Burning Amp BA-3

There is only one supplier of chassis kits which incorporate heatsinks in Europe to my knowledge and that is MODU in Italy. I have ordered a chassis from them in the past and they delivered without problems to UK. This is one way to go if metal work is not your thing. The build on these chassis is not perfect but at least you get metal parts of the right dimensions with drilled holes. However, I still ended up tapping holes in the heatsinks for the screws to hold the mosfets. The biggest issue I had was deciding how big the heatsinks had to be to cope with the heat output from the mosfets. It depends of course on how much bias you give the output devices but the bigger the better is the rule of thumb and to this end companies in the UK such as Birmingham Aluminium could supply some really big heatsinks and were very helpful on the phone. If you can then attach some metal work to this you are half way there. One aspect not mentioned is the importance of a sturdy base plate (which I think can be ordered from MODU). This is important because it needs to be stable and strong enough to support your transformers and large heatsinks. I have to say that the most difficult part of the build was the metal support infrastructure. The placing of the mosfets on the heatsinks is important also to allow them to dissipate heat for maximum efficiency and Andrew laid out some important information if you search the BA-2 thread.
Hope this helps.
Chris
 
If you look on their site they give the heat dissipation characterisitics of their heatsinks, however, the heatsink I think beyond 300mm (from front to back) consists of two 200mm heatsinks bolted together. I never did find out how this altered the heat dissipation characteristics. Instinctively (not scientifically) I would prefer the heatsink to be a single unit. In fact looking back on my experiences I would now start with some really big heatsinks with very efficient cooling characteristics and then look for some metal plates to bolt on after tapping some holes in the heatsinks. If the latter are fairly substantial they will physically take quite a load.
If this is your first build don't rush--good idea to get the parts and then you'll be better able to judge the chassis dimensions.
Chris
 
Another Black hole for me at this point is the power supply parts.

I have a 2x20V 630VA toroid Xformer (for two channels), and I am wondering about wicht value I should put on C3..C6.

I guess between 47.000 and 100.000 uF per channel and per rail ?
I know some like to get big bottles there.

As for the resistor for the RC lowpass, I guess it should drop about 1 Volt at 3 Amps, so three 1 ohm / 3Watt resistors in parallel should be OK.

Am I burning ? :rolleyes:
 
As for the resistor for the RC lowpass, I guess it should drop about 1 Volt at 3 Amps, so three 1 ohm / 3Watt resistors in parallel should be OK.

Am I burning ? :rolleyes:

Try four of the 3watt, 1ohm resistors as in the circuit diagram; then you'll have some headroom when things warm up.

As for the caps, the bigger they are the more difficulty the rectifier diodes have in charging them. I used two 10,000uF 63V Mundorfs per rail for each channel and they sound fine. More importantly in my view is to get the 10uF cap of the power supply right. Definitely go for a good polypropylene--I used an Ampohm (£10) and am currently getting good results with a Clarity Cap ESR rated at 250V (£8) which is smaller. I'm going to use the latter with the BA-3 front end.
Chris
 
the heatsink I think beyond 300mm (from front to back) consists of two 200mm heatsinks bolted together. I never did find out how this altered the heat dissipation characteristics. Instinctively (not scientifically) I would prefer the heatsink to be a single unit.
I think it depends upon how many power transistors are in your circuit, and whether they benefit from matched temperatures. If you have many power transistors that are independent, then cooling might be better with separate heat sinks.

The ADCOM GFA-555 has four circuit boards with (I recall) two power transistors each. There are four separate heat sinks on the back of the amplifier chassis, two per channel or one per board. I noticed as I took off the chassis cover that the heat sinks are actually separate, and merely bolted together by the screws that hold on the bottom plate and top plate. I do not know whether Nelson actually designed these aspects of the enclosure, but it seems like a relevant observation.
 
The ADCOM GFA-555 has four circuit boards with (I recall) two power transistors each. There are four separate heat sinks on the back of the amplifier chassis, two per channel or one per board. I noticed as I took off the chassis cover that the heat sinks are actually separate, and merely bolted together by the screws that hold on the bottom plate and top plate. I do not know whether Nelson actually designed these aspects of the enclosure, but it seems like a relevant observation.

It is a design matter really. The position of the mosfets on the BA-2 power board are not set by me which means that when I line them up against the two heatsinks bolted together at least one will end up close to the edge of the join. I'm sure this means that the ability of the heatsink to dissipate heat will be impaired because the mosfet is close to the edge. Of course you could cut the board in two and have three on one side and three on the other but it would all be very messy.
Chris
 
Hmmm wonder if Nelson will add to the Burning Amp story at this year's event.
Oh wait! He already did! But we can hope for a little spin on things can't we?
I just sayin' why wouldn't he, he's on fire with this new board!
Mark

It would be great to get over to the BAF but with my bad back and fear of flying it just ain't going to happen this year. Any chance of a EuroBAF, maybe at the Munich show. Good reason to visit a great German city. Does Nelson travel well? Fantastic beer.

I'll take the train from London
 
Member
Joined 2004
Paid Member
Gain different than expected

Hi Guy’s,
I fired up the BA3 input board with the FQP3N30 and FQP3P20 and I have only a gain of ~8X. I was going to reduce the gain so I started with R10 and R11 of 50 ohms and I had a gain of about 3.5X. So I started cutting back to the 22 ohms specified and could only get a gain of a little better than 8X. Is this just the difference in the Fairchild VS Toshiba parts? I thought they were equivalent maybe I should start hinting for the Toshiba's, any ideas?

Thanks,
 
The one and only
Joined 2001
Paid Member
For those of you who are not attending BAF, you will delighted to know that
I will be presenting the BBA-3, the balanced version.

Not to despair, I will put it online very shortly after. In the meantime,
hands-off to the usual smarty-pants who are probably already working on
this...

:cool: