diyAudio (
-   Solid State (
-   -   Heavy Duty 150W Amp (

ChocoHolic 21st June 2004 08:23 PM

Heavy Duty 150W Amp
2 Attachment(s)
Hi Everybody!
Carlos pushed me a little bit... and so I decided
to publish my 150W-Oldy/Goldy for our Brazil DIY-Oldy/Goldy

The attached file shows the schematic of one chanel
of my old 8x150 power amp, which I designed in 1993.
It uses mainly standard components, is quite rugged,
does deliver 150W into 4 Ohms and might give some ispiration
to you..... hope you can read it, I had to shrink it to an inconvinient low resolution.....

Some design considerations:
The heavy duty output stage is done with 10x BDW83D
and 10x BDW84D. The current limit is simply achieved by
the emitter resistors. Each resistor is 4 Ohm, altogether
this makes 0.4 Ohms, which still allows a low output impedance.
It can easily deliver highest output currents for demanding speakers,
as there is no additional current limiting.
For thermal compensation of the output stage I used Q7 and Q8.
Q7 & Q8 are placed with isoaltion on the main heat sink close to the
output transistors. Also the ouptut transistors are isolated.
The main heat sink is connected to signal ground. C_Qxx are positioned
close to the output transistors.
Adjustment of idle current is possible by an additional resistor
across the supply of the OP amp (from pin 4 to pin 7),
I did not put this to schematic, but in fact I am using about 10k.
Today I would prefer an adjustable Vbe-multiplier.....

The output transistors are selected and matched for same Vbe.
But also without that, the emitter resistors would ensure good current sharing.
I adjusted the idle current of the output stage to 5 x 20mA, because I had to
work with low heat and comparable small heat sinks....
Please note, I run eight of this units in an 44cm x 40cm x 11cm casing.
But already with that small idle current the amp is definitely not bad!

The power supply is a traditional design with 2 torroids of 1kVA and
4x 50 000 uF caps. ...seperate casing has the same size as for the 8 amps.

The topology uses a standard NE5534 as input stage. The supply current of this
drives the VAS. Q1 and Q2 are just used for thermal compensation of the VAS.
Q1...Q6 are NOT mount on the main heat sink, but Q1...Q6 together on a second
heat sink. Q1...Q6 are mounted with isolation. Their heat sink is connected to
the output!

Feedback philosophy:
Staggered internal loops.
R7 and R8 define a current feedback, which sets the voltage ratio of the VAS
vs. the OP output to about 14.7. So every 1V output of the OP amp will give
quite linear 14.7V at the output of the VAS.
R9 together with R10 sets the gain of OP amp stage + VAS together to 391
(frequency compensation with C9).
The overall feedback sets the gain of the entire amp to about 10.
Overall feedback is only decent 32db by this, but stable up to close 20kHz.
Lead compensation in the VAS with C5 and C6 improved the phase margin.
In fact this amp is stable at any complex load.

Additional I matched the frequency response and gains of all the 8 units by carefull trimming of C7, C66 and R_Feedback. This in combination with the excellent stableness allows paralleling of the units!

As the amp's gain is only 10 it needs a large input signal, but with modern OP amps this can be achieved easily.... also an inverted input signal for bridged operation can be done easy this way....

Have fun

destroyer X 21st June 2004 11:21 PM

Wonderfull Markus, never saw schematic alike yours
Some differences i will have to reflect a lot to understand, not darlington, but alike darlington...interesting.

A golden schematic, good to share it with us... i will have fun simulating and assembling.


thank you the attention to put it here for us.


Eva 21st June 2004 11:33 PM

The schematic looks interesting but somewhat undeadable. Could you convert it to 16 color GIF? [To create GIF stuff I use an old version of Paint Shop Pro that is just fine for that purpose]

ChocoHolic 22nd June 2004 04:59 PM

2 Attachment(s)
Hi Eva!
...hope this version is better. I zipped the unshrinked *.jpg.....
By the way:
Yes, the first schematic was unreadable, but you are also right
if you say "undeable" ... ;)
Nice syntax error :D your jokes :clown:
Or was it a real syntax error?? :cool:

Yes, but that heavy battery of darlingtons is some sort of
brutal force design... :rolleyes:
Anyway, I still like it.

Hi Carlos,
the output transistors (BDW83D and BDW84D ) are normal darlingtons.
The often used sign in schematics with the doubled collector are just simplified signs. Real power darlingtons have two transistors and a reverse diode inside as shown in my schematic. Typically also two resistors between each base-emitter are integrated, too.


ChocoHolic 22nd June 2004 05:04 PM

...seems to be a crazy word...
My first "undeable" ==> undeadable !!!
:bawling: :bawling:

ChocoHolic 22nd June 2004 05:15 PM

Hi Carlos!
If you plan to assemble it, then I would propose a slightly reduced
rail voltage, because of the less stable mains voltage in your country.
Please note BDW83D are 120V types. At +/-55V they can see 110V.
If your mains starts to make funny 300V instead of 240V...
...poor darlingtons....
In Germany I dared to keep only 10% margin, we have pretty stable 230V. But in Brazil you might need more safety margin.

Or you can look for darlingtons with higher rating.
The driver stage has good margin against overvoltage. Only the output stage is running close to max. voltage.


Eva 22nd June 2004 05:34 PM

Ops, sorry, it was a typing error, I was meant to write 'unreadable'

ChocoHolic 22nd June 2004 07:34 PM

BJT Battlefield
2 Attachment(s) need to say sorry. It was a cool error :nod: ...

Attached is a picture which shows the assembly.
4 amps are placed isolated on an aluminium plate (drilled to allow
air moving)
The collectors get their rail voltage through the aluminium bars which are mount from top. These aluminium bar are also drilled top/down which allows air convection.
There is a second aluminium plate placed just mirrored downwards.
Well, it is a battlefield of BJTs....

I tend to overdo the things a little bit, but learning the limits is fine.
If you go for larger heat sinks and do not press 1.2kW into a
44.5 cm x 40 cm x 11 cm casing, then you can allow a higher idle current.... always good for lower crossover distorsions....

ChocoHolic 22nd June 2004 07:58 PM

overall view
2 Attachment(s)
Hi Carlos,
if I understood your mail right, then you were also interested in my
entire HiFi Set...

Here is a view, which shows the amp with further components in the rack. Normally the amplifier is closed, of course. For this I use a black metal grid. The fans only turn on at high power load.

The power supply is placed below.

Above the amp there is a NAD CD/DVD/MP3 Disc Changer T 571.
Sound is OK. But slow disc handling. Let's say ultra slow!!! NAD seems to think, that snails have something aristocratic. No matter which command you send to that player.... it takes centuries, until something goes on.
If you press the ON button of the remote control:
- First 2 seconds nothing happens at all !
- Then the display turns on: "Hello" (No joke!!!)
- About 15 seconds later you can manage to get some signal at the line output....
Tiring beast in this regard.
Also small disc defects and CD copy protection drives this disc changer to very long studies :rolleyes: and result in "Disc Error"....

The Receiver above is a NAD T750 Receiver.
Well after adjusting the idle current of the power amps, it started to sound OK....

Speakers are some small old Dynaudios, still enjoying them very much!


destroyer X 22nd June 2004 09:06 PM

Wonderfull, really pretty, i like, real word, like it!
Klein wunder, Small wonderfull equipment, very pretty, simetrical, wind tunnel, a lot of transistor side by side.... one of the most beatifull i have seen, want a better picture to my colection of the bests.
All details you have i want to know.... put your white face to us too. good to see people that can do so good work.

Also NAD, this is a presentation letter from you, i was not wrong, you are a gold ear man...NAD is very good to hear.



All times are GMT. The time now is 12:57 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