Aleph J-X Amp Project

[SCAIFF001]

Originally Posted by Peter Daniel

When you check traces it shows that GND and SG are connected; SG means signal ground and is the same as GND.

I used two separate pads: one for input signal ground, the other to connect ground reference from the power supply.

I can't be much help in troubleshooting as I didn't built AJ-X yet, however the boards were tested and work fine if properly set up.

Hi,

when you say "ground reference from the power supply" are you referring to the two Bridge Rectifier(s) positive & negative outputs that are connected together along with connecting to the positive & negitive from the power supply capacitors, then all of these are connected to the Ground on the "front end pcb", I believe that this is what I have seen on the schematics ? A reply to this will help me out because connecting the "+" & "-" together with the ground seems counter intuitive to me, so it is responsible for my project taking so long to build because I dred reaching this point.

Thank You

Ian

[oiphy]

scaiff001: tips. solder 1 or to points then clean tip with sponge. if it dont help fill the soldertip with solder before the sponge treatment. if this dont help I guess your tip is already broken. worked for me in 20 years both prof. and private. and I still have tips that are 10-15 years old and they are fine.

[Itsmee]

I find this rather handy for cleaning iron clad tips.

Buy Soldering Iron, Tips and Accessories TTC-LF Soldering iron tip cleaner/tinner Multicore 706397 online from RS for next day delivery.

[mravinsky]

Samoloko, you have a PM

[jerfer]

Hi all,

I'm contemplating building an Aleph-J and then converting it to a J-X channel once I'm comfortably experienced. A few questions before I jump in, though.

Dissipating around 30W per fet and, ideally, thermally connecting all of the fets (4) of one channel means that I would need to find a heatsink with

C/W = 30 deg. C / 120 watts / 1.25 (de-rating factor)
or
C/W = 0.2

Looking through the various heatsink manufacturers' websites nothing comes close, even at 12" lengths, unless I blow air across the sink. So here are the first questions:

1. How do you do it, do you just build using multiple widths of smaller heatsinks thus not end up thermally joining the fets?

2. Where do you draw the line, in the A-J's case would you go down to one-fet-one-sink?

HeatsinkUSA has a nice heatsink on offer with a C/W of 0.8. Going to a 12" length should yield a C/W of around 0.4.

3. How much would black anodizing further improve the C/W of this heatsink?

One thing I've been thinking of is building the case with some sort of chimney to generate a stack effect, like in buildings and chimneys, where rising hot air escaping from the top of the stack draws cool air in from the bottom. Building a case with the heatsink flat, fins pointing up, and covering the sink with a chimney assembly should cause an appreciable amount of flow. By my calculation a 20"-high chimney topping a 12"x12" heatsink (sealed on all but the fin edges) should cause a ~50cfm airflow. This assumes a heatsink temperature of 50 C, and an external temperature of 20 C. I can't be the first to think this would work, so:

4. What am I missing this with line of reasoning?

Being that I want to ultimately build a J-X I'd like to size the transformer to be good for two A-J channels or one A-JX. Antek offers a 600VA 18V toroid which, from comments here, will be under-powered. They also offer a 800VA 20V toroid which would probably be sufficient. Here's my PSUD simulation:



Note that the only reliable values in this simulation are the ESR values for the parallel caps (3x56mF). The transformer and current sink I'm not sure I configured correctly.

5. Can I use the 800VA/20V trafo in a CRC configuration without having to cascode (and learn how to do so, at least not now) the JFETs?

Thanks for any help and advice you can offer, it is much appreciated.

Jerry

[h_a]

Hi,

typically one uses (especially for the monoblocks) 2 heatsinks and puts one fet bank (the paralleled fets) on each of one, so - in case of the Aleph J - one pair per heatsink. Alternatively, contact a large heatsink manufacturer (like R-Theta) directly as these do offer 0.2 heatsinks (but typically not available at the large distributors).

Black anodizing mostly looks good and gives some isolation, no real benefit from a thermal point of view (heatsinks are about heat convection not heat radiation).

Last, with that amount of caps in the PSU you'll likely get problems with inrush currents and max diode currents. Either reduce them or think about soft-start.

Enjoy! Hannes

[qusp]

placing the heatsink with its fins pointing up reduces the efficiency quite markedly

[john65b]

The heatsinks fins should be parallel to the airflow rising against it like a chimney.

The Black anodizing allows more radiative heat transfer to environment, but is dwarfed by the conductive heat transfer to air at these amp temps. Once you get into very high temps (1000F) then radiative starts to govern and really matter. When Radiative heat transfer is high enough, your bare skin will feel the heat, like sitting next to a campfire.

[jerfer]

Originally Posted by h_a

typically one uses (especially for the monoblocks) 2 heatsinks and puts one fet bank (the paralleled fets) on each of one, so - in case of the Aleph J - one pair per heatsink. Alternatively, contact a large heatsink manufacturer (like R-Theta) directly as these do offer 0.2 heatsinks (but typically not available at the large distributors).

I assumed (wrongly, it seems) that since the four FETs in each A-J channel need to be matched to 0.1 Vgs they would benefit from being thermally connected. This is good news! It opens up a whole new set of possibilities...

Originally Posted by h_a

Black anodizing mostly looks good and gives some isolation, no real benefit from a thermal point of view (heatsinks are about heat convection not heat radiation).

Originally Posted by john65b

The heatsinks fins should be parallel to the airflow rising against it like a chimney.

The Black anodizing allows more radiative heat transfer to environment, but is dwarfed by the conductive heat transfer to air at these amp temps. Once you get into very high temps (1000F) then radiative starts to govern and really matter. When Radiative heat transfer is high enough, your bare skin will feel the heat, like sitting next to a campfire.

Interesting, so there is no real C/W-effect with black anodizing?

Originally Posted by h_a

Last, with that amount of caps in the PSU you'll likely get problems with inrush currents and max diode currents. Either reduce them or think about soft-start.

I was thinking a CL-60 in series with each primary. Not enough?

[vmcfer]

Originally Posted by jerfer

I assumed (wrongly, it seems) that since the four FETs in each A-J channel need to be matched to 0.1 Vgs they would benefit from being thermally connected. This is good news! It opens up a whole new set of possibilities...






Interesting, so there is no real C/W-effect with black anodizing?



I was thinking a CL-60 in series with each primary. Not enough?

I group members,
Does anybody an address to buy PCB's to build the Aleph-X amplifier at a correct price?
I am interested in buy 3 of them .
Thanks,
Victor