[f] AMB on Facebook

β24 in pure class A

beta24

Re: β24 in pure class A

Postby amb » October 31st, 2017, 3:51 pm

Yes you're correct. Just build the high-power class AB version but bias the output stage to about 1.4A quiescent. To do that you need to change R25, R26, VR2 and VR3 to the pure class A variant's values. All other parts can remain the same as the high-power class AB version. If you want a max of around 80-100W output then you can reduce the σ22 output voltage to ±30V and the unregulated supply to ±25V.

Note that you will need even more heat dissipation capability due to the higher rail voltages than the lower-power pure class A variant! The dissipation will be about 140W per channel while idling!
User avatar
amb
Site Admin
 
Posts: 10378
Likes: 175 posts
Liked in: 482 posts
Joined: March 29th, 2010, 9:49 pm
Location: Sunnyvale, CA
Country: United States (us)

Re: β24 in pure class A

Postby MASantos » November 1st, 2017, 11:00 am

I'll take the big hifi2000 case and build the AB version, and then increase bias as much as allowed by the heatsinks.

Regarding board assembly on the heatsink, I am considering placing mosfets a bit apart on the 400mm heatsink by using solid copper wires to connect the pins to the pcb.
I would take the board's inner mosfets ( the ones closer to the diodes) and place them farther apart from the board using 3 or 4 inches of wire to connect them. This might help with the extra power dissipation needed for the high bias and high rail voltages.

Another option is an adicional heat spreader on the heatsinks but it might not be as efective.
MASantos
 
Posts: 75
Likes: 0 post
Liked in: 1 post
Joined: April 9th, 2010, 6:59 am
Country: Portugal (pt)

Re: β24 in pure class A

Postby amb » November 1st, 2017, 2:53 pm

Don’t lengthen the MOSFET pins by any means. The added inductance could make the amplifier unstable.
User avatar
amb
Site Admin
 
Posts: 10378
Likes: 175 posts
Liked in: 482 posts
Joined: March 29th, 2010, 9:49 pm
Location: Sunnyvale, CA
Country: United States (us)

Re: β24 in pure class A

Postby MASantos » November 2nd, 2017, 2:35 am

Ok, I'll study some sort of heat spreader or even an active cooling solution, Such as a silent 200mm fan blowing air through the heatsinks. More complicated but might be necessary.

Is the 1.4 bias a limit imposed by the circuit design or parts choice or could I bias it higher with enough heatsink?
MASantos
 
Posts: 75
Likes: 0 post
Liked in: 1 post
Joined: April 9th, 2010, 6:59 am
Country: Portugal (pt)

Re: β24 in pure class A

Postby amb » November 2nd, 2017, 3:50 am

1.4A is the bias current needed for full output for the low-power pure class A variant. Since you want your amp to stay in class A at up to the same power level, then the same biasing will accomplish it. Above that, the amp transitions into class AB. The 1.4A is not a "limit". It's just a bias point for class A operation for a specific output current level. You can increase it at the expense of more heat. Of course, at some point the heat dissipation will exceed the output device's safe operating condition no matter how big the heatsink is, due to thermal resistance of the device package itself and the isolating pad.
User avatar
amb
Site Admin
 
Posts: 10378
Likes: 175 posts
Liked in: 482 posts
Joined: March 29th, 2010, 9:49 pm
Location: Sunnyvale, CA
Country: United States (us)

Re: β24 in pure class A

Postby shonne » November 2nd, 2017, 1:04 pm

amb wrote:Yes you're correct. Just build the high-power class AB version but bias the output stage to about 1.4A quiescent. To do that you need to change R25, R26, VR2 and VR3 to the pure class A variant's values. All other parts can remain the same as the high-power class AB version. If you want a max of around 80-100W output then you can reduce the σ22 output voltage to ±30V and the unregulated supply to ±25V.

Note that you will need even more heat dissipation capability due to the higher rail voltages than the lower-power pure class A variant! The dissipation will be about 140W per channel while idling!


300VA per channel, will be enough for this combination?
shonne
 
Posts: 5
Likes: 0 post
Liked in: 0 post
Joined: August 4th, 2013, 4:48 am
Location: Aleksinac
Country: Serbia (rs)

Re: β24 in pure class A

Postby amb » November 2nd, 2017, 1:45 pm

shonne wrote:300VA per channel, will be enough for this combination?

Are you referring to the power transformer? If so, yes.
User avatar
amb
Site Admin
 
Posts: 10378
Likes: 175 posts
Liked in: 482 posts
Joined: March 29th, 2010, 9:49 pm
Location: Sunnyvale, CA
Country: United States (us)

Re: β24 in pure class A

Postby shonne » November 2nd, 2017, 4:41 pm

amb wrote:
shonne wrote:300VA per channel, will be enough for this combination?

Are you referring to the power transformer? If so, yes.


Yes, power transformer.
I have enclosure and PSU from A3 and I'll try this combination, very soon. Think, this is what I need. Pure AB is too much power, pure A is too less...30-40W in A and 80-100 in AB is great.
post-103-0-60556000-1318677627.jpg
You do not have the required permissions to view the files attached to this post.
shonne
 
Posts: 5
Likes: 0 post
Liked in: 0 post
Joined: August 4th, 2013, 4:48 am
Location: Aleksinac
Country: Serbia (rs)

Re: β24 in pure class A

Postby obee » November 22nd, 2017, 11:47 pm

Hi TK,

I just bought the beta24 PCB and I'm going to start building it by end of this year. Could you advise the following issues?

1. Regarding beta24, is the v1.01 the latest one?

2. The PDF page 14 mentioned that default (recommended) bias is 200mA and the default BOM allows 300mA max. I also read in forum that it is possible to make it to pure class A. What if I would like to make the bias higher but not full class A? I read that 40W/ch is max for class A. How about if I want to do it class A+AB with first 10W as class A or first 20W as class A? What are the corresponding settings (Resistors, VR, transformer voltages / rating)?

3. Regarding the size of the PCB, I still have not on hand, and it doesn't mentioned in the PDF its size. Could you tell me the size of the PCB? And possibly dimension after putting on the MOSFETs?

4. There's a power on circuit recommended in the PDF which consists of a solid state relay. I checked that relay in mouser but found it very expensive. May I know is it really necessary to have a solid state relay? Is there any alternative that is much cheaper to achieve?

Thanks in advance.
obee
 
Posts: 153
Likes: 0 post
Liked in: 2 posts
Joined: September 5th, 2012, 4:03 pm
Country: Hong Kong (hk)

Re: β24 in pure class A

Postby amb » November 23rd, 2017, 1:12 am

obee wrote:1. Regarding beta24, is the v1.01 the latest one?

Yes.

2. The PDF page 14 mentioned that default (recommended) bias is 200mA and the default BOM allows 300mA max. I also read in forum that it is possible to make it to pure class A. What if I would like to make the bias higher but not full class A? I read that 40W/ch is max for class A. How about if I want to do it class A+AB with first 10W as class A or first 20W as class A? What are the corresponding settings (Resistors, VR, transformer voltages / rating)?

The main concern is heat dissipation, which is directly related to the rail-to-rail supply voltage and the current through the output MOSFETs. The pure class A version of β24 has much lower supply voltage, therefore turning up the bias to full class A is OK (assuming an adequate heatsink) for up to 40W output. If you want to retain the full 170W+ per channel of class AB output power, then you need to run at the higher supply voltage, but that then restricts how heavily you could safely bias the output stage without overheating. You can experiment with the bias setting once you're done building it, watching the heatsink temperatures near the MOSFETs. But do so very carefully if you're running without the safety temporary current-limiting resistors. With the resistors, they will drop significant voltage as you turn up the bias, and the power dissipation on the current-limit resistors may become too high.

3. Regarding the size of the PCB, I still have not on hand, and it doesn't mentioned in the PDF its size. Could you tell me the size of the PCB? And possibly dimension after putting on the MOSFETs?

The board size is 4.5" x 7". The MOSFETs will protrude approximately 1" on each side, depending on where you locate the MOSFET mounting holes relative to the board.

4. There's a power on circuit recommended in the PDF which consists of a solid state relay. I checked that relay in mouser but found it very expensive. May I know is it really necessary to have a solid state relay? Is there any alternative that is much cheaper to achieve?

If you want to use a regular (electro-mechanical) relay, then you need to find one that has similar contact rating as the specified solid-state relay. Unlike regular relays, SSRs need very little current to activate. Since the control of this relay will be wired in parallel with the coil for the bleeder relay, if both are regular types, then the combined coil current may be too high for the ε24. In a high current application, SSRs won't have the potential problem of the contacts oxidizing or wearing out.

Mouser and Digi-Key prices for SSRs are high, but you will find much less expensive ones on eBay. You can even consider used ones (removed from equipment), as SSRs are nearly indestructable.
User avatar
amb
Site Admin
 
Posts: 10378
Likes: 175 posts
Liked in: 482 posts
Joined: March 29th, 2010, 9:49 pm
Location: Sunnyvale, CA
Country: United States (us)

PreviousNext

Return to β24 fully-differential power amplifier

Who is online

Users browsing this forum: No registered users and 4 guests

Quick Links

AMB audio DIY resources
AMB audio shop