TGM8 - an amplifier based on Rod Elliot P3a - Page 50 - 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 23rd July 2014, 02:23 AM   #491
diyAudio Member
 
Join Date: Apr 2013
Quote:
Originally Posted by Bigun View Post
He could provide room on the pcb to allow for a CFP VAS, which would have a lower output impedance and is therefore, better able to drive the FET gates.
That's a great suggestion. Another (perhaps technically better?) solution would be to buffer the VAS itself with an EF or CFP driver stage.

The CFP drivers would be more in keeping with Rod Elliott's philosophy, but I would want to investigate whether one of these two methods would allow us to retain the simple trimpot for biasing the output.

I have another idea for a quasi-complimentary output stage utilising the TGM8 front end. A switching MOSFET would handle the positive swing and a CFP BJT quasi-compliment the negative. I've turned the front-end upside down for this one, with an NPN VAS transistor, so that the action of the VAS bootstrap helps better saturate the MOSFET, theoretically making the clipping more symmetrical.

The BJT compound pair on the -ve side means that the VAS only has to contend with the gate capacitance of one MOSFET - not two.

I have no idea whether this will even work, let alone sound any good, but its a concept I'm keen to explore further. Gareth, I have a feeling that quasi designs aren't your cup of tea but I'm hoping Hugh will take an interest since he recently expressed to me an interest in quasi designs.

Anyway, I don't want to clutter this thread with my babble, I'll take it to my thread over here when I post a conceptual schematic in a few days.

Very simple quasi complimentary MOSFET amplifier

cheers

Christian
  Reply With Quote
Old 23rd July 2014, 02:46 AM   #492
diyAudio Member
 
still4given's Avatar
 
Join Date: Oct 2004
Location: Victorville, CA
Quote:
Originally Posted by Ranchu32 View Post
Hi guys, I can open the Spice file but it complains about the missing pot component and these devices are missing from the model.

The R values that make up the input transistor biasing circuit don't match the schematic. For example, R12 should be 4k7, R6 is labelled R4 in my schematic, R6 in the Spice model doesn't exist in the schematic.

C10 is open circuit in the model - it should be across the gate to source.

There is no good reason why there would be any DC offset at the output once VR2 is adjusted correctly. My changes are confined to the output stage, the front end is Gareth's tried and proven design.

cheers
Hi Ranchu,

The pots need the files I attached. Maybe Bimo can set it up with resistors in place of the pots. I know how to replace VR1 but VR2 needs something more that one resistor. I had to change R4 to 1K to get enough adjustment. I know there is a way to add the devices to the file but I haven't learned that yet. Maybe Bimo or Gareth can do that for you.

I am attaching the repaired asc file. The offset is good now. THD is .04

Blessings, Terry
Attached Files
File Type: asc TGM8 MOSFET 7-22.asc (12.3 KB, 27 views)
__________________
X-BoSoZ, B1 Mezmerize, P101, Symasym, KSA 50, Aleph-X, Leach Low TIM, Leach Superamp, Dx MKIII, Super A, Honey Badger, JLH 1969, Gainclone. FetZilla, Peeceebee, VSSA PMI & jkuetemann, SKA GB150, Ovation nx & sx, AX&SX-14, TMG8, Slewmanster w/CFA-XH, Wolverine, Spooky, Symasui, Low TIM hybrid
  Reply With Quote
Old 23rd July 2014, 04:44 AM   #493
diyAudio Member
 
Join Date: Apr 2013
Thanks Terry, much appreciated.

There is no R4 on my schematic or your model. Do you mean R12?

You have uncovered an important issue I had forgotten. R6 and R11 were selected to give adequate adjustment range with the two transistor VAS. Could you please simulate the effects of reducing R6 and R12 to 1k and 3k3, respectively? It would also be useful to reduce R11 from 8k2 to 4k7 at these voltages.
  Reply With Quote
Old 23rd July 2014, 02:28 PM   #494
AndrewT is offline AndrewT  Scotland
diyAudio Member
 
Join Date: Jul 2004
Location: Scottish Borders
Quote:
Originally Posted by Bigun View Post
I would suggest you don't get hung up on the DN at this point, you can simply connect the grounds and safety earth to the chasis and try out the amplifier - then if all is well you can sort out the DN later, if needed to ward off any hum from ground loops.
yes.
The DN is an option. A direct connection often works OK.

Only fit the DN if it improves performance.
I have advised adding a switch as an option across the DN. This way you can use the switch to insert, or bypass the DN. Use whichever switch setting works better.
This is a SAFE way to wire a ground lift switch. It bypasses the DN.

Some Ground Lift switches actually break the MAG to Chassis connection, or break the PE to Chassis connection. Both of these types are potentially lethal.
__________________
regards Andrew T.
  Reply With Quote
Old 23rd July 2014, 04:22 PM   #495
diyAudio Member
 
still4given's Avatar
 
Join Date: Oct 2004
Location: Victorville, CA
Quote:
Originally Posted by Ranchu32 View Post
Thanks Terry, much appreciated.

There is no R4 on my schematic or your model. Do you mean R12?

You have uncovered an important issue I had forgotten. R6 and R11 were selected to give adequate adjustment range with the two transistor VAS. Could you please simulate the effects of reducing R6 and R12 to 1k and 3k3, respectively? It would also be useful to reduce R11 from 8k2 to 4k7 at these voltages.
You are correct, I meant R12. I changed R6, R11 and R12 per your request. You can dial out the offset but right at the end of the pots range. I'm not sure how accurate the pot models I have are. Real life may show something different but probably not something that some easy resistor changes can't take care of. Here is the harmonic profile from the error log.

Harmonic Frequency Fourier Normalized Phase Normalized
Number [Hz] Component Component [degree] Phase [deg]
1 1.000e+05 2.685e+01 1.000e+00 -9.92 0.00
2 2.000e+05 9.875e-03 3.678e-04 160.55 170.48
3 3.000e+05 3.628e-03 1.351e-04 30.42 40.34
4 4.000e+05 5.174e-03 1.927e-04 -63.53 -53.61
5 5.000e+05 2.174e-03 8.097e-05 -88.50 -78.58
6 6.000e+05 2.436e-03 9.074e-05 -69.88 -59.96
7 7.000e+05 5.801e-04 2.161e-05 -37.74 -27.82
8 8.000e+05 2.416e-03 8.998e-05 -68.04 -58.12
9 9.000e+05 8.842e-04 3.293e-05 -133.45 -123.52
Total Harmonic Distortion: 0.046379%


.fourier quantity "V(vin)" not pressent in data.
N-Period=1
Fourier components of V(vout)
DC component:-0.241438

Harmonic Frequency Fourier Normalized Phase Normalized
Number [Hz] Component Component [degree] Phase [deg]
1 1.000e+05 2.685e+01 1.000e+00 -9.92 0.00
2 2.000e+05 9.875e-03 3.678e-04 160.55 170.48
3 3.000e+05 3.628e-03 1.351e-04 30.42 40.34
4 4.000e+05 5.174e-03 1.927e-04 -63.53 -53.61
Total Harmonic Distortion: 0.043665%
Date: Wed Jul 23 08:16:20 2014
__________________
X-BoSoZ, B1 Mezmerize, P101, Symasym, KSA 50, Aleph-X, Leach Low TIM, Leach Superamp, Dx MKIII, Super A, Honey Badger, JLH 1969, Gainclone. FetZilla, Peeceebee, VSSA PMI & jkuetemann, SKA GB150, Ovation nx & sx, AX&SX-14, TMG8, Slewmanster w/CFA-XH, Wolverine, Spooky, Symasui, Low TIM hybrid
  Reply With Quote
Old 24th July 2014, 04:37 PM   #496
diyAudio Member
 
mrsavage's Avatar
 
Join Date: Nov 2011
Location: Columbus, Ohio
Default Need Gerbers files for each PCB's

PDF versions of the PCB's are fine for making the PCB's yourselves at home,
but I would like to send the PCB's to a vendor. I have made PCB's myself but
I do not like the mess and nasty chemicals.

Unless someone knows how to change PDF PCB images into Gerbers or
something that software, like Eagle, etc., can understand, then I need
Gerbers files for each PCB. Another option is Group Buys for each PCB's.
Let me know what possible. Take care.Click the image to open in full size.
  Reply With Quote
Old 1st August 2014, 01:15 AM   #497
diyAudio Member
 
Join Date: Apr 2013
Default HEXFET version

Hi guys

Since I posted the LATFET schematic at #463, I've been thinking about this a little more. One of the issues bugging me is the fact that the VAS, with its bootstrap, can only source extra drive voltage into one of the outputs. I've ensured the other rail holds up under load by placing a series diode and generous cap to the front-end; however, the VAS will not be able to crest the rail on the positive swing.

We could use separate boosted rails, but this would require an extra pair of transformer windings and power supply components; I'd rather not resort to this added cost and complexity. I can readily buy toroidals locally with 25V and 40V secondaries - but nothing in between.

What's needed is some way to bootstrap the other side. A friend sent me this link to a design illustrating precisely that: the function of its "dual bootstrap" technique is described in detail:

DIY 40W MOSFET AMP

The second concern is that LATFETs are relatively expensive and difficult to source. Vertical MOSFETs are inexpensive, abundant, and have some useful qualities, but are more difficult to thermally stabilise: bias servo design has so far eluded me.

This article from Rod Elliott and app. note has some useful information and solutions that apply to HEXFETs:

Using HEXFETs in High Fidelity Audio
http://www.irf.com/technical-info/appnotes/an-948.pdf

So here it is... any comments or suggestions? It shares much in common with Bigun's original design, including the output transistors (but only the HEXFET pair...).

By happy co-incidence I bought several of these HEXFETs last Xmas with the intention to build Bigun's high power version. Well I never got around to it and the BJT-only version sounds so good at high volumes without them that I probably never will.

R12, R13 and VR1 will probably need some tweaking to ensure thermal stability. R18 & R19 may need to be increased to stabilise oscillations at the outputs; however, it may also be possible to decrease it slightly after experimentation.
Attached Images
File Type: png TGM8-HEXFET_sch.png (21.1 KB, 113 views)

Last edited by Ranchu32; 1st August 2014 at 01:20 AM.
  Reply With Quote
Old 1st August 2014, 01:27 AM   #498
Bigun is offline Bigun  Canada
diyAudio Member
 
Bigun's Avatar
 
Join Date: Jan 2009
Blog Entries: 2
This is an excellent approach fully worthy of the TGM8 badge

Maybe try a Spice simulation to confirm parts values.
__________________
"The test of the machine is the satisfaction it gives you. There isn't any other test. If the machine produces tranquility it's right. If it disturbs you it's wrong until either the machine or your mind is changed." Robert M Pirsig.

Last edited by Bigun; 1st August 2014 at 01:30 AM.
  Reply With Quote
Old 1st August 2014, 02:06 AM   #499
AKSA is offline AKSA  Australia
diyAudio Member
 
Join Date: Sep 2001
Location: Melbourne, Australia
Quote:
bias servo design has so far eluded me
Christian, you have around 8.2V bias across the generator which you cover with a single Vbe multiplier.

This will create a very high tempco, around 8.2/0.65 times the usual Vbe junction. This is way over more than the hexfets require, and with warming up the output stage would actually shut down.

You can ameliorate the multiple by putting a couple of leds series in the emitter of the transistor. Then, if you use greens, for example, you will drop 4V across the two leds, then another 0.65 across the be junction, so the resistor you use between base and emitter is 10K (5K trimmer and 8k2) and collector to base around 8k2 (which will drop around 3.5V, with 4.65V dropped across base to emitter.

This approach, if the leds are NOT thermally bonded to the outputs, only 3.5V of the Vbe thermally connected to an output device will adjust the reduced tempco over the output devices. Choose leds that can accommodate the hefty current through VAS.

Where did you find the double bootstrap?

Cheers,

Hugh
__________________
Aspen Amplifiers P/L (Australia)
www.aksaonline.com

Last edited by AKSA; 1st August 2014 at 02:12 AM.
  Reply With Quote
Old 1st August 2014, 02:38 AM   #500
Bigun is offline Bigun  Canada
diyAudio Member
 
Bigun's Avatar
 
Join Date: Jan 2009
Blog Entries: 2
Quote:
Originally Posted by AKSA View Post
Where did you find the double bootstrap?
No Miller cap, Double bootstrapping amp.

http://www.dextrel.net/diy-40w-mosfet-amp

http://en.wikipedia.org/wiki/Bootstr...r_follower.png

and First Watt F4 has a bootstrap for the JFET front end if I remember correctly.
__________________
"The test of the machine is the satisfaction it gives you. There isn't any other test. If the machine produces tranquility it's right. If it disturbs you it's wrong until either the machine or your mind is changed." Robert M Pirsig.

Last edited by Bigun; 1st August 2014 at 02:55 AM.
  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
Rod Elliot P3A Layout - Critics paranoid06 Solid State 338 14th September 2014 05:29 PM
Fixed line level BSC based on Rod Elliot's circuit arjscott Multi-Way 10 8th November 2006 06:57 AM
Rod Elliot P3a boards stuffed FS! Mattyo5 Swap Meet 8 3rd October 2006 02:45 PM
Rod Elliot's P3A considirations Sander Solid State 14 25th September 2005 12:38 PM
Rod Elliot P3A amp noises tktran Solid State 2 18th March 2003 08:04 PM


New To Site? Need Help?

All times are GMT. The time now is 09:03 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