Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › β24: A discrete, cascoded, fully-differential power amplifier
New Posts  All Forums:Forum Nav:

β24: A discrete, cascoded, fully-differential power amplifier - Page 21

post #301 of 309
3/30/10 at 9:48pm
Quote:
Originally Posted by digger945 View Post
On the PS schematic the center taps are connected to ground. Does your transformer have center tapped secondaries?

If not, then don't connect any part of your dual secondaries to ground, anywhere.

EDIT: My bad if you're talking about 4 wires(in series) for only one secondary of one transformer, and using the two wires as center tap. Maybe I need to shower and hit the sack now.
They are separate secondaries but since they are in series, it's basically a center-tap. At least that's my current (har har) understanding.
Reply
post #302 of 309
3/30/10 at 9:58pm
Quote:
Originally Posted by Iniamyen View Post
They are separate secondaries but since they are in series, it's basically a center-tap. At least that's my current (har har) understanding.
I see now. My bad.

May I say that I think you have a great looking build going there. I am anxious to know what you think of it when it's finished and you're listening to it.
Reply
post #303 of 309
4/2/10 at 8:21am
I must say after having brought my beta24 into my apt from the car, I have a much greater appreciation of the efforts of Amb to bring his amp to the last CanJam I went to.
Reply
post #304 of 309
4/2/10 at 9:11am
Thread Starter 
Quote:
Originally Posted by MASantos View Post
AMB, I tried to find the information, but didn't succeed...

With standard biasing, how many whats are biased into class A at idle?
Sorry for the late reply, I missed this post somehow.
Anyway, "standard" biasing is 200mA quiescent on each bank of output MOSFETs. A push-pull output topology will remain in class A as long as the peak output current is less than twice the quiescent. Into 8 ohms, this means about 1/3W peak. You could turn up the bias to make the transition to class AB higher, but the MOSFETs will run very hot.

Quote:
Is there any problem with making this a "full" class A amplifier by increasing the biasing, apart from the need of much larger heat sinking?
Not without drastically reducing the supply voltage and power output.
A pure class A speaker amp of an equivalent size and heatsinking to my β24 should realistically be a 40W/channel amp, not 180W/ch.

The TO-247 MOSFETs in this amp cannot take a huge amount of increased heat dissipation, even if the heatsink size is increased. A traditional method of increasing heat dissipation capability is to parallel multiple sets of output devices, but the PCB design doesn't lend to such a modification, and the gate capacitances of the output MOSFETs will be multiplied by the number of added sets, which is not desirable.
Reply
post #305 of 309
4/5/10 at 2:52am
Made some progress today after not touching this for weeks.

Front panel:



Boards mounted on heat sinks:
Reply
post #306 of 309
4/8/10 at 12:25am
Posted some new pics over at Amb's forum site. I'll probably be using that thread for updates ongoing:

Here's the link.
Reply
post #307 of 309
4/8/10 at 2:58am
Nice!
Reply
post #308 of 309
7/25/10 at 9:50am
Thread Starter 

I just posted about a 30W/ch pure class A variant of β24.  Some of you might be interested in building such a beast.

Reply
post #309 of 309
7/25/10 at 10:28am

Nice looking job on the case...

Reply
New Posts  All Forums:Forum Nav:
  Return Home
Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › β24: A discrete, cascoded, fully-differential power amplifier