Modified version of Leach amplifier

This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Here is a picture of my leach amplifier. It is quite different from Mr Leach design since I wanted to build a more 'up-to-date' amplifier -- that agrees more with todays design. First I used Klass Malman's design ( you can find the link on Mr Leach website) since the transistors were on the board. I added capacitors on the board (smaller ones in parallel). The amp schematics is exactly the same as Mr Leach's 4.5 version. I only changed the layouts. I believe it's a lot easier to build my amp since you don't have much wire to solder...everything is on the board.
Here is the pictures, let me know what you think...
Don't pay attention to the wirering, this is just to test...

An externally hosted image should be here but it was not working when we last tested it.
Retired diyAudio Moderator
Joined 2002
Wow! that is a great update to the prof. leach's design. Why did you use so many big caps? What is the advantage of using series and parallel combinations instead of one big one, besides the ability to fit the design in a low profile case? How low profile case can you fit this amplifier in? It looks great. Where did you have your boards made? Where did you get the heatsinks?


EDIT: It sure was a bitch wiring the power transistors to the board, with the original design. that is a much better design!
The caps are not that big (about 2 inch high 5600uF). I had places in my case to put 8 caps per rail...I know that this is a lot but when you look in hi-fi magazines, the most expensive amps usually have around 70 000 + uF...I really don't know if it makes a difference but it was not more complicated to do it. I also read that smaller caps discharge faster, so it gives more precision than a big one...Honestly, I'm not an audiophile so I can't really tell you if it is true. I'm a hardware guy so I thought : 'just do what they say...'

I etched the board myself and they look pretty good I think. It's a lot cheaper and you get the same result.

I made the heatsink myself...Actually this was a very big heatsink used in a truck or something...I cut it to fit the amp.

As for the case, I'm currently working on it. It should be not more than 2¾ high.
I'll set up a webpage in the near future with all the layouts, part lists and more pictures...
Retired diyAudio Moderator
Joined 2002
Sounds good. So, you have all of the caps in parallel, giving you 44800uF per rail? The boards look great. Can you take a picture of the bottom of the boards? This looks like a great design.

So, etching wasn't that bad? What did it entail? Do you have any links to pages describing the process?


Retired diyAudio Moderator
Joined 2002
It would be great to modify this design for the Leach SuperAmp, it would not be too bad, except for the placement of the output transistors, since there is 8 instead of 4. Does anyone think this is possible? Could you get away with having a double row of output transistors on the end of the board and just having a larger heatsink?

Etching is quite easy. I use MG chemicals products. It takes no more than 30 min overall for both boards. Drilling is a bit more tedious, and you need something like a Dremel to do a good job...takes about 2hours overall.
Here is a link to mgchemicals webpage that explains the process.

As I said, I'll make a website when the amp is finished with more pictures. If anyone really needs the layout now, just email me. I use Orcad Layout.

I was not planning to modifie the Leach super amp. I don't know what it would involve... If anyone tell me that it makes a very good subwoofer amp, I'll do it guaranteed. I still have a lot of spare parts so I could reuse them.
Leach's Superamp should make a GREAT subwoofer amp. :)

Really, it should. Though I think by pushing the rail voltages
to 75 volts and using three or four output pairs, Leach's lower
power amplifier design should work just as well. You could use
MJL21193/4 devices (plastic case) and simplify heat sink design
by not having to drill so many holes for TO-3 devices.

The Superamp is configured to safely use lower voltage output
devices by placing them in series/parallel so that current is
shared and voltage halved across each device. This makes for
a slightly more complex design, including additional driver
transistors, but that was the safest way to get that much power
out of existing transistors at that time.

At any rate, eliminating the extra wiring and socket contacts
as you've done, is always A Good Thing. Should help stability
and get rid of some problematic ohmic contacts.

I run with 30,000 uF/rail on my dual-mono chassis because I like
LOTS of capacitance to lower ripple and ESR. I should add that
I found it advisable to include a surge resistor in the AC line; may
have saved me a bridge rectifier or two over the years. I also
added 500 ohm/rail bleeder resistors for safety, and to ensure
symmetrical discharge, which seemed to help with a stability
problem I had with the older driver boards after the amplifier
was turned off. The ver. 4.5 boards don't seem to have that


Nice set of amps! Having built a 3 channel version, I can say this sure beats the chore of connecting 14 cables per board! Please set up your pages with the layouts etc ASAP! Is it possible for you to change the layout for the output devices so that Toshiba 2SA1943/2SC5200 devices can be used? Thanks in advance!
Hi Damon Hill!
Are you saying that i'd get a better subwoofer amp by simply add ouput transistors to my design than to go through the chore of redoing the entire layout for the superleach amp?

Bawang :

It might take a while before I set up the web page...I'm doing hard classes this semester so it might not be before the end of april...Can you tell me what is the advantage of using Toshiba transistors?

Thanks a lot

It should be possible to drive three pairs in parallel with Leach's
lower-power amplifier, and it's simpler to implement on your
layout than to redo his series-paralleled Superamp. Well, I know
it is because I've worked with a commercial version of Leach's
design that was done this way. With the MJ21193/4, it would
be a hardy design although I don't know how far it can be pushed
with MJE15030/31 driver transistors. They may run out of
headroom into very low impedance loads and much more than
75 volt rails.

As Audiofreak points out, the Toshiba transistors are more
linear and in parallel should have plenty of margin, but I won't
guarantee stability without testing first. They're much faster
and may oscillate more easily without some modifications. I'd
certainly like the devices to solder directly to the driver board;
it might help stability.

I've got this great big 'wind tunnel' heat sink that's hard-wired
for five pairs of output devices; with the MJ21193/4, it should be
good for at least 500 watts per side into 8 ohms; I think it would
do a kilowatt into 4 ohms but I'm certain I'll have to use TO-3
case drivers. I can't be sure until I've actually built a prototype
and tested at those power levels. In that case, I need a board
with just the first two stages and predriver stage.
Layouts are in Orcad Layout 9.0
I'm pretty sure you can export them in Protel and almost any other pcb layout program.
As I said, if anyone is interested in having the layouts right now, contact me via email as it might take a couple of months before I set up the webpage.
For those who thinks Orcad is too expensive (9000$), have a look at :
http column // orcrack dot cjb dot net
Don't wait too long, new versions will come out soon.
Have fun
Hi Bawang:
I'll try to print and scan the layout into a jpeg file...I think it's the only way to convert an Orcad layout file into an image of correct size. I should be able to work on this next week.

Vuamp2, my design is inpired from the one you submitted. Actually, I used Klas Malman's design (he was kind enough to send it to me) and I modified it. I didn't like the triangular shape and I thought that capacitors would fit well in the corners. I added fuses btwn capacitors and power transitors and also lighten up the design since I thought some traces were so close to eachother that they might be hard to produce with the 'home etching process'

Oh and by the way, I'm sorry for all of you europeen and american readers...
Canada will win the hockey gold medal in Salt lake...

This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.