OPA861 for I/V

Status
Not open for further replies.
Jump to Latest
A

AudioLapDance

Member
Joined 2003
#1
Noticed there was no specific thread for this ...



From TI's informative 2009 Current vs Voltage Feedback presentation (OPA861 pg. 8)

Features:

WideBandwidth OTA(>80Mhz)
900V/usecSlewRate
2.4nV/√HzInputNoiseVoltage
Very Flexible Circuit Building Block
Ideal Complementary Transistor Function
Tuneable Transconductance
Externally settable supply current

http://www.ti.com/lit/ds/symlink/opa861.pdf

http://www.ieee.li/pdf/viewgraphs/current_feedback_vs_voltage_feedback_amplifiers.pdf

Pedja's AYA II circuit:



www.audialonline.com/?wpdmact=process&did=MS5ob3RsaW5r

And some good threads:
(go back a few pages)
http://www.diyaudio.com/forums/digi...ltiple-tda1541a-no-tube-no-transformer-2.html

http://www.diyaudio.com/forums/digital-line-level/162330-yet-another-i-want-make-dac.html

http://www.diyaudio.com/forums/digital-line-level/161377-output-level-tda1545-5.html


DIYHiFi.org • View topic - Discrete I/V
 
Last edited by a moderator:
#2
Good stuff from TI's 2009 Current vs Voltage Feedback presentation:

V vs C opamps.JPG

V vs C opamps2.JPG


V vs C opamps chart.JPG
 
#4
Hi,

I simmed the OPA861 for PCM179x DACs as sources quite a while ago with a cascoded NJFET-Buffer (LSK389C or BF862 + 4391) that is THD and Noise transparent.
pros:
- simplicity
- very constant input impedance into the 10s of MHzs
- very fast --> clean square response
- nice THD spectrum with clean decay of harmonics

cons (not really, but rather mediocre performance):
- THD-niveau (close to -60dB/0.1% at 0dBfs)
- Noise figures just ok, though depending to a large dgree on the value of the IV-resistor.

It should be noted that DACs with offset currents require a current source or - sink to cancel that offset.
A simple JFET ccs may suffice here as best solution (noisewise) but will inevitably introduce some noise.
In fact it might be the dominant noise source of the whole circuit!

jauu
Calvin
 
#9
I was using the OPA861 as per the above schematic as part of the AYA II DAC.

It's good, and to my ears better than passive I/V into WE396A (with a good CCS fed shunt reg B supply). This was with the 2mA current injection as per the scheme.

I was nulling for 0VDC at the buffers output and driving directly a rather decent TVC. I did find that for some reason it didn't like driving a purely inductive load, directly, so went back to nulling for 0VDC at the 1541a output and coupling to TVC with a high quality capacitor, after trying many, preferring actually BG Hi NX.. Which surprised me.

Although I didn't like the idea compared with direct connection, it actually sounded a bit better and it was stable. I had fried a few OPA driving the inductance, for no other reason that I could discern and those SMD parts are a PITA to replace.

Recently I tried another tube stage, 6072A biased by the current offset of the 1541a itself through a 33R resistor, same as what Thorsten proposed. Near 0V bias, 10R carbon in the cathode cct. Pimm CCS loaded with mu output of that device and parafeed into the TVC connected as AVC. Removed the CCS fed shunt reg and used tube regulated choke input filter. To my ears is a better stage and will probably stay that way.

I'd encourage a build using OPA, for sure. And if you have the space, try Thorstens tube stage, the recent one as hybrid mu follower.

Regards,
Shane
 
#11
Ceglar said:
I'd encourage a build using OPA, for sure. And if you have the space, try Thorstens tube stage, the recent one as hybrid mu follower.

Regards,
Shane
Click to expand...

The transcoductance (current amplifier) with AD811 is better than OPA861, and much better than AD844 (I only ever tried single 844 in diamond buffer config, no feedback, but this had a lot of distortion - too much for my liking, couldn't listen...). The OPA861 sounded better compared to an AD844. The DAC was initially 1541. Once I settled on 811, I implemented the same I/V stage concept to dual differential 1704's (inside heavily modified Denon 1650 CD player) with very good results.

I will probably give stacked 844 a go... just need to find the NOS 1541 DAC that I still have... gathering dust somewhere...

But these days I can't be bothered to look for, and place the CD's inside the player(s) any more... it's iMac and modified W4S DAC2 DSD that keep me happy, and the sound isn't half bad neither.

Nick
 
#12
Extreme_Boky said:
The transcoductance (current amplifier) with AD811 is better than OPA861, and much better than AD844 (I only ever tried single 844 in diamond buffer config, no feedback, but this had a lot of distortion - too much for my liking, couldn't listen...). The OPA861 sounded better compared to an AD844. The DAC was initially 1541. Once I settled on 811, I implemented the same I/V stage concept to dual differential 1704's (inside heavily modified Denon 1650 CD player) with very good results.

I will probably give stacked 844 a go... just need to find the NOS 1541 DAC that I still have... gathering dust somewhere...

But these days I can't be bothered to look for, and place the CD's inside the player(s) any more... it's iMac and modified W4S DAC2 DSD that keep me happy, and the sound isn't half bad neither.

Nick
Click to expand...

Maybe it was implementation but I had a commercial PCM63 DAC with AD811 output and it was terrible, impossible to keep those from oscillating in that service I think.


As far as the PCB. I think we need to focus on the important things:
The SMD resistor pads & placement for the TDA
The Layout to keep the EMF/buzzes at a minimum
NOS

The DIY community doesn't have a quality PCB with attention to RF.
 
#13
😀 The dissidents thread... Hi guys,

Regal... we try to do it with goodwilled people elswhere...

We are going to ask a special next Linear Audio review just for the story of all the I/V outputstages for curent DAC chip 🙂


I am not the one who are thinking the brand new is always better than the older design, but is somebody here or elswhere compared the elegant CEN/SEN optimised for the TDA 1541 to the also good OP861 Pedja's output with the same source and with the component choosed by P. Rodric ????
 
Last edited:
#14
regal said:
Do you have a link I missed that one?
Click to expand...

You must be meaning T's tube stage as the OPA861 is shown above.

B supply 200-250V, 'ring of two' CCS plate load, output taken from Mu output via ~2.2uF with 1M shunt to ground. Grid circuit has I/V resistor to ground in parallel with shunt C for LPF (maybe un neccersary), scale to suit, with series grid stop (100R to 1k). Cathode circuit 10R only.

6922 12mA Ip ~ 70V plate
6072A 3.5 Ip and ~100V plate.

hth,
shane
 
#15
Extreme_Boky said:
The transcoductance (current amplifier) with AD811 is better than OPA861, and much better than AD844 (I only ever tried single 844 in diamond buffer config, no feedback, but this had a lot of distortion - too much for my liking, couldn't listen...). The OPA861 sounded better compared to an AD844. The DAC was initially 1541. Once I settled on 811, I implemented the same I/V stage concept to dual differential 1704's (inside heavily modified Denon 1650 CD player) with very good results.

I will probably give stacked 844 a go... just need to find the NOS 1541 DAC that I still have... gathering dust somewhere...

But these days I can't be bothered to look for, and place the CD's inside the player(s) any more... it's iMac and modified W4S DAC2 DSD that keep me happy, and the sound isn't half bad neither.

Nick
Click to expand...

can u post your AD811 I/v schematic ?
 
#20
Pardon me for bring up this old thread again. I've built the OPA861 I/V circuit with ADA4627 as the buffer. The sound is almost as good as the other stacked AD844 I/V in that OPA861 I/V suffers from higher DC offset which is around 30mv on my implementation. I don't like adding a cap on the signal path, so I just leave it as is. What's the DC offset on your implementation? Is 30mv too much on an OPA861?
 
Status
Not open for further replies.
Top Bottom