Go Back   Home > Forums > Source & Line > Digital Line Level

Digital Line Level DACs, Digital Crossovers, Equalizers, 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 11th May 2010, 10:08 AM   #121
anbello is offline anbello  Italy
diyAudio Member
 
anbello's Avatar
 
Join Date: Sep 2004
Location: Italy
Hi All,

I would like to know how critical are the values and kind of some of the components on the schematic posted by Owen on #61:
can i use 182 in place of 185, 390 in place of 400 and 220 in place of 221 for the resistors?
can i use some other kind of capacitor for the 10uF 25V X7R like film of what else?

Thanks
Andrea
  Reply With Quote
Old 11th May 2010, 07:01 PM   #122
opc is offline opc  Canada
diyAudio Member
 
opc's Avatar
 
Join Date: Oct 2004
Location: Ottawa, Ontario, Canada
Hi Andrea,

None of the parts you mentioned are going to be very critical. The 182 ohm is fine, and going from 220 to 221 won’t make a difference either.

If possible for the 400 ohm, go a little higher rather than lower. 390 ohm will be fine, but you’ll give up just a little bit of gain. If you can go up to either 410 or 420, that would probably be best.

The cap types don’t really matter much, and film would be fine in place of the ceramic. You could also use polymer or electrolytic, but a larger value might be in order.

For everyone else, I have some neat stuff to post, and I’ll get around to it hopefully tonight to tomorrow.

Cheers,
Owen
  Reply With Quote
Old 12th May 2010, 08:22 AM   #123
anbello is offline anbello  Italy
diyAudio Member
 
anbello's Avatar
 
Join Date: Sep 2004
Location: Italy
Thanks a lot Owen,
really kind

Ciao
Andrea
  Reply With Quote
Old 12th May 2010, 12:55 PM   #124
opc is offline opc  Canada
diyAudio Member
 
opc's Avatar
 
Join Date: Oct 2004
Location: Ottawa, Ontario, Canada
Default Updates

Hi Guys,

As promised, I have a few things to report from the last couple days of testing. This transconductance thing is tricky, and there's a little more to it than I initially thought. I wrongly assumed that if a mosfet has high TC at higher currents, that it would more likely have higher TC at lower currents. I actually sat down and made a test setup to measure several fets, and here's what I came up with so far:

Using 50mA, and 10V DS, 200 ohm source resistance:

FQA46N15: 0.355 S
IXTQ96N15: 0.276 S
FQA19N20C: 0.397 S
ACD101NDD: 0.216 S
FQA28N15: 0.340 S
IRF1324: 0.297 S
IRF540: 0.357 S
IRF740: 0.258 S

These are just a few that I had on hand which I thought would be interesting, and as you can see, the TC performance at such a low bias current is a complete toss up. There's no trend or other characteristic that helps to predict TC at such low levels.

The clear standout so far, however, is the FQA19N20C as used in Nelson's F5. I matched a pair of these for Vgs along with a pair of the IRF710s, and I'm going to do a head-to-head comparison. That's basically the best in the group against one of the the worst as far as TC goes.

Another fun exercise that I did was to crank up the rail voltage to get higher bias current. It's a neat circuit because you can basically just drop the gate voltage all the way down (adjust the pot to read 0V at the gate) then you can ramp up the rails as high as you want (limited by resistor and FET dissipation) then re-adjust to get the 1.65V at the source and you're off to the races.

So, I built up one channel using 50W power resistors, and put the mosfets (FQA19N20C) on heatsinks. I used a lab supply for the +/- rails which made adjustments easier.

Sure enough, the higher you go with rail voltage, the better and better performance gets. I'm not talking by a small amount either. I made it up to +/- 75V which gives about 180mA bias current per leg, for a total of 54W dissipation per channel. At this bias level, I was getting -112dB THD+N, and the frequency response was ruler flat out to 80kHz which is as high as I can measure with a 192kHz input.

That kind of performance is phenomenal for a single stage with a pair of mosfets. That's better than most people can do with three $20 op amps and a plethora of supporting circuitry/power supplies. If you have heatsinks, and don't mind owning a DAC that dissipates 100W, then I'd give that setup my full blessing.

I'll post my measurement and some pictures of "The beast" (probably has a record for the world's hottest DAC) later today or tomorrow.

Cheers,
Owen
  Reply With Quote
Old 12th May 2010, 01:41 PM   #125
qusp is offline qusp  Australia
diyAudio Member
 
qusp's Avatar
 
Join Date: Oct 2009
Location: Brisbane, Australia
so looks like the 540 still performs rather well, ok i'm gonna give that a whirl in the next few days. will crank up the sigma 22 to +/-50v and finish off my slab-o-copper heatsinks, bolt the lot to them and let her rip. anything else I should be aware of with the buffer in the circuit in these circumstances Owen?
  Reply With Quote
Old 14th May 2010, 01:04 PM   #126
opc is offline opc  Canada
diyAudio Member
 
opc's Avatar
 
Join Date: Oct 2004
Location: Ottawa, Ontario, Canada
Hi qusp,

If you're going to up the voltage, you'll have to lose the buffer. To be honest, the buffer is really not doing anything for you if you're using the resistance values I suggested. Without the buffer, you've got an output impedance of 200 ohms, and with it you jump to 1k. Not really much of a buffer, so just get rid of it. Cheaper, easier to build, and better performance.

With the higher voltage, you'll have 25V across each jfet and at 10mA that's 0.25 watt's per device which I think is too high for a TO-92 in the long term.

Cheers,
Owen
  Reply With Quote
Old 14th May 2010, 01:10 PM   #127
lauret is offline lauret  Netherlands
diyAudio Member
 
lauret's Avatar
 
Join Date: Oct 2007
Location: Amsterdam
Quote:
Originally Posted by opc View Post

I'll post my measurement and some pictures of "The beast" (probably has a record for the world's hottest DAC) later today or tomorrow.
Your findings are really cool! Any pictures of the beast yet?
  Reply With Quote
Old 14th May 2010, 05:33 PM   #128
diyAudio Member
 
Join Date: Feb 2004
Location: Davis,CA
Quote:
Originally Posted by opc View Post
Hi qusp,

If you're going to up the voltage, you'll have to lose the buffer. To be honest, the buffer is really not doing anything for you if you're using the resistance values I suggested. Without the buffer, you've got an output impedance of 200 ohms, and with it you jump to 1k. Not really much of a buffer, so just get rid of it. Cheaper, easier to build, and better performance.

With the higher voltage, you'll have 25V across each jfet and at 10mA that's 0.25 watt's per device which I think is too high for a TO-92 in the long term.

Cheers,
Owen
Owen:


Great project ! Nelson and company have done great things and you are definitely continuing their tradition.

Is post # 61 considered to be without the B1 buffer?

Thanks

Bob
  Reply With Quote
Old 14th May 2010, 05:53 PM   #129
qusp is offline qusp  Australia
diyAudio Member
 
qusp's Avatar
 
Join Date: Oct 2009
Location: Brisbane, Australia
Hi Owen, i've only got Jfets as one thing to try, I'm actually thinking the IRF540 for higher voltage as you suggested. this will work fine at 50V yes? if I was to use JFETS I would use lower current, or the sic devices like R550

Last edited by qusp; 14th May 2010 at 05:56 PM.
  Reply With Quote
Old 14th May 2010, 06:58 PM   #130
opc is offline opc  Canada
diyAudio Member
 
opc's Avatar
 
Join Date: Oct 2004
Location: Ottawa, Ontario, Canada
Hi Guys,

Ive got some graphs to post finally, so feel free to take a look below. I also measured several more fets, so here's a re-cap:

FQA46N15: 0.355 S
IXTQ96N15: 0.276 S
FQA19N20C: 0.397 S
ACD101NDD: 0.216 S
FQA28N15: 0.340 S
IRF1324: 0.297 S
IRF540: 0.357 S
IRF740: 0.258 S
IRF640: 0.256 S
IRF510: 0.054 S
IRFP4310Z: 0.374 S
IRF1310: 0.356 S
IRF644: 0.281 S
FQA32N20: 0.419 S
ST75NF20: 0.199 S
ST40NF20: 0.189 S
IRFP3415: 0.395 S

What you can take away from this is that you shouldn't use the IRF510, and that you have a plethora of other choices that are good. Looks like the FQA32N20 is the best of this group at 50mA and 10V DS.

As for the graphs, I used the same circuit, and adjusted the rails to various levels, taking measurements at each step. The voltages were 18V, 25V, 30V, 35V, 40V, 45V, and 50V.

As you can see, things just keep getting better the higher the voltage gets.


Bob:

The schematic in post #61 is the non-buffered version, and now that I look back, you should omit the 100k resistor across the two outputs before the cap. I thought it would help with start-up transients, but that turned out to not be a problem.

That's the schematic used in the measurements below, minus two filter caps which need to be adjusted.

qusp:

The IRF540 should be fine for the higher voltage version, just make sure you've got a heatsink. At 50V, the top resistor dissipates about 3.5W, the bottom resistors about 6.5W and the FET will dissipate about 3W.

Cheers,
Owen
Attached Images
File Type: jpg THD+N Ratio FR VOLTAGE.jpg (67.9 KB, 966 views)
File Type: jpg THD+N Ratio LEVEL VOLTAGE.jpg (58.1 KB, 953 views)
File Type: jpg FR VOLTAGE.jpg (62.3 KB, 941 views)
File Type: jpg THD+N Ratio 50V.jpg (89.7 KB, 928 views)
File Type: jpg Distortion Product Ratio 50V.jpg (90.8 KB, 916 views)
File Type: jpg D1.JPG (213.3 KB, 483 views)
  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
Pass Labs S/Ns ? dejanm Pass Labs 8 26th January 2007 07:35 AM
My opinion on Pass Labs and Mr. Pass (Nelson) himself b_online Pass Labs 11 21st May 2003 12:39 AM


New To Site? Need Help?

All times are GMT. The time now is 12:07 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