• These commercial threads are for private transactions. diyAudio.com provides these forums for the convenience of our members, but makes no warranty nor assumes any responsibility. We do not vet any members, use of this facility is at your own risk. Customers can post any issues in those threads as long as it is done in a civil manner. All diyAudio rules about conduct apply and will be enforced.

my_ref revC group buy (Mauro Penasa LM3886 design)

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Jump to Latest
schro20

schro20

Member
Joined 2008
multires.caltech.edu/
AndrewT said:
at least twice what National specify in the datasheet.
Click to expand...
While some have reported fine results with rather meager sinks I am also in the camp that too much is never enough when it comes to heat sinks. A good read on sizing these beasts is the article by Rod Elliott http://sound.westhost.com/heatsinks.htm

I can tell you that the heatsink you see in http://picasaweb.google.com/schro20/FinalBuildOfRevCFromKit#5296425580447681922 is far too small. It got hot enough that I could barely put my finger on it for more than a few seconds without deep meditative breathing! :) (I ran the amp at full tilt for half an hour) Technically that's still ok (say 50 C above ambient), but the chip performs better and lives longer if you don't go to such extremes.

Uriah (who is offline for a few days) has had good experiences with CPU coolers. I haven't tried that myself, so I can't comment.

peter
 
A

AndrewT

R.I.P.
Joined 2004
if that sink that schro says is too small then believe him.

I'd guess it is about 4C/W.
Table, on page10, shows ~60Vdc for an 8ohms speaker. This will run the chip very hot.
Adopting an 8C/W design, but using a double size sink (4C/W) would be ~ 46Vdc for an 8ohm speaker.

I suspect Schro used a lot more than +-23Vdc for his supply.
 
schro20

schro20

Member
Joined 2008
multires.caltech.edu/
troystg said:
The heatsink in your picture would probably be fine if the fins were vertically oriented. Efficiency is LESS THAN HALF with horizontal orientation.
Click to expand...
Correct. When I did the heat test I oriented it sideways. I drove my large FE207e Metronomes pretty loud. Now these babies are quite efficient, so they don't actually stress the amp that much... One thing I still need to do and haven't is to check on any high frequency oscillations since they could explain seeing far more heat than one "hears." Though this design is supposed to be quite good at not burning through a lot of power outside the audible spectrum.

Peter
 
marc brown

marc brown

Member
Joined 2007
schro20 said:
One thing I still need to do and haven't is to check on any high frequency oscillations since they could explain seeing far more heat than one "hears." Though this design is supposed to be quite good at not burning through a lot of power outside the audible spectrum.

Peter
Click to expand...

I'm curious to hear about your results. My next project is a NOS mod to an old Magnavox TDA1541-based cd player. Removing the OS has the potential to drive some amps into HF oscillation or burn up a lot of power.
 
schro20

schro20

Member
Joined 2008
multires.caltech.edu/
AndrewT said:
I'd guess it is about 4C/W.
Table, on page10, shows ~60Vdc for an 8ohms speaker. This will run the chip very hot.
Adopting an 8C/W design, but using a double size sink (4C/W) would be ~ 46Vdc for an 8ohm speaker.

I suspect Schro used a lot more than +-23Vdc for his supply.
Click to expand...
I used a 25-0-25VAC toroid which is rated at 300VA. In practice that means that I was actually seeing 27-0-27VAC. Multiply by sqrt(2) and you get some pretty beefy DC on the rails...

BTW, anybody who says that bigger toroids give them better bass first has to show me that they are running with the same *actual* voltage. Most of the time one only draws a small fraction of the total wattage and hence operates most of the time in the higher voltage region that a larger toroid will be designed for...

I have heard big differences in bass delivery, but I suspect most of it is due to simply running at effectively a higher voltage. But I would be interested to hear of any comparisons *at the same operating voltage* and higher wattage. Anyway, a different argument then heat sink size.

Did I mention that ApexJR has some nice and BIG heatsinks? Very good prices too. I schlepped off with some serious metal for my final build! :)

Talk to Steve of ApexJR. Not everything he has in big metal is on the website.

peter
 
schro20

schro20

Member
Joined 2008
multires.caltech.edu/
troystg said:
See his last sentence! :D

http://www.diyaudio.com/forums/showthread.php?postid=1758777#post1758777
Click to expand...
:)

Maybe we can get Dario to come up with the ultimate component selection... Dario? You game? (Just doing a dare here... I am not about to do another one; this was fun, don't get me wrong, but I have so many other things I want to try right now; several cabinets that await finishing, several more amps that want to be built and compared, etc. etc. etc.)

Peter
 
ClaveFremen

ClaveFremen

Member
Joined 2005
inserra.it
Paid Member
schro20 said:
Maybe we can get Dario to come up with the ultimate component selection... Dario? You game?
Click to expand...

The component selection of the GB is quite good and is a good balance of performance vs. cost, IMHO.

Adding some bucks it could be upgraded, of course... ;)

I think that I wouldn't be able to have the patience you and Uriah had managing the GB. :D

Also I don't have enough knowledge to modify Eagle files too... :rolleyes:

But for support, beta-testing and similar I could be in. :smash:
 
ClaveFremen

ClaveFremen

Member
Joined 2005
inserra.it
Paid Member
marc brown said:
Based on your research I'm going to pull out all the AVX's and put in some Vishay MKP's and bypass the big caps with MKP's.
Click to expand...

Hello Marc,

pay attention to the fact that my tests was for a signal path bypass cap, AVXs could be good in PS...

I suggest you to use sockets too, it could be that Vishays MKP aren't any better... :rolleyes: I've had some surprises too...
 
schro20

schro20

Member
Joined 2008
multires.caltech.edu/
ClaveFremen said:
The component selection of the GB is quite good and is a good balance of performance vs. cost, IMHO.
Click to expand...
This is mostly Brian's work. Whenever I tried to deviate from his recommendation to get a better price I soon returned to his suggestion (or an equivalent) realizing just how well the trade-off between price and performance was done by him. Kudos to Brian!

Peter
 
schro20

schro20

Member
Joined 2008
multires.caltech.edu/
ClaveFremen said:
pay attention to the fact that my tests was for a signal path bypass cap, AVXs could be good in PS...

I suggest you to use sockets too, it could be that Vishays MKP aren't any better... :rolleyes: I've had some surprises too...
Click to expand...
One important word of caution. When talking of parts just giving the company name doesn't tell the story. One has to be careful to specify the exact series the part comes from. I discovered that when spec'ing parts. In particular because some firms have been bought up by others and you can get some real surprises by that.

peter
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Top Bottom