my project: class B with ThermalTrak - diyAudio
Go Back   Home > Forums > Amplifiers > Solid State

Solid State Talk all about solid state amplification.

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 30th January 2006, 11:53 PM   #1
mafoe is offline mafoe  Germany
diyAudio Member
 
Join Date: Jan 2006
Location: Stuttgart
Default my project: class B with ThermalTrak

Hi all.

I freshly registered here and I would like to start posting with an introduction to my amplifier project. The name of the project is simply "MyAmp2006".

After months of reading books (D. Self, R.Slone, ...). I am still in the design and simulation phase. I use the free "SwCAD III" from LT for that.

This is the latest version of my concept:

Click the image to open in full size.

It's a 120W@8Ohms BJT class B design using three paralleled pairs of OnSemi's ThermalTrak output devices (which I got already as samples).

Results of simulation:

max input level: 2.828V (output 120W@8Ohms)
min input level: 0.044V (output 30mW@8Ohms)
DC offset: -22V
slew rate: 54 V/s (symmetric)
bandwidth (-3dB): 2Hz to 400kHz
output noise: 63nV/sqrt(Hz)
quiescent current through each output device: 10.6 mA.

These are THD values by simulation:

Click the image to open in full size.


What do you think about this design? Do above numbers look acceptable?
During simulation, I could not see any signs of instability. Do you see any instability or blow-up risks with my concept?

I would like to omit the usual output current sensing and limiting, just keeping the rail fuses. Each of the three paralleled output devices should withstand max 15A (by datasheet). 3x15A is much more than my transformer can do. Thus, rail fuses of 10A or so should make it. Do you agree?

Thanks for having a look on it.

Regards,
mafoe
  Reply With Quote
Old 31st January 2006, 12:30 AM   #2
clem_o is offline clem_o  Philippines
diyAudio Member
 
Join Date: Jun 2005
Location: Manila
Hi - looks good; how did you verify stability? Try kicking it with square wave inputs and a reactive load... Also, the rising THD vs lowering frequency (0.01% 150Hz and lower) seems odd...


Cheers!

ps: try dropping the bias currents and see if the simulation tallies. I would have thought 10mA is a bit low, so I'm wondering how 'true' the simulations are...
  Reply With Quote
Old 31st January 2006, 01:08 AM   #3
diyAudio Member
 
Join Date: Sep 2005
Good to see someone is trying the Thermaltrak transistors out!

A question for you:

You said that your output devices are biased at 10.6ma per device. What is your method for adjusting the bias current since there is no Vbe multiplier adjust? (Sorry if the answer is in your schematic; I can open the link to your schematic but I can't see the schematic picture itself).
  Reply With Quote
Old 31st January 2006, 07:50 AM   #4
mafoe is offline mafoe  Germany
diyAudio Member
 
Join Date: Jan 2006
Location: Stuttgart
Hi,

I got several people reporting that the schematics is difficult to see. After clicking the thumbnail link, a strangely compressed view of the big picture appears. This picture can be zoomed in to see all details by simply left-clicking. Sorry, it's not my invention, seems it's how the "xs.to" image hosting site is dealing with images of huge size.

Back to my project:

to check stability, I fed the input with square waves up to maximum magnitude (+2.828V, -2.828V), with 1ns rising and falling edges. Even having set the max timestep of the Spice transient analysis down to 10ns I cannot see signs of voltage over-shoot or ringing at V(SPEAKER) output. Should I check something more?

THD: At the beginning I simulated only at 1kHz and 20kHz with full power, assuming that all other values at lower input freqs and levels would be equal or below. (maybe I misunderstood Mr. Randy Slone in his book High-Power Amplifier Construction Manual.) When I started to simulate a bit more to fill such THD table, I detected that my original design showed several weak spots in this table, e.g. like 0.03% at 2400Hz, 0.177V input, but having good 0.00x values in all the cells around. Yet, do not understand the mechanism that leads to this effect.

After weeks of playing around I found that the capacitor to buffer the V_bias (made by 4 ThermalTrak diodes) plays some role. I tried values of 1F to 220F, but this lead only to a shift of the bad points in the THD table. My latest approach with several small capacitors along the V_bias diodes improved the figures significantly. I don't know why.

The setup is not yet optimized. I took the 2.2n along the diodes just like that. I'll continue to find better values, and I will try to tune the quiescent current by adjusting the R11/R33/R34 chain. You said 10mA could be a bit low. Which value would you assume?


mafoe
  Reply With Quote
Old 31st January 2006, 09:24 AM   #5
clem_o is offline clem_o  Philippines
diyAudio Member
 
Join Date: Jun 2005
Location: Manila
Hi mafoe,

You may be chasing simulation gremlins. When spice-like simulators begin to give you funny values (i.e. changing capacitors that shouldn't have any effect but does!), then you can make a guess that it's time to go to a real-world breadboard....

Regarding bias - I believe D Self does some in-depth analysis and shows that its not really the current that counts but rather resulting bias voltage as measured (I'm just recalling here, please check the chapter to make sure) from the emitter resistors. Therefore, the lower the emitter resistor value, the higher the resulting "correct" current for bias. In your case, you've chosen 0.1ohm resistors, so I expect the "correct" bias current to be pretty high. Of course, this is on the assumption that D Self is correct.

On the other hand, most designers seem to favor bias per stage at 25 to 100mA... (other guys and gals, correct me if my memory is faulty!!)

Cheers!
  Reply With Quote
Old 31st January 2006, 09:54 AM   #6
diyAudio Member
 
darkfenriz's Avatar
 
Join Date: Jun 2004
Location: Warsaw
Regarding complementary differentials with mirrors:
You're walking on the thin ice man, in sims all transistors of the same make are identical, in reality not allways.
Try simming 2 different transistors in differential pairs, like mpsa06 with 2n5551 and mpsa56 with 2sa970 and slightly different emitter resistors (+-5%). Try several times with differnt setups (hint: currents in VAS, balance) and don't be angry on me if you get dissappointed, sorry.
  Reply With Quote
Old 31st January 2006, 10:09 AM   #7
diyAudio Moderator Emeritus
 
lumanauw's Avatar
 
Join Date: Oct 2002
Location: Bandung
Send a message via Yahoo to lumanauw
Hi, Mafoe,

Welcome to the forum

Quote:
Regarding complementary differentials with mirrors:
I've built this, and it doesn't work AT ALL in real cct. I don't know why it works in SIM.

Mafoe, try to replace the current mirrors with plain R, then it will work in real cct.
  Reply With Quote
Old 31st January 2006, 10:19 AM   #8
clem_o is offline clem_o  Philippines
diyAudio Member
 
Join Date: Jun 2005
Location: Manila
Hi mafoe,

One other thing I forgot, that you may want to try - put in "funny" values for the compensation capacitors in the amp, and load it with a reactive load. The objective - make the amplifier oscillate in simulation!!

If it doesn't oscillate, then you know the simulation can't be 100% believed... (ugh, ask me why I know...)

Cheers!

Clem
  Reply With Quote
Old 31st January 2006, 10:22 AM   #9
lineup is offline lineup  Sweden
diyAudio Member
 
lineup's Avatar
 
Join Date: Dec 2005
Location: the north
There is nothing to define the current of VAS stage in this circuit.
Not as far as I can see ....



Maybe I do not understand the idea of this circuit.
I have not seen this before.

I am talking about:
R19, R16, Q9 and Q12.

see detail in my attachment
Attached Images
File Type: png thermaltrakvas.png (43.2 KB, 991 views)
__________________
lineup
  Reply With Quote
Old 31st January 2006, 09:54 PM   #10
mafoe is offline mafoe  Germany
diyAudio Member
 
Join Date: Jan 2006
Location: Stuttgart
Hi,
thanks for your interest and hints!

The roots of my project can be found in the book of Randy Slone that I mentioned above (ISBN 0-07-134119-6), on page 351. The differential input stage with current mirrors is almost copy&paste from there, including the values of the degeneration resistors. This gives me some confidence that it works. Since my rails run on lower volage (48V instead of 55V in the book) I substituted the 2N5551/2N5401 with MPSA06/MPSA56 types. A good choice?

The symmetric VAS is made of two Darlington circuits with active loads (whatever that means). In the original version in Slone's book, the cascode transistors (like Q9) are biased by a 4.4V zener diodes, each connected to its near rail. All my Spice simulations (Orcad/PSpice as well as SWCad III) had heavy convergence problems with that. So I finally decided to get rid of the zener and find another way of biasing the cascode. After trying funny circuits showing more oscillation than amplified signal, I came to the version that I posted here. I have not yet completely understood how this symmetric cascode thing works. Honestly, I don't know what defines the VAS current. I just see in the simulation that there is some current (about 45mA).

I'm on a business trip the rest of the week. I'll be back here on the weekend.

Greetings,
mafoe
  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
On Semi ThermalTrak davidallancole Solid State 639 1st May 2013 02:29 AM
Practical use of ThermalTrak diodes ceharden Solid State 5 9th April 2009 06:00 PM
OnSemi ThermalTrak Arius Solid State 8 26th August 2005 04:09 PM
ThermalTrak circuit design Mambo Solid State 4 12th July 2005 11:03 AM


New To Site? Need Help?

All times are GMT. The time now is 01:24 AM.


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