New Aleph Mini PCB GB

Status
Not open for further replies.
Jump to Latest
#942
randytsuch said:

....................................
Hi Zen
When I looked at using inductors in my PS filter before, I was modeling a CLC filter, and could not get enough voltage drop. So, I ran PSUD with an LC filter.

BTW, I am not sure what “1,41” means? 1,410 uf, or 14,100 uf, or 141,000 uf? 14,100 uf seems like the most appropriate, but 1,41 would be a strange way to say that, and the other values either seem too low or high.

So, I looked at Hammond chokes for the L.
From the original mini aleph thread, they draw around 1 amp at 15VDC. So, I figured I need a choke rated for at least 1.5 amps. Best Hammond I could find was the 159ZC, it is rated at 2A, 60Mh and 0.7 ohms. This guys weighs 2.5 lbs, and costs about $20 each.

If I plug these parameters in, with a 14.1 mf cap, I get 16VDC output from the simulation. My transformer is rated for 18V at 6.25 A, 7% load regulation.

Does all this seem reasonable?
I was thinking I could lower the current, so I could use a chock with a lower current rating, but lowering the current would increase distortion, so I am not sure I want to do that. And, I am not sure I want to spend $40 on two inductors, to build a LC supply.
.....................

Randy
Click to expand...


when you have filter with C bank on first place, you compute DC voltage as Uac x 2squared = Uac x 1,41 without load.......depending of load it can be pretty lower,and in A class amps you can often have just 1,2 instead of 1,41`

when you have series choke as first element in filter ,and you must add C bank,of course, you compute DC voltage somewhat harder,but there is always Duncan's proggie ........but in any case you will not have DC voltage more than 0,9 x Vac.

with choke almost always you have better regulation of filtered voltage,than in any other case......and price you already know-higher overall losses and somewhat emptier wallet

I really don't remember which member is already big fan of L input filter,but I'm sure that he tried it from same reasons-too much voltage in xformer sec for needed purpose....after that he just loved that approach and ,again, if memory serves me well- he continued to use this approach.
I know-chokes are not cheap , but just for moment think about commercial value of amp you going to make.......

buy Pass original one,and there you'll have adequate xformer,adequate C bank , and-off course adequate load for supply
:devilr:
 
#943
Thanks for the encouragement, guys. I appreciate it.

Last night I switched C1's polarity and fried the same resistors again.

For the 10PF 300V MICA RADIAL, I'm using a MICA 10PF 300V SMD instead, and the connection is UGLY. Can this be a big problem?

I still don't understand how my other board doesn't run at all. I've thoroughly inspected the wiring, especially the voltage in/out leads, and everything appears to be correct. At this point, I'd prefer it if the board just ran and blew up.

Also, with a multimeter, I've checked the FETS, and they appear to be working properly.
 
#948
spzzzzkt said:
Brian GT did confirm that the polarity marked for C1 is incorrect in a followup post:

http://www.diyaudio.com/forums/showthread.php?postid=630253#post630253

insert C1 so it is the "wrong" way with reference to the board markings 😉

Paul
Click to expand...

There is no sound difference at all if it is in wrong. Its just good to have it in correctly. i build 1 channel with it in wrong and compared and heard nothing new or old.
 
#950
Here's the board that actually sort of worked for awhile; now I get no sound, just hot FETS:

http://members.shaw.ca/GMTMASTER/Cr-ap1.jpg

Notice the toasted resistors. The C4 cap is hideously connected via wires, a few inches away from where its board position is supposed to be.

This is the one that is probably wired correctly but doesn't work at all as far as I can tell:

http://members.shaw.ca/GMTMASTER/Cr-ap2.jpg

Show me some sign of life, please?

$400 poorer and nothing to show for it.
 
#951
Here's the board that actually sort of worked for awhile;
Click to expand...
Did you P2P wire that, using a printout as template? Nice idea. Looking at your setup, I think you might have a bad connection or short. The traces on the board are running pretty close to each other on both sides, and its very easy to have a short if 2 leads are bent or something. If everything is connected correctly, these boards usually works. You might want to control everything again, following the partnumbers one by one. Look at a clean printout while doing it, all the leads could cause confusion.


Steen🙂
 
#952
I also didn;t see any insulators between your FETs and the heatsink. Shorting them together (metal surface on the back of the FETs via the heatsink) would be bad (hot FETs, little or no sound). It would also burn your output resistors (0.47ohm).

EDIT: Also, the clips are a neat idea for a quick power up test, but they will not provide enough clamping pressure for actual use.
 
#954
The FETS' pins don't actually make contact with the clips/heatsinks, so I think that shorting is not an issue, or at least I hope.

What worries me the most is that the 2nd board that I wired is the one that doesn't work at all. I actually printed out the rear of the PCB on that one just to make 100% sure that I wouldn't mess up somewhere. My guess is that one or two res/caps must be incorrect or DOA.

If anyone has a spare board, just one, I'd love to buy it 🙂 The only reason I did P2P is because I jumped onto this project well after Brian had sold out of his boards.
 
#957
Speaking of insulation between the back of the FET and the heatsink, I wanted to mention that, one time I was troubleshooting the FET for Aleph 30 that I was building, I lift the FET off the heatsink while the circuit is "hot", in less than 20 seconds, the FET (& one resistor) are fried. I doubt the paper clip can provide enough grip to transfer the heat. This is probbaly not the reason why your amp is not working but I thought I mention it.
 
#958
I think Brian already found your problem, on the IRFP240's the part against the heatsink is tied to the drain, you should be able to confirm this with a meter.

BTW, did you check recently with BrianGT to see if he has any more boards. I think he sold out of SM boards, but he might have a th board left. I bought some from him a couple months ago.

Zen,
I am using the Duncan program to simulate my power supply, that is how I know it will put out 16VDC at a 1 amp load, using those hammond chokes.

16VDC is really higher than I want, I now am thinking I could add a little value power resistor, along with the choke, to bring down the voltage to closer to 12 VDC

Randy
 
#960
Thanks for the advice.

I've tried taping the FETS, and now the 2nd board works somewhat at low volume.

When I connect both boards together, only one channel works though.

Also, when I mix the CGND wiring, I get either centre mono or one-sided stereo.

Board 1 gets HOT very quickly and runs loudly now but distorts and crackles at moderate volume, while board 2 doesn't get much volume but never even gets lukewarm.

Perhaps the culprits are some resistors on board 1, and maybe some caps and/or busted FET on board 2?

P.S. Hipcheck, please check your e-mail in a few minutes. I might have some of those parts that you're missing. I'm using a Mica SMD CAP (which is easy to connect to a PCB but a pain with P2P) instead of radial for C4.
 
Status
Not open for further replies.
Top Bottom