12V DC Regulated Power Supply

tl431

Hi.

The TL431 is an excelent shunt regulator, very low noise. by the looks of the series resistors and the voltages you've measured total current draw from the 12V is about 130mA ish. a 0.2A supply at 12V should do fine.

An important issue is keeping the rectifier noise low, at these currents use 2 1n4148 signal diodes in parallel for each arm of the rectifier (a series resistor of a few ohms will make sure they share the current.) A small toroidal transformer, CRCRC smoothing will remove any remaining noise and then a cheapy 7812 regulator will do the rest.

If you want i'll sketch a circuit and email it to you.

Chris
 
Chris, thanks for the info. If you could sketch a circuit that would be great. My e-mail is:

michael AT luacheia DOT com

What if I wanted an "overdimensioned" PSU, say capable of 1A supply at 12V? Just to keep my audiophilia nervosa from flaring up you understand ;)

Michael.
 
rating

It's best to keep the PSU as small as possible but with sufficient overhead for the job.

The larger the PSU the bigger the radiated fields, the less rejection for noise from the mains (larger transformers have better regulation and therefore filter less mess). Larger rectifiers are slower and give more switching noise.....

A new preamp i've just finished has a 3 watt supply and the circuit uses 2W, sounds great, residual output noise is less than 10uV in 100kHz bandwidth, nothing nasty coming from the PSU there!!!

Power amps are a different matter though.

I'll try and sketch something before tonight.

Chris
 
you could go exotic,the switching can be automatic or manual,im using something simaler on my phono stage,i just happened to have a couple of 12volt 7amp fire alarm batterys looking ineed of a use when i built it,but a 1A sealed lead acid can be found for less than a fiver
 

Attachments

  • psublock.jpg
    psublock.jpg
    11 KB · Views: 436
Tube_Dude said:
For the inquiring minds, here is a opinion of John Curl (Parasound designer ) about both types of power transformers...


Hi Tube Dude.

John is a nice bloke, but we must not take what he says out of context. He's talking about making comercial equipment, the transformers are designed specifically for audio use, I use both EI and toroids in my comercial designs depending on application for the same reasons. Unfortuantely the transformers most oftenly available to the hobbyist are not made to the same high standards they are built down to a price. Cheap EI transformers often run at a very high flux density pushing the core to its limits, the laminations are often not very well stacked and even sometimes not interleaved at all but just welded, these cheap manufacturing tecniques lead to external magnetic fields being many times more than that of a EI transformer I (and probably john) would let anywhere the insides of my equipment. Some aren't even impregnated leading to mechanical vibrations/rattles.

For the amature building at home the amount of RFI break through on a small torroid is a lot easier to filter out than trying to screen the excessive hum field from a cheap EI, MUmetal is a lot more expensive than a small choke and capacitor:)

When I used to make amplifiers as a hobby i always designed my own transformers and used my local coil winder to make them for me (I was lucky the industial electronics i was designing at the time had a lot of wound components which i sourced from him so he was always willing to help me with a home job). Now I'm building amps commercially live is a lot easier, when placing an order for 100 transformers you have a lot more leaverage on the quality you will get, so long as you don't try to push the price too low.

Chris
 
Tube_Dude said:
Hi Michael

Your Dac is something like this...

http://jwg.student.utwente.nl/fedde/images/nonoz3.gif
Yes, it's a similar circuit. That circuit only has 2 regs though, one for the DAC and one for the CS8412. Mine has 2 regs for the CS8412, one for the digital part (VD+) and one for the analogue (VA+) and then 1 reg for the DAC so 3 regs in total.

Michael.
 
And the regs are shunt rather than series.

Now I'm wondering what's analog about the CS8412, or does it have some form of onboard signal isolation between the digital in and digital out? Which the circuit referenced above doesn't exploit? Time to check out a datasheet.

Paul
 
Chris Jennings said:
An important issue is keeping the rectifier noise low, at these currents use 2 1n4148 signal diodes in parallel for each arm of the rectifier (a series resistor of a few ohms will make sure they share the current.)

Yes ,I agree that we must keep the regulator noise low, but using a 1N4148 is in my view ,pushing the luck to far , because at the moment of power on when the electrolytic filter capacitor is discharged the diode see a virtual short . And as you know the ratings of the 1N4148 are for a small signal aplication.

[[ The larger the PSU the bigger the radiated fields, the less rejection for noise from the mains (larger transformers have better regulation and therefore filter less mess). ]]

I can not agree with you, in the less rejection from the mains noise in a big transformer , because with a big iron the high frequencies are more attenuated by the higher inductance.
Have you wandered why the pulse transformers for high speed
transmission (as SPDIF interfaces) are so small...

[[ I use both EI and toroids in my comercial designs depending on application for the same reasons. Unfortuantely the transformers most oftenly available to the hobbyist are not made to the same high standards they are built down to a price. Cheap EI transformers often run at a very high flux density pushing the core to its limits, the laminations are often not very well stacked and even sometimes not interleaved at all but just welded, these cheap manufacturing tecniques lead to external magnetic fields being many times more than that of a EI transformer I (and probably john) would let anywhere the insides of my equipment. ]]

Have you read the thread from the beginning?
We are talking about a remote power suplly and the EMI interference is not a issue ,but only the rejection of the mains borne interference ,that is proportional to the capacity between primary and secondary .
About the transformers for the professional market and for the hobbyist , seems to me ,more as a "buy ready made units because we use better parts" warning.

For the general hobbyist...if they want a "superior" trade mark transformer, take one from a old CD or Preamp, that we can buy for peanuts in a repairing shop...

Cheers :MILD:
 
Last edited by a moderator:
Hi TD.

1n4148 diodes have as good a short pulse rating, as do other diodes, also i prefer the first cap to be small and then use RC th smooth the ripple this improve PF and reduces current spikes during cap charge.

Th inductance that is important in removing HF in transformers is leakage inductance this is larger in small transformers, also the capacitive coupling is less in small transformers than large ones.

EMI interference is always an issue. Measure you equipment in a properly screen room and you will find out how much.

My point was that the toroids available to the hobbyist are far better than the EIs, since the saturation point is far steep on a toroid manufactures tend to need to leave more margin.

Chris:)
 
michaelab said:
What does a shunt reg do differently to a series one and which is better?

The series use more feedback and the output impedance is lower.
I prefer the series regulators , because I am, a high feedback proponent ...but be prepared for contrary opinions from the no-feedback brigade

By the way ,the amp that you have liked to ear at my house...use tons of feedback... :MILD:
 
Now I'm wondering what's analog about the CS8412 It's the PLL doing the clock recovery. Noise on this pin gets rolled directly into the jitter of the recovered clock; certainly quite small alterations here make for quite audible changes in the output.

As to TL431s - yes they're shunt regs with (1) quite a high output impedance - about 1 ohm, where the LM317 does at least 100x better (10mOhm up to c. 2Khz ish, rising gently above) and (2) wideband noise output, with a fairly 'white' characteristic. Personally I wouldn't use 'em in new designs, but I could imagine that this second characteristic matches very well with the 8412's PSU susceptibilities to produce a pleasing result.

Micheal - check out the TNT-Audio site; Werner Ogiers has just finished a short run of articles on basic voltage regulatoin you might find interesting.
 
Last edited by a moderator:
felix, just read the TNT articles you refer to. Are you saying that you wouldn't use a TL431 in new designs? The measurements Werner Ogiers made seem to suggest that the TL431 is overall significantly better (less noisy) than the LM317 if used properly.

TD - I know that your amp is a high feedback design, however, I'd prefer to use the regulator that has a better measured performance rather than the one that uses the theoretically better system ;)

Michael.
 
Tube_Dude said:
The series use more feedback and the output impedance is lower.
I prefer the series regulators , because I am, a high feedback proponent ...but be prepared for contrary opinions from the no-feedback brigade


The LM317L output impedance is lower at 100 hz, but the TL431 still has low op imp at 100kHz. The LM317L starts well with 0.01 ohm at 10Hz, by 100Hz its up to 0.1 ohm (just half of that of the 431). If you look at figures above 500Hz the story is much different; at 100kHz the 431 is still near 0.2 ohms whereas the 317L is nearly 20ohms !! For most of the audio band the 431 wins on output impedance.

Noise should also be far better with the 431 since its reference is driven from the already regulated output supply.

The TL431 has 55dB of gain in its error amp, how much feed back do you require before it stops being a low feedback design.

Chris:)
 
Hi.

stop the willy waving!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

I was comparing the 100mA TL431 with the 100mA LM317L, without adjustment this has 20 ohms, the data sheet you linked to shows 2 ohms without ajustment. If i buffer the 431 to be a 1 amp device the output impedance will fall to well below that of the 1A LM117.

I only came here to offer a professionals advice, but since you know far more about electronics than me i'll leave you to it. I don't need your praise i get more than enough of that from reviewers and customers.

TTFN

Chris
 
Chris Jennings said:
I was comparing the 100mA TL431 with the 100mA LM317L, without adjustment this has 20 ohms, the data sheet you linked to shows 2 ohms without ajustment. If i buffer the 431 to be a 1 amp device the output impedance will fall to well below that of the 1A LM117.


But why someone will use a LM 317 without the adjustment capacitor?


[[ I only came here to offer a professionals advice]]

An advice can be correct or not ,but it don't become correct , because it is a "professional" advice.


[[ but since you know far more about electronics than me i'll leave you to it.]]

I don't know if I am better or not than you in electronics...and frankly I even don't care.

[[ I don't need your praise i get more than enough of that from reviewers and customers.]]

Good for you...but I am only in Audio by passion and invoking professional knowledge as a argument ,in a technical conversation is not very fair...

Cheers :MILD:
 
Guys, let's not get into an argument about this please! Chris, thanks a lot for your input (and the PSU circuit diag.). TD, maybe things got lost in translation somewhere but there's no need to be so adverserial about things :) .

Michael.
 

Latest posts

Back
Top