JGs Nobrainer and Nobrainer Discrete - Page 10 - diyAudio
Go Back   Home > Forums > Source & Line > Analogue Source

Analogue Source Turntables, Tonearms, Cartridges, Phono Stages, Tuners, Tape Recorders, 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 1st August 2012, 12:21 AM   #91
diyAudio Member
 
Joachim Gerhard's Avatar
 
Join Date: Mar 2009
I started now to build the two other channels. I will build them non inverting like the first ones and do the Elliott trick. I have put this little amp now over 80Hz on my 95dB panels i have build for the ETF. Now decent volume is possible. I must say that i can not complain.
The amp sounds neutral and well resolved with a small measure of warmth.
When we do the PCB we can include the option of inverted and also current drive but at the moment i have no desire to change it.
  Reply With Quote
Old 1st August 2012, 01:29 AM   #92
diyAudio Member
 
Joachim Gerhard's Avatar
 
Join Date: Mar 2009
I studied Douglas Selfs Audio Power Amplifier Design Handbook for the 100 x time.
He is an advocate of the CFP stage i use here too. This stage works best in Class B and Class A. Class A/B curiously gives the highest distortion. Also it is quite clear that lowering the value of the emitter resistors ( actually here they are on the collectors of the output transistors ) to the lowest practical value that gives thermal stability has an advantage in terms of crossover distortion and also drives low impedance loads better. How the bias is set is somewhat unclear. In the mathematical example with 0.1 Ohm resistors it is 3.06mV over both resistors. That is only 18mA of bias. A bit hard to set because there is some 10% tolerance in the emitter resistors and even some mV of offset ( fortunately low ) can screw that up badly. Stabilizing that thermally is not easy ether. In his Trimodal Amp he sets that value to 10mV in praxis. That seems more reasonable also in the view that a CFP output is somewhat less sensitive to over bias ( what a name ! )then a conventional double emitter follower ( Darlington ). So we end up with 50mA. That is reasonable. Lets see what sounds better, more or less bias.
  Reply With Quote
Old 1st August 2012, 08:21 AM   #93
FdW is online now FdW  Netherlands
diyAudio Member
 
FdW's Avatar
 
Join Date: Feb 2009
Location: Both feet on the ground (near Gouda)
Quote:
Originally Posted by Joachim Gerhard View Post
I studied Douglas Selfs Audio Power Amplifier Design Handbook for the 100 x time.
He is an advocate of the CFP stage i use here too. This stage works best in Class B and Class A. Class A/B curiously gives the highest distortion. Also it is quite clear that lowering the value of the emitter resistors ( actually here they are on the collectors of the output transistors ) to the lowest practical value that gives thermal stability has an advantage in terms of crossover distortion and also drives low impedance loads better. How the bias is set is somewhat unclear. In the mathematical example with 0.1 Ohm resistors it is 3.06mV over both resistors. That is only 18mA of bias. A bit hard to set because there is some 10% tolerance in the emitter resistors and even some mV of offset ( fortunately low ) can screw that up badly. Stabilizing that thermally is not easy ether. In his Trimodal Amp he sets that value to 10mV in praxis. That seems more reasonable also in the view that a CFP output is somewhat less sensitive to over bias ( what a name ! )then a conventional double emitter follower ( Darlington ). So we end up with 50mA. That is reasonable. Lets see what sounds better, more or less bias.
To measure bias you can insert a resistor in the power line and measure it there.
  Reply With Quote
Old 1st August 2012, 02:03 PM   #94
diyAudio Member
 
Joachim Gerhard's Avatar
 
Join Date: Mar 2009
I will do the emitter resistors from several 1 Ohm metal films in parallel. I think that way i can measure the idle indirect by measuring the voltage over them. Your idea is good too.
I am more concerned if i can stabilize that low amount of idle and how it sounds. I think i have to measure distortion in the end.
  Reply With Quote
Old 1st August 2012, 02:22 PM   #95
FdW is online now FdW  Netherlands
diyAudio Member
 
FdW's Avatar
 
Join Date: Feb 2009
Location: Both feet on the ground (near Gouda)
Quote:
Originally Posted by FdW View Post
To measure bias you can insert a resistor in the power line and measure it there.
The thing about this is, you can use (almost) any value (even up to 10 Ohms) if you add a buffer capacitor on the south side (PSU on the north side ). So even with an less expensive multimeter you get good results (I'm sure you have a good one (I did see a picture of your UNI-T somewhere )). Using a high'ish resistor will also help keep the smoke in the output device when things go pear-shaped. http://en.wikipedia.org/wiki/Pear-shaped

Last edited by FdW; 1st August 2012 at 02:26 PM.
  Reply With Quote
Old 1st August 2012, 02:25 PM   #96
diyAudio Member
 
Joachim Gerhard's Avatar
 
Join Date: Mar 2009
Yes, i have a big UNI-T. I bought it to measure input offset on my phono stages. I think it goes down to 0.01mV.
  Reply With Quote
Old 7th August 2012, 03:58 PM   #97
diyAudio Member
 
Joachim Gerhard's Avatar
 
Join Date: Mar 2009
I am building two more channels of the SBP to make a bridged amp. I also experimented with Bias and Output Inclusive Compensation.
I got now better and bigger heat sinks so i set the Bias at around 500mA. That gives 4W in Class A into 8 Ohm per module. I will listen if that sounds better. I will also try Class B as recommended by D.Self. That is then less then 50mA of Bias.
The Output Inclusive is kind of interesting. I was afraid that it could make the amp unstable into capacitive load but the contrary happened. Putting a 0.33uf across a 6.6Ohm resistor there was barely any change visible with a 10kHz square. What it also does is slowing down the amp. The -3dB frequency in the treble is aproximately cut in half. It also seems that the rolloff has a higher order so until 20kHz there is not much loss compared to the miller compensated version. Both effects have not being described by Self in the Linear Audio 0 article and may be due to the fact that i use a different topology. I could of cause half the Miller caps and get the speed back.
As simple as this little thing is, building the SBP is fun and i am still learning.
  Reply With Quote
Old 7th August 2012, 04:09 PM   #98
FdW is online now FdW  Netherlands
diyAudio Member
 
FdW's Avatar
 
Join Date: Feb 2009
Location: Both feet on the ground (near Gouda)
Quote:
Originally Posted by Joachim Gerhard View Post
I am building two more channels of the SBP to make a bridged amp. I also experimented with Bias and Output Inclusive Compensation.
I got now better and bigger heat sinks so i set the Bias at around 500mA. That gives 4W in Class A into 8 Ohm per module. I will listen if that sounds better. I will also try Class B as recommended by D.Self. That is then less then 50mA of Bias.
The Output Inclusive is kind of interesting. I was afraid that it could make the amp unstable into capacitive load but the contrary happened. Putting a 0.33uf across a 6.6Ohm resistor there was barely any change visible with a 10kHz square. What it also does is slowing down the amp. The -3dB frequency in the treble is aproximately cut in half. It also seems that the rolloff has a higher order so until 20kHz there is not much loss compared to the miller compensated version. Both effects have not being described by Self in the Linear Audio 0 article and may be due to the fact that i use a different topology. I could of cause half the Miller caps and get the speed back.
As simple as this little thing is, building the SBP is fun and i am still learning.
I think that 50mA is a bit low (but I'm not D. Self ). Check the datasheet of the power device and like here ON Semiconductor MJL4281A: NPN Audio Bipolar Power Transistor you may see a 'Gain Linearity from 100 mA to 5 A' like specification (or you should interpret the graphs in the datasheet), select this minimum as your minimum for 'B'-like operation of the amp (in this case 100mA).
  Reply With Quote
Old 7th August 2012, 04:15 PM   #99
diyAudio Member
 
Joachim Gerhard's Avatar
 
Join Date: Mar 2009
Yes Frans, that is true for an EF follower ( Self give somewhat over 100mA, depending on the emitter resistors, that has to do with that famous 24,5mV ) but this is a CFP ( Szikley ) with 100% feedback so optimum bias is much lower if you do not bias high into Class A.
  Reply With Quote
Old 7th August 2012, 04:26 PM   #100
diyAudio Member
 
Joachim Gerhard's Avatar
 
Join Date: Mar 2009
What is different to Selfs output stage is that i degenerate the drivers much more. I use 100
Ohm and he uses 10 to zero Ohm. That of cause lowers the voltage gain ( in my case it should have 1x, so no voltage gain) but it helps to linearize, a trick i copied from JLH.
Another reason i do that is to lower the current that is supplied by the drivers that go into the load. That way i can use high speed small signal transistors. What effect that has on the optimum BIAs i can not tell at the moment. In JLHs books i have not found how to bias that variety optimum.
  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
My first discrete amp danfo098 Solid State 60 30th July 2009 05:40 AM
Best discrete regulator ? ANADIGIT Solid State 1 1st December 2008 02:55 PM
discrete DAC Bernhard Digital Source 6 26th September 2008 12:24 PM
My discrete opamp udip Solid State 51 8th August 2006 07:33 AM
which discrete preamp ? sopapo Solid State 12 13th January 2006 08:11 PM


New To Site? Need Help?

All times are GMT. The time now is 01:09 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