CD/DAC Output Stage for Purists - diyAudio
Go Back   Home > Forums > Source & Line > Digital Source

Digital Source Digital Players and Recorders: CD , SACD , Tape, Memory Card, etc.

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 29th January 2002, 11:33 PM   #1
Dave is offline Dave  New Zealand
diyAudio Member
 
Join Date: Oct 2001
Location: New Zealand
Lightbulb CD/DAC Output Stage for Purists

Here is an idea I had for those of you who want to try something a little different in your DAC/CD output stage. Take a look at the schematic which is attached. Basically it is a common base amplifier used as an I/V stage suitable for PCM1702/1704/63 or similar current output style DACs. This is STILL VERY MUCH A SKETCH at the moment but I thought some of you may wish to have a play and tweak it.

T1 forms the actual gain transistor and I suppose it can be considered the only transistor in the signal path. T2 creates a voltage reference for T1s base. T3 and T4 form constant current sources the current through these should be nearly identical.

The emitter of T1 will provide a low impedance point for the DAC chip. The DC voltage at this point will probably need trimming to zero by a pot arrangement around Rib enabling you to vary the bias current through T1 either side of the current level set by fixed current source T3.

I am not sure what the bias current should be but I would say that 10mA would be a good starting point.

Ri is where the actual I/V conversion takes place and the value chosen for this resistor is a direct compromise between output level and output impedance. Its fairly clear to see that its value is the output impedance. The maximum output level (RMS) of the circuit is equal to the maximum RMS current from the DAC chip multiplied by the value of Ri. For example if it is being used with say a PCM1704 (1.2mA) peak and Ri=1K then the maximum drive level is 0.8485v RMS with an impedance of 1K. Clearly this is not your usual CD Player drive but should be suitable for amplification by a good pre-amp, a ZEN for example.

It will be important to watch the voltage drop across Ri to ensure that T1 has an adequate Vce drop, you may need to raise the supply voltage. Speaking of supply it will have to be very well filtered as any noise on the positive supply will appear at the output.

Lastly this design has a very wide bandwidth and digital noise will be passed through to the output. The bandwidth can be decreased by adding a capacitor across Ri. There is lots of room for experimentation here and results will probably vary for different amplifiers and systems.

I would really like some opinions on this so let me know what you think.
Attached Images
File Type: png simpleiv.png (1.3 KB, 1828 views)
  Reply With Quote
Old 30th January 2002, 08:16 AM   #2
diyAudio Member
 
hill's Avatar
 
Join Date: Nov 2001
Location: G.Z. China
Ha, you'er a genius if you had never knew the current mode technology.

Yes, it's a good idea, but it isn't prefect.

1. There're tow LEDs which be used in constant current source. It will caused un-match current between these two CCS. In this case, the collect current of T1,T2 must be matched.

2. The rbb1, rbb2 will cause high distortion. rbb is resistance of base area of transistor. You may needed to generate a negative resistance to reduce it's affort.

3. You may needed to add a low pass filter to reduce noise.

4. You may needed to add a output buffer to iso this stage, so it can reduce distortion. The output impendance of this stage is Ri // Rload_impendance.

My english is not well. I hope that you can understand what thing i want to say.
__________________
I change therefore I am.
  Reply With Quote
Old 30th January 2002, 10:30 AM   #3
Dave is offline Dave  New Zealand
diyAudio Member
 
Join Date: Oct 2001
Location: New Zealand
hill thanks for your info.

I ran some quick simulations in the multisim 2001 demo and came up with the following

THD was around 0.0035%

This was obtained under the following conditions.

T1 & T2 were 2N2222s. Input current 1.2mA peak (2.4mA p-p ie. full drive of PCM1704/2) at 10KHz analyzed up to the 9th harmonic. Bias current = 10mA, Ri = 1K and the supply rails were +/- 20v.

Of course the output of the DAC was assumed to be a perfect current source so I decided to change the model for a voltage source in series with a 1K resistor - the output of the voltage source was 2.4v p-p.

The THD was now 0.0055%

The impedance at the emitter of T1 is roughly 3.5 ohms but clearly this is not resistive but even so the above voltage source/ resistor demostration shows that it may not affect THD at the output that much.

The bandwidth, around 40MHz. This can be greatly decreased by a capacitor in parallel with Ri. In this case 3nF in parallel with 1K gave a bandwidth of about 50KHz.
  Reply With Quote
Old 30th January 2002, 11:03 AM   #4
diyAudio Member
 
ftorres's Avatar
 
Join Date: Apr 2001
Location: Limoges, France
Default What about this one ?

Hi,

Reminds me of something I thought of a few years ago. Basically, it's a common gate mosfet, followed by a buffer (Thanks Nelson for the idea ).

The Jfet diff pair is here to maintain the source of M1 (i.e. the current input) at 0V. In a previous version, I used a BJT diff pair, but I'm not sure how the base currents drawn by the BJTs would affect the I/V conversion...

I really enjoy the previous version (with PCM63), but want to try this "new" version (differences are the Jfets instead of BJTs, and a DC servo at the output). May be I will try it with a cascoded output instead of the simple buffer stage.

Any comments welcome
Attached Images
File Type: gif ivconv.gif (8.3 KB, 2914 views)
__________________
François
"Learning French is trivial: the word for horse is cheval, and everything else follows in the same way."
  Reply With Quote
Old 30th January 2002, 03:59 PM   #5
tvi is offline tvi  Australia
diyAudio Member
 
tvi's Avatar
 
Join Date: Apr 2001
Location: Moonee Ponds, Vic, Australia
The Audio Amateur 2/89 had an interesting I/V by JM Didden

Bakkuan(sp) in Japan uses a similar ides but a bit more complicated.

Regards
James
Attached Images
File Type: jpg iv.jpg (28.1 KB, 2777 views)
  Reply With Quote
Old 30th January 2002, 11:11 PM   #6
Dave is offline Dave  New Zealand
diyAudio Member
 
Join Date: Oct 2001
Location: New Zealand
Both those look like nice designs the one by JM Didden seems to apply a simple feedback loop around the whole stage.

ftorres,

Do you know how similar your one is to the Passlabs D1 DAC?

I have also been throwing ideas around similar to your 2SK389 approach to maitain 0v at the source of the I/V MOSFET. In my current DAC I am using a OP-AMP setup as a DC Servo with a very low cut off frequency to eliminate feedback at audio frequencies. Your approach however will apply feedback around the MOSFET at audio frequencies and thus lower the impedance seen by the DAC chip.
  Reply With Quote
Old 30th January 2002, 11:13 PM   #7
Dave is offline Dave  New Zealand
diyAudio Member
 
Join Date: Oct 2001
Location: New Zealand
ftorres,

Sorry, I forgot to ask in my previous post - what sort of offset values were you getting at the source of the MOSFET with the 2SK389 ??
  Reply With Quote
Old 30th January 2002, 11:45 PM   #8
diyAudio Member
 
ftorres's Avatar
 
Join Date: Apr 2001
Location: Limoges, France
Dave,

Quote:
Do you know how similar your one is to the Passlabs D1 DAC
Yep, that's what I had in mind when I began to think of it I'm not quite sure the whole design is the same, since I never had the opportunity to approach a D1

Concerning the feedback, I must admit it's a question I asked myself (and not answered it...) : Is it better to have the Mosfet source all the time at 0V, or just to have a kind of DC servo as you use, only to correct the long term drifts (very low frequencies) of the Mosfet, just like at output ? Have you tried both ?

Concerning you last question, I'm sorry, but I don't exactly understand what you mean (my english's too bad ) Can you please elaborate ?

P.S. : I tested Didden's design long ago, but found it hard to stabilize (slow drifts). Surely some additionnal servo-type feedback will help, will try to re-test it.
__________________
François
"Learning French is trivial: the word for horse is cheval, and everything else follows in the same way."
  Reply With Quote
Old 31st January 2002, 12:37 AM   #9
jam is offline jam  United States
diyAudio Member
 
jam's Avatar
 
Join Date: May 2001
Location: Auburn, CA, USA
Ftorres,

Great design! We need more of this stuff on the site.
Looks similar to the Pass design but with the addition of the servo. I would probably capacitor couple the output stage and leave out the servo if I built it.
Do you have a board layout by any chance?

Jam

P.S. I wonder if Grey will notice you liberal use of current sources in the design.
  Reply With Quote
Old 31st January 2002, 02:19 AM   #10
Dave is offline Dave  New Zealand
diyAudio Member
 
Join Date: Oct 2001
Location: New Zealand
Ftorres,

No, I never really came to a conclusion about feedback either but a couple of people who tried similar designs decided that a servo was the way to go.

By my last question I meant what is the voltage difference between ground and the source of M1.
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
3C24 in a SE output stage - output transformer impedance recommendation !!! aldovan Tubes / Valves 15 6th September 2008 11:18 AM
preamp output stage resistor from output to V- ? ? Bernhard Solid State 4 19th October 2007 03:55 PM
Output stage biasing, with the driver stage andrew_whitham Tubes / Valves 0 19th May 2007 12:14 PM
Adding secondary of output transformer into cathodes of output stage ? Eico ST-70 freddymac406 Tubes / Valves 3 21st April 2007 07:15 PM
TV Horzontal Output Transistor for output stage..? ronybc Solid State 46 10th October 2005 03:26 AM


New To Site? Need Help?

All times are GMT. The time now is 10:23 PM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright ©1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2