Kimbal-
I was thinking about your "hot" heatsink idea. Assuming the collectors are all at he same potential it will work, however aluminum forms an insulating oxide film immediately on contact with the air. A number of years ago, I built a light controller for the lights on the sides of a huge ferris wheel. There were something like 120 25 amp triacs involved. We thought we'd save some time and money by sandblasting the anodic coating off the heatsinks for direct contact.
Well, it worked for a while, but after a few months, the thing started blowing triacs like crazy. They were overheating because of the resistance of the oxide build-up on the aluminum. We ended up remounting all of the triacs with tinned copper lugs. Never had a problem after that.
So, I would suggest that if you use this approach, you consider mounting the transistors on a copper plate that is thermally coupled to the Al heatsinks. If you really want to get fancy, braze a copper tube to the plate and circulate a glycol/water solution through it to an external radiator. I once built a heatsink set for a guy who was building a 150 watt class A amp and didn't want massive heatsinks.
Be warned, however that copper is a real bitch to machine. It binds drill bits and taps alike- you have to work slowly with a lot of lubricant/coolant. If your tap starts to bind, you're screwed- you have to back the tap out before it binds. It's like 1/4 turn at a time, backing the tap out and cleaning the shavings out of the flutes completely. One tapped hole can take 10-15 minutes through a 1/4" plate.
Cheers,
J
I was thinking about your "hot" heatsink idea. Assuming the collectors are all at he same potential it will work, however aluminum forms an insulating oxide film immediately on contact with the air. A number of years ago, I built a light controller for the lights on the sides of a huge ferris wheel. There were something like 120 25 amp triacs involved. We thought we'd save some time and money by sandblasting the anodic coating off the heatsinks for direct contact.
Well, it worked for a while, but after a few months, the thing started blowing triacs like crazy. They were overheating because of the resistance of the oxide build-up on the aluminum. We ended up remounting all of the triacs with tinned copper lugs. Never had a problem after that.
So, I would suggest that if you use this approach, you consider mounting the transistors on a copper plate that is thermally coupled to the Al heatsinks. If you really want to get fancy, braze a copper tube to the plate and circulate a glycol/water solution through it to an external radiator. I once built a heatsink set for a guy who was building a 150 watt class A amp and didn't want massive heatsinks.
Be warned, however that copper is a real bitch to machine. It binds drill bits and taps alike- you have to work slowly with a lot of lubricant/coolant. If your tap starts to bind, you're screwed- you have to back the tap out before it binds. It's like 1/4 turn at a time, backing the tap out and cleaning the shavings out of the flutes completely. One tapped hole can take 10-15 minutes through a 1/4" plate.
Cheers,
J
Last edited:
Avoiding Heatsink Corrosion.
Agree with your comments - but as for the Ally oxidisation that can be minimized with maybe clear Silicon grease under the devices, and making sure the thermal contact faces are Milled before assembly - OR - use a cast Ally heatsink which you can purchase. And yes cast Ally is not my favourite approach either but it works well for most applications and looks professional.
Using copper - it will tarnish again much the same way as Ally does. Maybe even look at plating it with gold, or some other thermally efficient material - if your rich - this would need some investigation. Rather than tapp the copper use blind nuts on one side, with suitably sized holes and avoid the whole tapping exercise. Again, make sure the thermal contact faces are milled.
I also hate tapping Ally as it also has its own difficulties.
The whole thing with copper, brass and Ally is they all oxidize to various degrees depending on grade and additives etc. A high level of surface finishing may reduce this and the microscopic pores are smaller for oxide build up, but it's not a guaranteed solution since oxygen still attacks the surface, as you well know. Gee - what about immersing the whole assembly on Transformer oil, or just Nitrogen gas - Now its becoming a science lab project !
Oil cooled - other than Secondary Breakdown, it would be thermally very efficient. Heat pipes yes can be done, but a bit of a mechanical plumbing exercise. I like to keep construction simple where possible.
Painting, plating or sealing them in some way is the only way to reduce this - UNLESS you use maybe an electrical process attached to the heatsink, that "kills" oxidization much like a anti-rust preventer in a car, you buy from Auto shops. Such a device could be built into the Power supply and wired straight to the heat sink.
I have also toyed with the idea of using 312 food grade Stainless but this may posses other technical problems - such as thermal efficiency for one, and getting the desired shape for a second, and then cost for a third. As I said its just and idea.
Agree with your comments - but as for the Ally oxidisation that can be minimized with maybe clear Silicon grease under the devices, and making sure the thermal contact faces are Milled before assembly - OR - use a cast Ally heatsink which you can purchase. And yes cast Ally is not my favourite approach either but it works well for most applications and looks professional.
Using copper - it will tarnish again much the same way as Ally does. Maybe even look at plating it with gold, or some other thermally efficient material - if your rich - this would need some investigation. Rather than tapp the copper use blind nuts on one side, with suitably sized holes and avoid the whole tapping exercise. Again, make sure the thermal contact faces are milled.
I also hate tapping Ally as it also has its own difficulties.
The whole thing with copper, brass and Ally is they all oxidize to various degrees depending on grade and additives etc. A high level of surface finishing may reduce this and the microscopic pores are smaller for oxide build up, but it's not a guaranteed solution since oxygen still attacks the surface, as you well know. Gee - what about immersing the whole assembly on Transformer oil, or just Nitrogen gas - Now its becoming a science lab project !
Oil cooled - other than Secondary Breakdown, it would be thermally very efficient. Heat pipes yes can be done, but a bit of a mechanical plumbing exercise. I like to keep construction simple where possible.
Painting, plating or sealing them in some way is the only way to reduce this - UNLESS you use maybe an electrical process attached to the heatsink, that "kills" oxidization much like a anti-rust preventer in a car, you buy from Auto shops. Such a device could be built into the Power supply and wired straight to the heat sink.
I have also toyed with the idea of using 312 food grade Stainless but this may posses other technical problems - such as thermal efficiency for one, and getting the desired shape for a second, and then cost for a third. As I said its just and idea.
"Hot" heatsinks require two, for each ±V.
Grease must be used, and direct connections to the device collectors must be made.
Most of the Crown amplifiers are made this way (Microtech, Macrotech, Powertech, Powerbase, Comtech).
Early McIntosh solid-state amplifiers were made this way (MC2300, others).
Grease must be used, and direct connections to the device collectors must be made.
Most of the Crown amplifiers are made this way (Microtech, Macrotech, Powertech, Powerbase, Comtech).
Early McIntosh solid-state amplifiers were made this way (MC2300, others).
Agreed djk - two big hot heatsinks which cannot be external to the case.
That means fan forced cooling to keep up the efficiency or a fancy caged cabinet for natural air flow to get in.
I do recall another way as done with the Australian designed ETI-480 ( from Electronics Today Magazine - 1976 ) output stage, where the emitters of the 4 output devices were at the supply rails and collectors at Vout potential of the speaker and all being common to each other ( no Dc on the heatsink just an Ac audio component), if you did away with the insulators on the TO3 devices.
Again the heatsink is at audio potential and is live, but not as nasty as the first idea spoken about.
If this arrangement was adopted one would have to totally redesign the Leach Output stage to suit.
Ok your right - my fault for getting side tracked.
Ok - so before I say a definite yes - what is the approx wait time for these, I will have to build it into my budget.
Price was $14.99 each USD just for plain, no frills, drilled Fibreglass PCBs in 3 oz copper ?
And to confirm - Are these the Leach boards that require hard wired output transistors ( such as TO3 devices )?
If so I would like TEN ( 10 ).
Do you want a paypal deposit before ordering, and if so, how much ?
Last edited:
The turnaround is 24 hrs. To get the price down to $13;99 ea, 30 boards have to be ordered. We have 8 committed. The wait is going to be dependent on how long it takes to find homes for the remaining boards.
If the order is 20 units, the price is $17.33, same turnaround time.
So, is there anyone else interested?
If the order is 20 units, the price is $17.33, same turnaround time.
So, is there anyone else interested?
Regrettably I have just encountered a situation that makes it finacially impossible to go ahead with this buy. I apologize for getting hopes up, but my hands are tied at this point.
If another of you wish to pursue this, you can use the link that Damon Hill provided in Post #19 of this thread to down load the zipped Gerber files. These may be e-mailed to any competent board house for quotes and/or production.
If you can only get 10 boards committed, they shouldn't be more than $25-26 each. Just depends on how badly you want to build the amp.
The Gerber files also have the silkscreen and soldermask layers imbedded, but these will add to the board cost considerably because of the addtional set-up charges.
Again, Very sorry.
Jay
If another of you wish to pursue this, you can use the link that Damon Hill provided in Post #19 of this thread to down load the zipped Gerber files. These may be e-mailed to any competent board house for quotes and/or production.
If you can only get 10 boards committed, they shouldn't be more than $25-26 each. Just depends on how badly you want to build the amp.
The Gerber files also have the silkscreen and soldermask layers imbedded, but these will add to the board cost considerably because of the addtional set-up charges.
Again, Very sorry.
Jay
Whoever decides to run a group buy should set it up in the group buy forum not here in the solid state forum.. Title it properly (specific version of the amplifier the PCB is for) and you should have little difficulty in getting enough interested members to proceed.
For those who might be interested - I received my original Leach ver 4.5 boards ( 4 dark green fibreglass boards ) today in the mail. I was very pleased with them, considering the cost and difficulty in sourcing them. They have on the copper side - LEACH AMP 4.5 - printed in one corner along the far edge with a small oval logo saying - EEC 5.0 - on the opposite side in the middle.
I have 4 original boards with no silk screen printing, but they have solder masking and tinning on the copper side. So the Gerber files should carry at least that information for those wanting to get such produced.
I also have one only Super Leach Amp PCB ( which might make a good workbench Lab Amp monoblock ); which has Solder Masking, but no tinning on the copper side.
It does have silk screen printing for components on both component side and copper side of the board. Why it has silk screen printing on both sides, I see no point.
I don't think the Super Leach Amp board is fibreglass; but the earlier much cheaper compressed light brown paxilon material, as used during the 1960's and 70's before fibreglass became popular.
The Super Leach Amp has - WML SUPERAMP - printed along the far edge of the copper side, along with component silk screen printing.
After assessing the boards with and without silk screen printing - unless the silk screen printing is very neat and sharp, and in a colour that stands out form the background board; I really prefer the boards with no silk printing. The silk screen printing makes the board look far too busy to be of much benefit. A better quality print could rectify this issue. My board print was done in black ink and though it is quite readable the definition could be much better.
I have 4 original boards with no silk screen printing, but they have solder masking and tinning on the copper side. So the Gerber files should carry at least that information for those wanting to get such produced.
I also have one only Super Leach Amp PCB ( which might make a good workbench Lab Amp monoblock ); which has Solder Masking, but no tinning on the copper side.
It does have silk screen printing for components on both component side and copper side of the board. Why it has silk screen printing on both sides, I see no point.
I don't think the Super Leach Amp board is fibreglass; but the earlier much cheaper compressed light brown paxilon material, as used during the 1960's and 70's before fibreglass became popular.
The Super Leach Amp has - WML SUPERAMP - printed along the far edge of the copper side, along with component silk screen printing.
After assessing the boards with and without silk screen printing - unless the silk screen printing is very neat and sharp, and in a colour that stands out form the background board; I really prefer the boards with no silk printing. The silk screen printing makes the board look far too busy to be of much benefit. A better quality print could rectify this issue. My board print was done in black ink and though it is quite readable the definition could be much better.
Last edited:
I'm interested in four (4) original Leach Low-TM Version 4.5 Power Amplifier
PCB's.
If you can't find any original PCBs you may be able to copy the artwork from the Leach website and get them made up.
If you do, make and number of them and sell off the rest to offset your costs.
There will be buyers out there I'm sure. The originals cost about $10-12 USD each as far as I know. Leach had them made up before he passed away some years ago. for me to get the ones I had I was just Lucky as I haven't come across any others as yet.
Link to GB for bare boards of a Leach Low-TIM clone based on Jens Rasmussen's last published schematic (which in turn is a modified version of the v4.5 schematic), with 2 additional BJT pairs (8 + 8 total):
http://www.diyaudio.com/forums/grou...tim-8-8-bjt-outputs-bare-pcb.html#post3822518
http://www.diyaudio.com/forums/grou...tim-8-8-bjt-outputs-bare-pcb.html#post3822518
Leach Amp 4.5 original PCBs
I have 4 unused v4.5 PCBs still in packing that Prof Leach used when he sent them to me many years ago. I never built the amps obviously. Anyone interested?
I am new to this Forum and probably will not check it that often but I assume it will send me an email when some responds.
Regards, David
I have 4 unused v4.5 PCBs still in packing that Prof Leach used when he sent them to me many years ago. I never built the amps obviously. Anyone interested?
I am new to this Forum and probably will not check it that often but I assume it will send me an email when some responds.
Regards, David
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.