Heavy Metal Amplifier

[Ray P]

Just for a change I am starting on a build of a solid state amplifier - low power, class A, using Tokin SITs.

I was going to build a simple follower but have changed course and will buy a kit of parts for one of DIY Audio member Zen Mod's designs.

I'll need to dissipate around 80W per channel but managed to pick up some big heatsinks from a local solar panel installer, recycled, I believe, from dead inverters. Today I started to put together a chassis using two of the heatsinks, which are spaced apart by two 150mm wide 'U' channel sections. The idea is to mount the amp boards on the front 'U' channel section and the transformers/power supply on the rear one. The main active devices will be mounted on the sides.

I still have lots of metalwork to do but am really pleased with this afternoon's efforts, which involved drilling and tapping 20 M5 blind holes in the heatsinks - all done successfully! I'm learning as i was meticulous in drilling holes accurately and vertically and have invested in a decent tap guide.

Progress so far...

[IslandPink]

Mmm - nice starting point !

[Cressy Snr]

Cool, or should that be hot?

[Ray P]

I've just ordered some set/grub screws of the correct length to bolt it together properly.

Once I finish the metalwork I plan to get it all anodised black (for improved heat dissipation) and I'm waiting for prices from some local businesses.

I plan to do the front panel in wood, if possible, shaped to continue the curve of the heatsinks - I would love to use Olive again but I don't think I'll be able to get a piece thick enough for the curve.

[Dave the bass]

I plan to do the front panel in wood

Ah! 'Kindling' I think its called on a large SS Class A amp

[RhythMick]

Christ on a bike that's a big mofo heatsink !

[Ray P]

Christ on a bike that's a big mofo heatsink !

This one's not a MoFo, I've decided it's going to be a SissySIT.

Each channel needs to dissipate about 80W.

[Ray P]

Lots more metalwork again today, the chassis is coming together nicely.

Drilled and tapped another 20 or so holes today - that's around 40 all in and not a single failure (hope you're reading Phil) - the tap guide I bought was pricey but worth every penny. Some more pictures tomorrow.

[Ant]

I did some tapping yesterday, bought an angle plate to bolt my small proxxon dividing head to in order to be able to use it on the new milling machine. Didnt want to have to buy a new one. had to drill and tap blind holes in it to bolt the head on to use it stood upright.
Since i got it, ive always put my taps in my drill press chuck, then held the handle down while hand turning the chuck to tap bang straight threads.
Did it using the collet chuck in the milling machine this time, same way by hand turning
I hate using hand tap wrenches, a guide is a must.
I made a set of different sized guides which i still have somewhere

[pre65]

Lots more metalwork again today, the chassis is coming together nicely.

Drilled and tapped another 20 or so holes today - that's around 40 all in and not a single failure (hope you're reading Phil)

I'm glad to hear that Ray.

[Nick]

Out of interest, where is the active device going to go on those heatsinks?

[Ray P]

As close to the centre of the heatsink as I can reasonably get them - I'll try and post a picture or two tomorrow

[Nick]

On the u sections or actually on the heatsinks? Just wondering what all the tapping and fixing was for.

[andrew Ivimey]

Maybe he just likes tapping and fixing....

[Ray P]

On the u sections or actually on the heatsinks? Just wondering what all the tapping and fixing was for.

Sorry, I misunderstood...

Ideally I will bolt the active devices to the sides of the U channel, which is why half (20) of the taps are just to bolt the U sections to the heatsinks - I reason that for good thermal efficiency there needs to be good consistent contact between the channels and the heatsinks, plus I'll use heat transfer compound for the final assembly. That said, in case my arrangement isn't efficient enough thermally, I've also tapped some holes to locate the devices directly to the heatsink face but the downside of that is the need for longer wires to connect them to the PCB and they'll be located a litle further from the centre of the heatsink. To check out the heat aspects I plan to run one channel of the amplifier from a SMPS as a test - I've used the heatsink calculator website and the heatsinks should be more than big enough but you have to get the heat efficiently to the heatsink of course.

Incidentally, the main devices are a pair of Tokin THF-51S SITs. They're like TO3s on steroids and all the connections are on the top so you get a large flat contact area when bolting them down. I will have to make some mica or keratherm pads for them.

Tokin THF-51S.jpg (19.42 KiB) Viewed 5220 times

Other drilling and tapping was for assembling parts, such as locations for feet, socketry etc.