DIY ACA mini

My version of the Mini is up & running and playing music :) Big thanks Mr. Pass & everyone involved in making this fun and cool project available to all of us!
I always like to customize so I made my own pcb that can be fixed to a single bigger heatsink. Plan is to put it in a square(ish) case to protect it from the small kids and cats that are never far away...
If anybody wants the kicad files and/or gerbers, let me know, happy to share them
Hi Bro, can you share your kicad/gerber files? happy that you can send to me by email [*Redacted]. Thank you very much.
 
Last edited by a moderator:
  • Like
Reactions: 1 user
Firstly, I am by no means an experienced guy, just like to build some small stuff and have a good listen. So I maybe ask stupid questions...

So, I recently built a ACA Mini and like it very much, so much so that I decided to build another one in a case with heatsinks to be able to crank current up a bit.
So far so good, here is the current state...
IMG_20230122_144816_edit_60082030070518.jpg

Then I realized that the whole circuit is grounded through the 4 mounting holes, so the ground of inputs and speaker connection is connected to the case now.
This doesn't feel right...
Also because I want to use this power supply, which is also grounded by the mounting screws by the famous yellow/green wire.
Question: is it ok to mount the amp board with nylon screws for isolation, and connect the supply as on the picture, so the case is connected to safety ground by it?

IMG_20230122_145033_edit_60218944481956.jpg


And then...the supply can put out 26V, is it ok to run the amp like that, also with some more current?
 
Administrator
Joined 2007
Paid Member
It may run to hot. 24 to 26 volt may sound like a very small increase in voltage but because power (dissipated) has voltage squared as a term in the equation it makes a big difference.

You could see getting on for 20% more dissipation per device even if you don't alter anything current wise. Somewhere in the region of around 18.5 watts to 22 watts per transistor. That makes a big difference to the heat.

Give it just a 100ma or so more bias and that figure jumps even higher to around 27% increase per device.
 
@Mooly: thanks for those figures, I'll increase the current carefully then, but those sinks are pretty big compared to the small Aavid ones so there should be some headroom. Just don't want to destroy it:)
So the active devices can take 26V I assume then? Want to squeeze out a little more out of it.
@AJedi: they are in place, I moved them upright to the capacitor spot, and connected the caps to the resistor spot. Otherwise it's difficult to use bigger input caps. It's even a boutique Audio Note tantalum, which I really like for input duties.
 
  • Like
Reactions: 1 user
Administrator
Joined 2007
Paid Member
You could see getting on for 20% more dissipation per device even if you don't alter anything current wise. Somewhere in the region of around 18.5 watts to 22 watts per transistor. That makes a big difference to the heat.

I better correct that :) I was thinking of the standard ACA...

Same applies but the numbers are very different. You would see around a 10 to 12% increase in dissipation per device by increasing the supply and correspondingly more if you increase the bias as well.

Sorry about that...
 
Administrator
Joined 2007
Paid Member
thanks for those figures, I'll increase the current carefully then, but those sinks are pretty big compared to the small Aavid ones so there should be some headroom. Just don't want to destroy it

Heat is the enemy, the transistors themselves can handle far far more than they see here... as long as they are not running to hot.

You can easily put some numbers in to see how big things get for small increases.

If you have 12 volts across the transistor and it passes 400ma then you have 4.8 watts of heat for that transistor.

Increase the supply so you have 13 volts across it and you also up the bias from 400 to 500ma and you now have 6.5 watts of heat. That is per device and is nearly 36% of an increase.
 
Member
Joined 2014
Paid Member
Nylon standoffs and a ground lift should work well. CL-60 would be easy.

So I built a pair of dual mono ACA Mini boards up with a dual'ish mono power supply. Although I only had 24 volts, I cranked the bias up to something like .450 per device and with the large heatsinks, the MOSFETs didn't even get above say 35-40 Celcius with an ambient of around 17 Celcius. That was with the heatsinks faced down on my test bench. So crank her up and use the back of one of your finger knuckles to get a sense of how warm it is

This amp does not dissipate a lot of heat being push-pull and only 5 watts stock. In fact, I decided to make mine a nice summer amp using the large heatsinks for my workshop becuase of these results :)
 
  • Like
Reactions: 1 user
An even lower-hanging fruit is to space the components apart from the PCB (thus the soldermask) by the thickness of a popsicle stick. Apply stick, stuff component, solder leads, remove stick. Voila an air-gap 2mm wide. Air is an excellent insulator. Or, leave the popsicle stick in place. Wood is a pretty good insulator too.

Just in case "popsicle stick" is accidentally a regional idiom instead of a universal expression, here's an illustration.

_
And a tasty treat to be consumed before you begin soldering!
 
  • Like
Reactions: 1 users
Nylon standoffs and a ground lift should work well. CL-60 would be easy.

So I built a pair of dual mono ACA Mini boards up with a dual'ish mono power supply. Although I only had 24 volts, I cranked the bias up to something like .450 per device and with the large heatsinks, the MOSFETs didn't even get above say 35-40 Celcius with an ambient of around 17 Celcius. :)
VB at 0.35V that are 0,46A is the standard setting on guide ,for winter :)