
ppap64 x 2 V2
Peterma x 3 V2
kenlaumm x 2 V2
j.burtt x 2
crowlie x3 V2
ed linssen x2 V2
mravinsky x2 V2
palmito x2 V2
xaled x2 V2
noizas x4 V2
wlowes 1x V2
pchw 2x V2
badrisuper X 1
tagheuer x2 V2
Jaffrie x2 V2
Luke X2 V2
lll X2 V2
unixdeveloper x2 V2
BDL x1 V2
zany x2 V2
ppap64 x 2 V2
Peterma x 3 V2
kenlaumm x 2 V2
j.burtt x 2
crowlie x3 V2
ed linssen x2 V2
mravinsky x2 V2
palmito x2 V2
xaled x2 V2
noizas x4 V2
wlowes 1x V2
pchw 2x V2
badrisuper X 1
tagheuer x2 V2
Jaffrie x2 V2
Luke X2 V2
lll X2 V2
unixdeveloper x2 V2
BDL x1 V2
zany x2 V2
Just updated mine to V2.
Greg Stewart x2
ppap64 x 2 V2
Peterma x 3 V2
kenlaumm x 2 V2
j.burtt x 2 V2
crowlie x3 V2
ed linssen x2 V2
mravinsky x2 V2
palmito x2 V2
xaled x2 V2
noizas x4 V2
wlowes 1x V2
pchw 2x V2
badrisuper X 1
tagheuer x2
Jaffrie x2 V2
Luke X2 V2
lll X2 V2
Report Post
Greg Stewart x2
ppap64 x 2 V2
Peterma x 3 V2
kenlaumm x 2 V2
j.burtt x 2 V2
crowlie x3 V2
ed linssen x2 V2
mravinsky x2 V2
palmito x2 V2
xaled x2 V2
noizas x4 V2
wlowes 1x V2
pchw 2x V2
badrisuper X 1
tagheuer x2
Jaffrie x2 V2
Luke X2 V2
lll X2 V2
Report Post
Greg Stewart x2
ppap64 x 2 V2
Peterma x 3 V2
kenlaumm x 2 V2
j.burtt x 2 v2
crowlie x3 V2
ed linssen x2 V2
mravinsky x2 V2
palmito x2 V2
xaled x2 V2
noizas x4 V2
wlowes 1x V2
pchw 2x V2
badrisuper X 1
tagheuer x2 V2
Jaffrie x2 V2
Luke X2 V2
lll X2 V2
unixdeveloper x2 V2
BDL x1 V2
zany x2 V2
ppap64 x 2 V2
Peterma x 3 V2
kenlaumm x 2 V2
j.burtt x 2 v2
crowlie x3 V2
ed linssen x2 V2
mravinsky x2 V2
palmito x2 V2
xaled x2 V2
noizas x4 V2
wlowes 1x V2
pchw 2x V2
badrisuper X 1
tagheuer x2 V2
Jaffrie x2 V2
Luke X2 V2
lll X2 V2
unixdeveloper x2 V2
BDL x1 V2
zany x2 V2
Greg Stewart x2
ppap64 x 2 V2
Peterma x 3 V2
kenlaumm x 2 V2
j.burtt x 2 v2
crowlie x3 V2
ed linssen x2 V2
mravinsky x2 V2
palmito x2 V2
xaled x2 V2
noizas x4 V2
wlowes 1x V2
pchw 2x V2
badrisuper X 1
tagheuer x2 V2
Jaffrie x2 V2
Luke X2 V2
lll X2 V2
unixdeveloper x2 V2
BDL x1 V2
zany x2 V2
flyboi x2 V2
ppap64 x 2 V2
Peterma x 3 V2
kenlaumm x 2 V2
j.burtt x 2 v2
crowlie x3 V2
ed linssen x2 V2
mravinsky x2 V2
palmito x2 V2
xaled x2 V2
noizas x4 V2
wlowes 1x V2
pchw 2x V2
badrisuper X 1
tagheuer x2 V2
Jaffrie x2 V2
Luke X2 V2
lll X2 V2
unixdeveloper x2 V2
BDL x1 V2
zany x2 V2
flyboi x2 V2
PSU update.
Design has been modified to include PNP pass transistors. The 431 in this configuration is used as a shunt-controlling amplifier driving an emitter follower. Output impedance can be much lower in the audible frequency range than with a solitary TL431. It must be noted that there are many manufacturers of 431s with varying specs, the low noise ones are from On Semi and Fairchild.
The three shunt supplies are connected in series acting as a type of "zener string". The required current to the whole circuit is ~150mA, depending on how much heat we will be able to dissipate from the -15V PNP. (1W in this configuration).
The part of the schematics (attached) contained by the red marking is what i'm proposing to put on the PCB. Only one 24V pre-filtered and pre-regulated supply will be required. (~4V headroom for the CCS in the negative line).
Simulated ripple rejection from the source is -220dBV. Noise from the shunt circuit will likely halve this value.
Overall, considering the simplicity, performance, and cost - This will be a very good solution.
Design has been modified to include PNP pass transistors. The 431 in this configuration is used as a shunt-controlling amplifier driving an emitter follower. Output impedance can be much lower in the audible frequency range than with a solitary TL431. It must be noted that there are many manufacturers of 431s with varying specs, the low noise ones are from On Semi and Fairchild.
The three shunt supplies are connected in series acting as a type of "zener string". The required current to the whole circuit is ~150mA, depending on how much heat we will be able to dissipate from the -15V PNP. (1W in this configuration).
The part of the schematics (attached) contained by the red marking is what i'm proposing to put on the PCB. Only one 24V pre-filtered and pre-regulated supply will be required. (~4V headroom for the CCS in the negative line).
Simulated ripple rejection from the source is -220dBV. Noise from the shunt circuit will likely halve this value.
Overall, considering the simplicity, performance, and cost - This will be a very good solution.
Attachments
Hi Ryan, I'm good with your decision. I recall seeing this at: Simple Voltage Regulators Part 1: Noise - [English]
Its exceptionally low noise according to this, but they get 1ohm impedance which is much different to your simulation. Iguess it comes down to which parameter is more important, impedance or noise.
Either eway, Im sure its going to sound good and besides, we can bypass and try different supplies.
Its exceptionally low noise according to this, but they get 1ohm impedance which is much different to your simulation. Iguess it comes down to which parameter is more important, impedance or noise.
Either eway, Im sure its going to sound good and besides, we can bypass and try different supplies.
Your terrible Luke. lol. Think of the efford that Ryan has put in.
I would rather have low noise then low z any day. What is
important is reg stability underload at audio frequency with
any output cap. That way you can tailor whatever sound
to your liking with different brand of caps & uf as well
I would rather have low noise then low z any day. What is
important is reg stability underload at audio frequency with
any output cap. That way you can tailor whatever sound
to your liking with different brand of caps & uf as well
Really, I must be a poor communicator🙁. I am not sure which is more desirable, it was just a comment based on what I read. As I said days ago Ryan will never please everybody and there is an option to bypass and add your own reg. I intend to use the onboard reg🙂
As they say, the pen is mightier than the sword. Joke aside we all are caught up in dwelling & focusing on a perticular spot & trying to use & make the best etc etc there is. At the end of
the day it might not sound like what you had expected cause through the use of the best of etc
etc parts, one might have schewed & off balance the sound of whatever you are building.
Cheers
the day it might not sound like what you had expected cause through the use of the best of etc
etc parts, one might have schewed & off balance the sound of whatever you are building.
Cheers
Luke, thanks for posting the link to the TNT site. Some years ago I read that article and ended up using their Zener + Emitter Follower circuit on the +5v rail of my tda1541a. Always sounded pretty good and parts cost was about $5. Compared to all the shunts being built it felt like a good deal. Mine used thru the hole parts. With SMT the thing could be right on the pin.
I think having a power supply close to the pins is more important than the exact topology, rightly or wrongly, I like the idea. I also think low impeadance is more importance than noise. You have to look at the value of noise we are talking here, its so low I think its irrelevant. Engineering is about compromises, Ryan is doing the engineering and Im OK with that. I am going to use what he provides, and maybe try a couple of spare shunts I have as well by bypassing the local supplies.
I seem be spending all my time on HP amps at the moment, need to move onto my dac🙂
I just got a Sapphire 3.0 HP amp up and running and it uses a similar reg. Its a very good amp and I don't know that a fancier supply would make it any better. I think we sometimes obsess too much here on DIYA🙂
SMT is becoming a reality. More and more transistors are not being manufactured. If you look at the headcase forum they are creating new boards for what parts are available. They are now going to SMD transisitors I believe. Soon we will ahve no choice I think.
Last edited:
Seems to me that the user will be locked into a single winding and rectifier for all three supplies if they intend to use this PCB.
No galvanic isolation and 150mADC charging pulse into a 5kuF cap.. have fun!
Sun
- Status
- Not open for further replies.