NOS TDA1541 + D1 I/V & output stage

Jump to Latest
peppennino

peppennino

Member
Joined 2005
flickr.com/photos/194234898@N02/albums
Last edited:
peppennino

peppennino

Member
Joined 2005
flickr.com/photos/194234898@N02/albums
Thanks ZM!
Actually I use passive I/V with 90 Ohm resistor.

D1 schematic:

PL D1 OS.png
 
B

Ben Mah

Member
Joined 2012
Paid Member
From NP's article:

......... send the input to the Source pin. The Source pin has is a relatively low impedance, more or less the inverse of the transconductance of the gain device, which satisfies one of the conditions that a current source DAC is looking for.
 
Zen Mod

Zen Mod

Official Court Jester
Joined 2003
zenmod.in.rs/
Paid Member
peppennino said:
Yup, I know it's a common gate, but what's the impedance seen by dac chip?
Click to expand...

DAC chip output, being current source, regarding resulting THD, best load is 0 Ohms
but, resulting voltage swing - product of chip current swing across loading impedance, then is 0V, and we still want some voltage there, because without voltage there is no music :clown:

so, instead of using resistor of some moderate value, compromise between THD and voltage swing, we are feeding chip current in a middle of constant current source , point having low input impedance, while still conducting entire current swing of DAC chip

see where input point of D1 is, source of mosfet ; observe INR point

if you have 5mA (arbitrary figure, lazy to check 1541 data) of current swing out of DAC chip, it is entirely transferred up through D1 current path; said current path is R4,Q1,R5,R6,R7,R8

so , when your measly 5mA of current swing do its thing across R4, there is suddenly 5mA * 1K5 = 8Vpp (2V82rms) of voltage swing, which is sent to gate of Q2 source follower, then meeting outer world with low Rout, at OUTR point

Papa's words (tnx to Ben) , xconductance of IRF510 is in range of 0.5-1S (lazy to search datasheet showing xconductance at small currents; same always declared at damn manly levels)

which means Rin at its source is 1/S ...... few Ohms max

output at TDA1541 loaded with few Ohms max of normal resistor should be of decent quality, but no damn voltage swing

with decent I/V stage - practically - you're loading I output chip heavily (proper way) without robing it of possible current swing
 
  • Like
Reactions: 3 users
B

Ben Mah

Member
Joined 2012
Paid Member
As you said, the TDA1541A actually wants see a low impedance load for best performance.

The data sheet (page 7) says, "To ensure no performance losses, permitted output voltage compliance is ±25 mV maximum."

The current output of the chip is +/-2mA, with an offset of -2mA.

So for +/-25mV, maximum, the chip wants to see about 12 Ohm maximum. I'm sure that is why peppennino is interested in the D1's input impedance. :)

But then I load my dac with about 70R IV resistor and enjoy its sound. diyAudio member ecdesigns (John Brown), who has done a lot of work with the TDA1541A, has suggested that up to 150R may be used for IV.

ecdesigns IV resistor

But the D1 circuit is cool. The input mosfet is a current buffer. :)

I may try it but use a lower value for R4 since I don't need 3Vp output.
 
Last edited:
  • Like
Reactions: 1 users
Vunce

Vunce

Member
Joined 2014
Paid Member
Thank you for the detailed D1 circuit explanation ZM ;)
My application is slightly different than Peppennino's TDA1541A but from your comments, R4 should be doubled for a dac chip with Iout of +/-1mA and no offset. Would there be any other modifications?
Perfect timing with this thread Peppennino, Good luck with your TDA1541A dac project. (I'm using an AD1862)
:cheers:

(I should add: PCM63 Iout is +/-2mA)
 
Last edited:
  • Like
Reactions: 1 user
Zen Mod

Zen Mod

Official Court Jester
Joined 2003
zenmod.in.rs/
Paid Member
Vunce, in case of needing different circuit gain, I would go with circuit optimization - most likely change of positive rail, to retain same operating values ( Iq, voltage levels) for both input and follower mosfets

what Ben said, 1541 2mA sinking, with D1 stage - taken care of with proper setting of PR/PL, no need for small additional CCS
 
You must log in or register to reply here.
Top Bottom