aleph 30 project

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
so starting to put together an aleph 30 using briangt boards.

some questions...

planning to put one output board per heatsink, 4 of them, 2 to side as in picture. just barely fits width wise. are the sinks big enough? they are about 4.25 inch x 4.25 x 2 inch thick.

going to be a tight fit, 12 inch wide by 9.5 deep 'working' area. might need to get another .125 sheet and go 14 inch wide, overall chassis would be about 18.5 then though.

transformer is a 36-18-0-18-36 500VA. planning to use the 18-0-18... when they make transfo's like this and one uses the inner steps, still get full juice out of it? don't need to run any secondaries in parallel like that do i? how much juice does an aleph30 pull ? (on the secondary dc voltage side, not from the wall)

one last side question... often see 28-0-28 recommend for aleph 5 pulling down to 34v dc rails. I have 800va dual 31, can a simple zener emitter follower regulator be used to shave off some volts? so, also curious how much juice an aleph 5 would draw such that if shaving say 4 volts at X current, how much transistors burn off (thus want minimal extra heat sinking for the regulator). this is for a second project where i know i will need bigger heatsinks. doing the aleph 30 first to try.
 

Attachments

  • aleph30.jpg
    aleph30.jpg
    63.7 KB · Views: 2,358
All IMHO:

Your sinks look a little light to me, especially if you want to ramp the bias up.



still get full juice out of it?
Doubt it, actually no. You'd get 1/2 juice. I'd bet a dollar that if you cut that connector off of the toroid you would find 4 18v secondaries. I'd bet they are just in series right now to give that 36-18-0-18-36 spec. 4 18v secondaries is way more useful for the project and you'd get full 'juice'. I'd check if I were you.

The only practical way to drop a few volts in an amp like this would be a CRC or CLC in my opinion. Fortunately you'll lower your ripple too.:)

Good luck, it's a great project!
 
Your heatsinks are too small for 30watts output per channel. Also you would need about 20vac from the trafo before rectification.

An Aleph30 needs less than .3 c/w (corrected for ambient room temperature) per channel to disipate 100watts of power for 30 watts output.

I'm using one 13.8x6x1.9" per channel and they are about right. The sink temp after 2 hours is 150 degrees F.

I'd double up on the heatsinks or reduce the power output.

My Aleph30 uses http://www.conradheatsinks.com/products/flat100_350.html rated at .21 c/w at 80 degrees C. Applying a correction factor of 1.4 I get a .294 c/w rating for a typical room temp.

Here's a link to my Aleph30:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=70622&highlight=

Best of luck, The Aleph30 is fun to build.

-David
 
traw said:
what are the symptoms of overheating? or, how do i know it's too hot?

built it... sounds nice. gets hot, but probably too hot. should i measure temp some how?


Nelson's rule of thumb for DIYer's has been, "if you can leave your hand on the heatsink for about 5 seconds, the temp is around 150 degrees F."

Any hotter than this and you risk greatly reduced lifespans for the output FETs. Also leave a decent amount of ventilation in the case for the other parts to remain cool. The other parts like capacitors have shortened lifespans as well in high heat environments.

-David
 
You will also probably need a CRC filter so you don't get any excessive noise on the supply. So include an other set of capacitor and some provision for a voltage drop into the CRC resistor.

You can also use a CLC filter but choke are more expensive than power resistors. See my own CRC filter, one per channel to go a simili dual mono supply. The power resistors are mounted on the frame.

An again for the Aleph 5, you can use the CRC filter to remove some voltage from the supply...

Bye...
 

Attachments

  • crc.jpg
    crc.jpg
    11.3 KB · Views: 1,713
traw said:
will this chassis do better? current one works, but too hot after at about an hour. fans work but bah....

also switching to a 20x2 vac toroid as my current one sags down to 21 volts.


Very nice. When you build the chassis leave some room maybe 1.5" ish to allow air to rise through the fin's.

Jase
 
from crtsurplus on da'bay... think another diyaudio member here had used them as well. sold in 12x12x4 piece, 30 bucks. i chopped so i could use up some left over bar stock. remaining peice, 7.5 inch, for another project... maybe aleph 5 depending on how these 4.5x12 sections work out for the aleph 30. each one uncut is about 20 pounds. 66 fins. they aren't extruded... the fins are cemented with some dark gray compound (?) to the base, 1/2" thick.

think about trying to make home made sinks similar to the pass styling - horizontal fins but bent upwards. would be much less tedious than trying make something like these with the vertical fins. if works out, will post results of that endeavor. any easy way to measure ballpark performance?
 
up and running

got it done now, just need to pick up some sheet metal for the top... it's just about right as after running overnight can touch for 3-5 seconds firmly. transformer gets warm. not worried, will beat into ground and replace if have too. should have used a bit bigger. i used a 400va 20x2. CRC uses 2x0.3ohm and get 23.5 volt rails.

it was running good for a few days, then a channel went out after running a few hours after completed rear plate. was freaking out thinking damaged output devices from heat. seems a solder bridge on Q1 (one of the input diff pair) started acting up. did some searching and inquired with a diyaudio member directly via email giving me some starting points helping to identify the culprit. thanks to netlist post that list voltage points for aleph 3 (similar) and email responses from bob ellis. reached in with wick and iron getting the excessive solder. phew... sounds great. had a small hum on this second version (same boards new chassis)... moved the board ground from the cap bank as star ground to a new star ground at the chassis and went away.

phew... aleph 30 really has some bottom end.

attaching slew of pics
 

Attachments

  • a30 front angle.jpg
    a30 front angle.jpg
    48.1 KB · Views: 1,011
Aleph 30 - Distorsion on higher frequencies

Hi DIY-ers,

Another build of A30 using Brian's board has been completed. It was really good fun and results were quite pleasing.

Off-course I couldn't just left it there, so I started testing, measuring, tuning ...

I've noticed, that on higher frequencies (10K and more) and signal levels there is certain distorsion quite visible on my DSO.

I attached couple of DSO screenshots. First one is output (4ohm dummy load) at 14.4Vrms and Second one is near clipping and 15Vrms. Area in question is circled red. Is this normal type of distorsion for Aleph 30 ?

I am running with 22V rails, total of 2.5A (Ch1) and 3A (Ch 2) total bias (just for test ) and AC gain of 45%-55%.

Any help would be very appreciated.

Regards, Boris
 

Attachments

  • Distorsion on high signals 10KHz.jpeg
    Distorsion on high signals 10KHz.jpeg
    192.5 KB · Views: 452
  • Distorsion near clipping 10KHz.jpeg
    Distorsion near clipping 10KHz.jpeg
    188.7 KB · Views: 161
That is pretty lousy looking, isn't it. It looks like you are well into the area where it has run out of poop. 22V rails won't get you a clean 14Vrms output, you have to take the losses into account such as source R drop and Vgs of the fets. Also, the difference between 45% and 55% AC gain is pretty substantial. Just trying to be helpful.
 
That is pretty lousy looking, isn't it. It looks like you are well into the area where it has run out of poop. 22V rails won't get you a clean 14Vrms output, you have to take the losses into account such as source R drop and Vgs of the fets. Also, the difference between 45% and 55% AC gain is pretty substantial. Just trying to be helpful.

Thanks for your comment Bill. I am currently experimenting with different bias's so one channel is at 2.5A (total) with 55% ACG and second one is at 3A with 45% ACG - I still have to bump up that one. I am getting about 12Vrms without distortion - which should give me ~ 35W on my 4ohm Dali speakers. Enough for me .

My main concern was to see if there is anything else that I could do to improve distortion that occurs before clipping. ie do you guys have clean waveform before going into clipping or you are seeing similar behavior.

Cheers, Boris
 
Official Court Jester
Joined 2003
Paid Member

Attachments

  • Aleph CCS modification.pdf
    52.2 KB · Views: 215
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.