4-part Harmony, a 9 watt class A amp using a voltage regulator IC. - Page 3 - 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 20th March 2003, 02:15 AM   #21
diyAudio Member
 
mrfeedback's Avatar
 
Join Date: Feb 2002
Location: Perth, Australia.
Default Backhanded Encouragement

Graham,
When we say 'your crazy ideas', we mean that in the nicest possible way, of course.

Indeed it is very interesting to hear of your experiments.
Your expose on modifying MWO transformers was excellent - I have thought I using these myself, but had discarded the idea because I thought that they were permanently welded shut, but your closer observation reveals that they can be hacked.
I kick myself for not taking a close enough look.

Eric.
__________________
I believe not to believe in any fixed belief system.
  Reply With Quote
Old 21st March 2003, 08:36 AM   #22
diyAudio Member
 
Circlotron's Avatar
 
Join Date: Jun 2002
Location: Melbourne, Australia
Default Stuff and anti-stuff.

Complementary symmetry this time. As with the others, no need for global feedback because the output of the regs track the input within a fraction of a millivolt. Voltage gain is x1. Set bias by changing 2R4 resistor. Has constant 2.5 volts across it. Currently = 1 amp so operates in class A. Any oscillations can be damped by putting a 100nF between adj and out pins of each reg.
Attached Images
File Type: gif lm317-337 comp symm.gif (4.5 KB, 1215 views)
__________________
Best-ever T/S parameter spreadsheet.
http://www.diyaudio.com/forums/multi...tml#post353269
  Reply With Quote
Old 21st March 2003, 09:22 AM   #23
e96mlo is offline e96mlo  Sweden
diyAudio Member
 
Join Date: Apr 2002
Location: Helsingborg, Sweden
How does the 1,25 V ref. voltage behave compared to the 0,7 V base-emitter voltage on a BJT when talking about biasing?

I'm asking because to me it would be more interesting to run it in class AB.

/Marcus
  Reply With Quote
Old 21st March 2003, 11:08 AM   #24
diyAudio Member
 
Circlotron's Avatar
 
Join Date: Jun 2002
Location: Melbourne, Australia
If you compare the the '317 with an NPN, the base is 0.7v higher than the emitter, with the '37 the ADJ is 1.25v lower than the OUT. This 1.25v stays put come hell or high water though. No temp compensation necessary. Yep, class AB would be nifty, but with the cct I drew if you raise the resistor value the currents in the two output coupling capacitors would become assymetrical and the voltage across them would change and so would the bias. What it needs is some sort of constant voltage drop thingy between the OUT's instead of a resistor, something like a couple of forward biased diodes. Actually a *way* better setup would be to join the two OUT's, get rid of the resistor, and use just one output coupling cap. Separate the two ADJ's from each other and make the upper one lower by 1.25 volts, and the lower one higher by 1.25 volts. Capacitively couple them, bias them to half rail (well, the imaginary midpoint between them to half rail seeing they are 2.5 volts apart) and feed signal in as before. Only trouble is, unlike a BJT output stage where the current gradually comes on with bias, with this one the current comes on most suddenly once you exceed that 1.25 / 2.5 volts bias. Only way out that I see is to put a 0.1 ohm or so resistor between each OUT and output coupling cap just as in a conventional BJT output stage. But then the output impedance rises to 0.1 ohms, and would have to have a feedback loop to lower it if you really wanted to. In reality though it would probably be quite ok this way. I'll have to give it a go too because I really want to try class AB too. One you are running a bare minimum of current I can't imagine crossover distortion would be an issue at all courtesy of that 80dB feedback inside the regs.
__________________
Best-ever T/S parameter spreadsheet.
http://www.diyaudio.com/forums/multi...tml#post353269
  Reply With Quote
Old 21st March 2003, 11:33 AM   #25
e96mlo is offline e96mlo  Sweden
diyAudio Member
 
Join Date: Apr 2002
Location: Helsingborg, Sweden
Maybe this would work then...

/Marcus
Attached Images
File Type: jpg 317amp.jpg (23.9 KB, 1120 views)
  Reply With Quote
Old 21st March 2003, 11:42 AM   #26
diyAudio Member
 
Join Date: Sep 2002
Location: south yorkshire, UK
I am going to make a complementary symmetry buffer using an LM317 / LM337 combination

Ok, how about applying the same idea to a pair of switching regulators - like the 'simple switchers' to give us a nice hefty 12v powered monster amp?

hmmm.... it's set me thinking.....

ray
  Reply With Quote
Old 22nd March 2003, 08:23 AM   #27
diyAudio Member
 
Circlotron's Avatar
 
Join Date: Jun 2002
Location: Melbourne, Australia
Quote:
Originally posted by e96mlo
Maybe this would work then...

/Marcus
Hah! I ended up trying almost exacty that! I used a TL431 and a pot instead of a transistor, but it's the same principle. Definitely had to use a 10R in series with 100nF across the speaker to stop it oscillating sometimes. Also a 100nF from adj to out on the '317 only. At low levels I could get it run at 15mv p/p output with 10mA bias and zero xover distortion. At much higher levels the waveform would get a pair of notches in it unless you wound the bias up to about 50mA. It was a bit of a pig for oscillation sometimes though. My favourite has been the Tetrahedron version. That was very well behaved, that one.

Ray, the switching one sounds cool!
__________________
Best-ever T/S parameter spreadsheet.
http://www.diyaudio.com/forums/multi...tml#post353269
  Reply With Quote
Old 29th March 2003, 09:40 AM   #28
diyAudio Member
 
Circlotron's Avatar
 
Join Date: Jun 2002
Location: Melbourne, Australia
Default 4PH on steroids.

Righto, here it is. The original cct beefed up for some real power. The LM338 has a 5 amp continuous 12 amp peak rating. The fet above the reg bootstraps the supply to the reg to minimise the dissipation in it and also allows way higher supply rails and therefore output swings. In theory, if you put in a 1200v IGBT.... The fet may need a bit of capacitance between gate and source in case it wants to oscillate. I haven't actually built this one up to find out how it goes but you could probably get about 50-70 watts into 8 ohms with a 40v supply. To get the quiescent current set right, if it turns out the choke doesn't have exactly the right resistance, just raise or lower the bottom end of the 4k7 input resistor to get the bias current right. Again, I must stress I haven't built this one, therefore leaving it up to you to 'boldly go where no diyer has gone before'.
The reg is an LM338, NOT LM358 like the drawing shows!
Attached Images
File Type: gif 4 part harmony beefed up.gif (4.8 KB, 927 views)
__________________
Best-ever T/S parameter spreadsheet.
http://www.diyaudio.com/forums/multi...tml#post353269
  Reply With Quote
Old 30th March 2003, 09:27 AM   #29
diyAudio Member
 
Circlotron's Avatar
 
Join Date: Jun 2002
Location: Melbourne, Australia
Default more more more...

Time to crank up the wick a bit. Instead of going to a fancier reg for higher power output as per the previous cct, another way to increase the current capacity is to add a PNP transistor to the reg (back to an LM317 again) so the combination functions as a complementary feedback pair, or Sziklai pair. Actually, if you turn this part of the cct horizontal you will probably recognise it as a standard technique for getting higher output from a standard three terminal reg. The current is now only limited by the capacity of the PNP. The '317 reg can manage about 1.5 amps base drive so that should be enough for most situations. The good part about this cct is that the PNP only has about 6 volts across it so no matter how much current you pull through it, there is no worries about second breakdown. Also, you can use quite a slow transistor because the voltage across it changes very little unless you swing to within 6 volts of the positive rail. Quiescent current should be set to supply volts / load resistance. E.g. 32 volts / 8 ohms = 4 amps. One way to do this is to add a bit of voltage to the bottom of the input resistor as in the previous cct. Of course 32v x 4 amps equals 128 watts dissipation, most of which will go in the mosfet, but also the theoretical power output will be 64 watts. In practice a little lower.

Reality strikes! I had the thing running on 40v at about 2.5 amps and I was getting 18.75 watts into 6 ohms and I turned it off to set up some fans on the heatsink so I could wind up the quiescent current further, and when I turned back it on... foof! orange flames and black smoke leaked out of the fet. Actually the flames shot out about 30mm where the legs go into the plastic body. But other than that it worked quite well. I decided to give up on it for the night and come inside.

/Circlotron - remembers the old saying - "With power electronics you are always only a microsecond away from misery". Well, with high voltage switchmode type power electronics at least.

P.S. The IRF540 is a bit to small if you want to get *real* power out of this thing. Seeing it dissipates most of the power, the wattage rating should be at least supply volts x quiescent current.
Attached Images
File Type: gif 4 part harmony beefed up with cfp.gif (5.2 KB, 855 views)
__________________
Best-ever T/S parameter spreadsheet.
http://www.diyaudio.com/forums/multi...tml#post353269
  Reply With Quote
Old 30th April 2003, 02:37 PM   #30
diyAudio Member
 
Circlotron's Avatar
 
Join Date: Jun 2002
Location: Melbourne, Australia
Default 4PH with gain

All of the previous cct have been buffers, i.e. whatever voltage swing you wanted across the speaker you had to provide at the amp input, albeit at a higher impedance. This one makes more use of the '317' s internal gain. The quiescent current is set by 1.25 volts divided by (in this case) 1.2 ohms for about 1 amp. 14 volts supply is car battery voltage. With no signal output there is a slight hiss if you put your ear right up to the speaker so it would be no good as a headphone amp unless you used an attenuator between the amp output and the 'phones. Also, you may need a 10 ohm resistor in series with a 100nF cap across the inductor just to make sure it is stable in all circumstances. It worked ok without it but I'd feel more comfortable with it. Anyway, voltage gain is now load resistance divided by bottom resistor, so 1 volt peak will drive it just nicely. That's fairly average for most power amplifiers. I fired it up briefly tonight and fed it from the PC sound card speaker output and it sounded quite respectable considering what it is made from. I'm going to put together a second one and bring it inside and put it on the proper system and see what it is like.

So there you go. Probably the worlds simplest low distortion thermally stable true class A power amp using what really amounts to a single ended power amplifier chip. Not in the the same league as the JLH class A, but besides the choke it costs next to nothing and takes about 5 minutes to wire it up point to point. Next I'll have to give a report on the sound.
Attached Images
File Type: gif 4 part harmony with gain.gif (5.3 KB, 729 views)
__________________
Best-ever T/S parameter spreadsheet.
http://www.diyaudio.com/forums/multi...tml#post353269
  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
JLH 10 Watt class A amplifier tschrama Solid State 3019 19th June 2014 11:57 PM
10,000 watt class D amp BillyDoc Class D 19 8th May 2009 06:58 PM
Re: JLH 10 Watt class A amplifier DomKo Solid State 16 13th February 2004 06:27 AM
Eight Part Harmony: The Tao Headphone Buffer Steve Eddy Headphone Systems 4 13th November 2003 06:19 PM


New To Site? Need Help?

All times are GMT. The time now is 07:50 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