Double shunt reg PS

[387cbrb]

hi gang!

i'm new here so be nice!
i'm trying to design a good PS for TDA1541A
i'm a bit unsure of the use of series regs when audio
currents are involved. is there any way around the
standard LM317(337) -> TL431 combo?

how about a double shunt reg TL431->TL431?
i attach a pic of what i have in mind.
feel free to laugh at it
(sorry for the quality...)

cheers!

[387cbrb]

i forgot to ask: is there any optimal operating current for TL431 or anything 1-100mA woul do equally well? datasheets quote data at 10mA...

cheers!

[JonHarrison]

Hi,

Have you thought about using a capacitance multiplier ?

Could you post a higher quality scan ?

Cheers,

Jon

[387cbrb]

i attach a better resolution schematics of what i have in mind.
any comments/criticism?

John, i wanted to avoid active elements in series with
the supply rail. perhaps it's not a good idea when digital
currents are involved? i'm new to digital, hence seeking
advice

i rather thought of putting a transistor on top of tl431.

cheers!

[JonHarrison]

Hi,

i wanted to avoid active elements in series with
the supply rail. perhaps it's not a good idea when digital
currents are involved?

Why do you want to avoid active elements in the supply line, just curious ? In which case you wouldn't consider feeding the shunts with a constant current source ?

Cheers,

Jon

[387cbrb]

hi!

the reason is very simple -- i don't want the audio signal to go through the active elements. in tube world, where i come from, it's a well established fact that shunt tube reg's sound better than series tube reg's. current source behind a shunt is a bit different story and i will try it also.

but my question remains open: since shunts (discrete or integrated) are acknowledged as being superior to series regs, why not pre-regulating with a shunt too and getting rid of a series reg at all? perhaps it simply doesn't matter as a good shunt screens the regs behind? any ideas?

cheers!

[JonHarrison]

Hi,

but my question remains open: since shunts (discrete or integrated) are acknowledged as being superior to series regs, why not pre-regulating with a shunt too and getting rid of a series reg at all? perhaps it simply doesn't matter as a good shunt screens the regs behind? any ideas?

Please forgive my ignorance here (I'm learning) but a shunt regulator draws a constant current from the supply - what the load doesn't draw is dumped into the ground therefore the pre-shunt regulator doesn't actually have to 'regulate' anything as it simply supplies a constant DC voltage to the shunt regulator ? So, you want to feed the shunt regulator with something that supplies a clean DC and filters out AC noise ? You could use another (pre) shunt regulator but that will simply dump more current into the ground plus it also has a relatively wide bandwidth so will also couple noise into the second regulator ?

Cheers,

Jon

[rfbrw]

Originally posted by 387cbrb


but my question remains open: since shunts (discrete or integrated) are acknowledged as being superior to series regs, why not pre-regulating with a shunt too and getting rid of a series reg at all? perhaps it simply doesn't matter as a good shunt screens the regs behind? any ideas?

cheers!

As they are in parallel with each other there would have to be a current source or a dropping resistor between them so that they are at different voltages.

[Art]

Hi 387cbrb,
I replaced series R in normal TL431 circuit by LM317 in CCS mode. I observed a bit better sound in my non-OS TDA1541A and AD1865 DAC. I think about power J-Fet CCS too. I'll try it soon. Hope this benefit to you too. Any comment are welcome.
Regards,
Art.

[isdom]

hi sorry to change the subject but is it ok if you send me a pdf copy of that marantz cd46 manual / schematic, cheers.

my email is

removed by moderators

kind regards dom