The BULLET Amplifier.
This is the power supply.
Each channel output is biased with 900mA.
This is enough to give a bit more than 10 Watt Class A into 8 Ohm.
The transformer is 2x24VAC with one winding each for the channels.
The voltage of supply is like 32.1VDC.
This is the power supply.
Each channel output is biased with 900mA.
This is enough to give a bit more than 10 Watt Class A into 8 Ohm.
The transformer is 2x24VAC with one winding each for the channels.
The voltage of supply is like 32.1VDC.
Last edited:
The BULLET Amplifier.
There are 2 potentiometers.
R6 for adjusting bias to 900mA.
R11 for adjusting the voltage at the output. Set the output like VCC/2.
Actually optimal output voltage at 32.1V supply is 16.47V.
Formula. Output = VCC/2+0.42 Volt
The amplifier has very low distortion. THD is like 0.00007%
The amplifier is very fast. Upper bandwidth is like 3.5MHz.
This is typical for CFA, Current FB Amplifiers.
There are 2 potentiometers.
R6 for adjusting bias to 900mA.
R11 for adjusting the voltage at the output. Set the output like VCC/2.
Actually optimal output voltage at 32.1V supply is 16.47V.
Formula. Output = VCC/2+0.42 Volt
The amplifier has very low distortion. THD is like 0.00007%
The amplifier is very fast. Upper bandwidth is like 3.5MHz.
This is typical for CFA, Current FB Amplifiers.
Good amplifier.
Class AB 100W is impressive.
My BULLET needs rather big heatsink.
Left channel 29W to cool down.
Right channel 29W to cool.
Sum like 58 Watt heatsink.
Class AB 100W is impressive.
My BULLET needs rather big heatsink.
Left channel 29W to cool down.
Right channel 29W to cool.
Sum like 58 Watt heatsink.
This is the AC Analysis.
Shows that BULLET has bandwidth: 2.5 Hz - 3.5 MHz.
The phase also looks good.
Shows that BULLET has bandwidth: 2.5 Hz - 3.5 MHz.
The phase also looks good.
Have included the transistors that I finally settled on. Low gain and Hfe types used, I'm sure much to horror of the DC to light brigade but they sound better in this design. Regarding the choke you can adjust this to a value which you find best for listening or do away with it if your load is well behaved. also the voltage rail is slightly higher at 95V.
Your design is not the same as the OP. Why not starting your own thread?
Member
Joined 2009
Paid Member
Oh, why do I bother..... Jacques, the two circuits are topologically the same. There are differences in the transistor selection and the DC offset application but we both have a single ended transistor feeding a current sourced VAS into a lateral output stage, feedback to the input transistor emitter. Both circuits are powered by a single polarity supply, need I go on. The fact that one is Class A and the other is Class AB is merely due to how much you turn the biasing resistor. I posted the design in support of Lineup's amp topology, saying there are others using this unconventional topology. I was not saying you must change this and that etc.
Now to the Sesebe and Walor, lateral mosfets have a negative temperature coefficient above 100mA so you will not get thermal runaway if you bias above 100mA. Now to qualify that: providing you have adequate heat sinking and operating a single pair of mosfets. Now you say but manufacturers put source resistors on their amps, well they don't want the amp returned for any reason. It costs them money. When operating multiple pairs, source resistors stabilises the current, which is the case with bipolars and vertical mosfets, but with laterals it's more of a case of poor matching rather than the device itself. Source resistors alleviate the need for matching. The manufacturer decides they rather spend 20cents on a source resistor than $15 on extra transistors and the time spent matching. I have operated matched lateral mosfet amps since 1979 without source resistors and not one has exhibited thermal runaway.
Now there are other reasons for not using source resistors, namely when used, you reduce the transconductance by introducing local negative feedback. Without source resistors you extend the Class A envelope, I suggest you read the words of the Great Oz Nelson Pass on this. In my experience and many others on this forum, it simply sounds better.
Finally, I would say there is a different set of rules between manufacturers and DIYers. Primarily for a manufacturer is cost, reliability and look; if they get sound into that they are on a winner. For a DIYer it is sound, safety and satisfaction, all the rest is a bonus.
Apologies to Lineup taking up space, great design by the way. Keep them coming.
Now to the Sesebe and Walor, lateral mosfets have a negative temperature coefficient above 100mA so you will not get thermal runaway if you bias above 100mA. Now to qualify that: providing you have adequate heat sinking and operating a single pair of mosfets. Now you say but manufacturers put source resistors on their amps, well they don't want the amp returned for any reason. It costs them money. When operating multiple pairs, source resistors stabilises the current, which is the case with bipolars and vertical mosfets, but with laterals it's more of a case of poor matching rather than the device itself. Source resistors alleviate the need for matching. The manufacturer decides they rather spend 20cents on a source resistor than $15 on extra transistors and the time spent matching. I have operated matched lateral mosfet amps since 1979 without source resistors and not one has exhibited thermal runaway.
Now there are other reasons for not using source resistors, namely when used, you reduce the transconductance by introducing local negative feedback. Without source resistors you extend the Class A envelope, I suggest you read the words of the Great Oz Nelson Pass on this. In my experience and many others on this forum, it simply sounds better.
Finally, I would say there is a different set of rules between manufacturers and DIYers. Primarily for a manufacturer is cost, reliability and look; if they get sound into that they are on a winner. For a DIYer it is sound, safety and satisfaction, all the rest is a bonus.
Apologies to Lineup taking up space, great design by the way. Keep them coming.
@Ejam. Sorry if I offended you. I just meant that your design is interesting enough to discuss its specificities (post#6) in a separate thread.
W
Then we can start ordering PCB.
The circuit is not complicated. Should be fairly easy to make.
The thing with this amplifier is the transformer and the big heatsink.
We need someone to make PCB files, Gerber.
Then we can start ordering PCB.
The circuit is not complicated. Should be fairly easy to make.
The thing with this amplifier is the transformer and the big heatsink.
A little change.
The capacitor in the input filter C9 should be 68pF. Not 100pF.
This will make the filter works at it should.
Also an input volume potentiometer should be 10kOhm - 25kOhm.
10kOhm is to prefer.
Not 50kOhm!
The capacitor in the input filter C9 should be 68pF. Not 100pF.
This will make the filter works at it should.
Also an input volume potentiometer should be 10kOhm - 25kOhm.
10kOhm is to prefer.
Not 50kOhm!
Last edited:
We do not normally edit or delete posts for technical inaccuracies unless there are safety implications.
Many silently follow threads and would either miss changed content or alternatively think some conspiracy theory was going on.
If something is wrong or has been updated then quote the post that is incorrect together with a correction.
Cool😊 I just thought that if I was fast😄 And no we wouldnt want diyaudio to cause paranoia.
Cheers!😊
Cheers!😊