Maybe not a ground-breaking example of rocket science, but at least something healthy, nice and simple: much more than what is usually presented here, and therefore more commendable.
Nothing wrong with good, simple cheap and effective recipes
Edit:
I seen hadn't the 1Meg resistor.... some reason to despair about the human race
Nothing wrong with good, simple cheap and effective recipes
Edit:
I seen hadn't the 1Meg resistor.... some reason to despair about the human race
Last edited:
And even better to ears if you put the 10 k pot not common with output but // on the 10 nF ? Not of a electrical view but with a subjetive SQ I mean !
filtering the tl 431 wit 2 000 uF low esr + 0.2 uF MKT ?
Jfet instead BC 327 ?
Tip 142/147 are said to be good, but some elswhere struggle more for a discrete darlington : "measurement vs quality thread" in the Lounge part !
Thank you for sharing, I'm looking for someting like that : simple and unexpensive 🙂 maid on a veroboard.
filtering the tl 431 wit 2 000 uF low esr + 0.2 uF MKT ?
Jfet instead BC 327 ?
Tip 142/147 are said to be good, but some elswhere struggle more for a discrete darlington : "measurement vs quality thread" in the Lounge part !
Thank you for sharing, I'm looking for someting like that : simple and unexpensive 🙂 maid on a veroboard.
Clever ideas, with just one caveat: they are going to ruin the circuit, but who cares
😀 Oups, the circuit will burn ?
Just a simple question please :
Is going for something simplier equal to a worst sound here :
Forgetting the design above for just a CCS feeding a TL431, this one with 2 K uF just for a local feed back, with the upper resistor staying a pot but not coupled at the output of the tip147 but at the TL431 cathode .
The TIP147 just having 0.2 uF at output or 0R2 for stability.
It's certainly a naive question but I saw this design on a book elswhere (local feedback only, designer says it's better on measure with the pot at the output of TIP147 but worst at ears!). Is the OP elegant simple design is better (is it not a global feedback ? ).
I will keep the most musical !
@ Dzius : could maybe interest you to swap the tip142/147 : discrete seems have wider bandwidth if I understood well : http://www.diyaudio.com/forums/lounge/200865-sound-quality-vs-measurements-1712.html#post4161417
http://www.diyaudio.com/forums/lounge/200865-sound-quality-vs-measurements-1712.html#post4161510
and :
http://www.diyaudio.com/forums/lounge/200865-sound-quality-vs-measurements-1719.html#post4167637
No, static and dynamic regulation performances will be degraded
It will have mediocre regulation, but if that's what you are trying to achieve, it is probably simpler and more effective to ditch the 431 completely, and connect the pot wiper directly to the base of the darlington
A supply is not a musical instrument, its role is to supply a voltage that is as constant and as clean as possible, under all circumstances
Some explanation how it works:
1) 2k resistor provides the TL431 bias current and the Ibe current of T1
2) 2.2k resistor limits the Ibe of T2 and final current of the regulator
3) 10k pot and 1k resistor form control voltage divider at the output.
4) 10nF capacitor is a kind of an error correction
In other words circuit is suitable only for bipolar transistors !
I'm not designer of that just found it on the internet and couldn't find it here.
Probably Elvee understand better how it works than me. I don't know which transistor would play better at the output . Neither I know why discrete darlington has better bandwith (strange to me).
1) 2k resistor provides the TL431 bias current and the Ibe current of T1
2) 2.2k resistor limits the Ibe of T2 and final current of the regulator
3) 10k pot and 1k resistor form control voltage divider at the output.
4) 10nF capacitor is a kind of an error correction
In other words circuit is suitable only for bipolar transistors !
I'm not designer of that just found it on the internet and couldn't find it here.
Probably Elvee understand better how it works than me. I don't know which transistor would play better at the output . Neither I know why discrete darlington has better bandwith (strange to me).
I can get it in full could you draw it ?
A possible improvement would be to connect the 10K control as a potentiometer rather than as a variable resistance (wiper to the 431). This would reduce the noise, especially when the pot is operated, and improve the thermal stability somewhat.
It would also be a good idea to place a resistor (a few K) in series with the wiper, to let the compensation cap act even when the voltage is at the minimum.
It would also be a good idea to place a resistor (a few K) in series with the wiper, to let the compensation cap act even when the voltage is at the minimum.
You mean for example than a global feedback which is more stable and has an equal "sounding quality" like the amp maid in the 70-80s'?
Hummm I believed the shunt was more for allow current than voltage stability... I checked and some find the pot of the shema I described less stable but better for sounding ! they call it more a stabiliser than a regulator ! the guys knowing the regulation to be worst but the listning result to be better like that ! Maybe not experienced diyers 😕
ok thanks for the inputs.
but I never tried both of them yet 🙄 (I mean with pot after the Darlington or on the cathode of the tl431 before the Darlington.).
Yes of course, give me some times: I described already this shematic on posts near the ones I linked. It is so simple and old than this very simple design was more for the discusion about design.
I come back soon as I don't know how to draw it but with a pen and a paper.
It will be more easy for the experienced designers & tech to talk about it !
Hi Dzius,
sorry for to be late : http://www.diyaudio.com/forums/powe...ed-back-emitter-follower-reg.html#post3887530
the shematic I talk about is in this post. Took from a french book by Francis Ibre while being very common. Of course this is not a regulator, more a stabilisator but it is said to give better listening results... the local feedback only on the tl431 is a will !
sorry for to be late : http://www.diyaudio.com/forums/powe...ed-back-emitter-follower-reg.html#post3887530
the shematic I talk about is in this post. Took from a french book by Francis Ibre while being very common. Of course this is not a regulator, more a stabilisator but it is said to give better listening results... the local feedback only on the tl431 is a will !
Looks like you suggest to use Tl431 as regular zener diode. One can find such source/emiter circuit in many Nelson Pass designs so it's nothing new. But if you insist it sounds better then no future worry about what to do.
Another source follower with Tl431 controlling voltage at the output is here:
Ultra Low Drop Linear Voltage Regulator | EEWeb Community
Another source follower with Tl431 controlling voltage at the output is here:
Ultra Low Drop Linear Voltage Regulator | EEWeb Community
few things :
Never say it was new 😕
I don't insist, I'm not technician, i just want to understand more : it is showed as I want to understand the difference, that's all !
Maybe instead employ such a bad word you have to try yourself to make your own idea, no ?😛 .
Just show you as you ask and to be talked as i'm a beginner and to understand the tradeoffs between a more quieter zener diodes vs a lower output impedance provided by such a bandgap... It seems in the links which show a design I copy than the regulation is forgived because the sound quality is better with a local feedback on the TL431 only ! Just want to understand why, not wanting to know it is better in regulation term if the pot output were after the tip142/147 but more wanting to know why the regulation performance was given up while this designer knew perfectly it is better after ! He tryed both !
Was used before Nelson Pass which is not the only good designer in the World, japoneses and french use such design for maybe longer years !
That's all.... and thanks for the link 🙂
Never say it was new 😕
I don't insist, I'm not technician, i just want to understand more : it is showed as I want to understand the difference, that's all !
Maybe instead employ such a bad word you have to try yourself to make your own idea, no ?😛 .
Just show you as you ask and to be talked as i'm a beginner and to understand the tradeoffs between a more quieter zener diodes vs a lower output impedance provided by such a bandgap... It seems in the links which show a design I copy than the regulation is forgived because the sound quality is better with a local feedback on the TL431 only ! Just want to understand why, not wanting to know it is better in regulation term if the pot output were after the tip142/147 but more wanting to know why the regulation performance was given up while this designer knew perfectly it is better after ! He tryed both !
Was used before Nelson Pass which is not the only good designer in the World, japoneses and french use such design for maybe longer years !
That's all.... and thanks for the link 🙂
Last edited:
What you mentioned is just regular zener controlled transistor voltage regulator (like here http://www.circuitstoday.com/wp-con...olled-Transistor-Series-Voltage-Regulator.jpg) but with TL431 instead of zener. (It's not 7 year old but close to 70, I guess. Used in series triode regs for example with glow tubes. Have a look here http://www.talkingelectronics.com/Download%20eBooks/Principles%20of%20electronics/CH-17.pdf - Fig 17.23, page 18) .
Following your way of thinking one can say that follower circuit without zener diode or TL431 is even better in audio. Good example you can find here: http://linearaudionet.solide-ict.nl/sites/linearaudio.net/file/Didden LA V3 PK lr.pdf - Fig 1.
Nonetheless circuit I enclosed at the very begining can be very useful for estimating the best working point for particular transistor working in SE A-class amplifiers similar to L'SIT amp for instance (or Burning Amp). Regulator is very stable since it control base current of power bipolar transistor in a much better way than simply zener controlled transistor voltage regulator. I mean TL431 can pass just a little current, but BC337 much more which in turn can give much more current at the regulator output also.
Your good sounding proposal http://www.diyaudio.com/forums/powe...ed-back-emitter-follower-reg.html#post3887530
is much weaker at this point.
Following your way of thinking one can say that follower circuit without zener diode or TL431 is even better in audio. Good example you can find here: http://linearaudionet.solide-ict.nl/sites/linearaudio.net/file/Didden LA V3 PK lr.pdf - Fig 1.
Nonetheless circuit I enclosed at the very begining can be very useful for estimating the best working point for particular transistor working in SE A-class amplifiers similar to L'SIT amp for instance (or Burning Amp). Regulator is very stable since it control base current of power bipolar transistor in a much better way than simply zener controlled transistor voltage regulator. I mean TL431 can pass just a little current, but BC337 much more which in turn can give much more current at the regulator output also.
Your good sounding proposal http://www.diyaudio.com/forums/powe...ed-back-emitter-follower-reg.html#post3887530
is much weaker at this point.
There's a catch to this circuit:
it requires a minimum current to work properly.
With the component values given in the OP it drops out of regulation below approx. 150mA.
Exact value depends on gain (hfe) of power transistor; the higher the gain, the higher the min. current ...
Reason is that TL431 needs min. voltage across K-A of 2.5v, which depends on voltage drop across 2.2k (bottom left), which depends on base current of power transistor, which depends on beta.
Therefore if load is disconnected or has idle mode, output v rises towards rail.
it requires a minimum current to work properly.
With the component values given in the OP it drops out of regulation below approx. 150mA.
Exact value depends on gain (hfe) of power transistor; the higher the gain, the higher the min. current ...
Reason is that TL431 needs min. voltage across K-A of 2.5v, which depends on voltage drop across 2.2k (bottom left), which depends on base current of power transistor, which depends on beta.
Therefore if load is disconnected or has idle mode, output v rises towards rail.
Clever observation: the 1MΩ B-E made me uncomfortable, but I didn't look further: this was a mistake.
There is clearly a problem there, although in practice it is probably not as bad as what can be inferred from the datasheet: generally, TL431's are able to pull their output terminal lower than 2.5V if the control input is higher, but that is undocumented behavior, and one should not rely on it.
The 1MΩ was probably a typo, and 1KΩ is more likely, but even with 1KΩ there is no guarantee the circuit will behave gracefully. 100Ω would be safer
Hi,
Assuming the last input :
What is, please, the minimal load cunsumption needed if the top 1 M ohm is swaped by 100 ohms ?
For a symetric voltage regulation does I swap the BC337 by a PNP as I have to do for the Darlington ?
My load is a mini diamond buffer with 6 discrete BC transistor (3 per rail) for each (L/R) channel than Pedja Rogic drawed some times ago... so I need an inexpensive +/- 15V to 16 V... certainly less than 100 mA but don't know exactly.
thanks in advance to say me if this shematic is good enough like it is with the resistor correction.
regards
Assuming the last input :
What is, please, the minimal load cunsumption needed if the top 1 M ohm is swaped by 100 ohms ?
For a symetric voltage regulation does I swap the BC337 by a PNP as I have to do for the Darlington ?
My load is a mini diamond buffer with 6 discrete BC transistor (3 per rail) for each (L/R) channel than Pedja Rogic drawed some times ago... so I need an inexpensive +/- 15V to 16 V... certainly less than 100 mA but don't know exactly.
thanks in advance to say me if this shematic is good enough like it is with the resistor correction.
regards
If the base of T1 must not go under 2.5V the current must not drop below 1mA.
The emitter has 2k2, gives 2.2V+0.6V=2.8V base voltage.
We have to make the 1M so small that 1mA is needed to get T2 conducting,then it should funtion.
With the BD244 we need 1mA to get 0.6V , lets say 560 Ω.
Mona
The emitter has 2k2, gives 2.2V+0.6V=2.8V base voltage.
We have to make the 1M so small that 1mA is needed to get T2 conducting,then it should funtion.
With the BD244 we need 1mA to get 0.6V , lets say 560 Ω.
Mona