Class A amplifier for my speaker testing and Audition

[arj]

Infact what i noticed is something different. For the first few minutes about 10 minutes the amplifer SQ is not that smooth. But after the amplifier heats up it sound very very open and smooth and the entire sound stage has lot of presence and warmth. I feel after getting warmed up and after the quiscent current is stabilized it becomes class A. Before the warm-up it works in class B as the current is yet to build-up in the circuits.... just like a Tube amplifier.

The heat sink is kept up-side down because the entire bottom portion is ventilated and the heat sink is directly touching the chassis. Hence the chassis also acts like an heat sink and the heated chassis escapes the heat to the ambient. This could be one of the reason the amplifier not getting hot.

Hi hari, great to see. From what I know the power expended as heat depends on the efficiency of the speaker...the more the speaker draws, the less dissipated as power.

Also class A gets to optimal sound only after a warmup of 20-30 mins..was the case with the Sugden I owned as well..


Sent from my iPad using Tapatalk

 

[aks07]

Dear Shri Iyer

The heatsink in your build seem appropriate for standing current of 120mA. And I am convinced it should be sounding darn good as well, although the mentioned current seems too low for Class A operation of a solid state power device irrespective of the voltage.

I dont have much idea on these but the last solid state Class A amplifier I built was based on a very popular John Linsley Hood design. It was 15 years ago and I vaguely remember it used a pair of 2N3055 output BJT, a standing current of approx 2.5A and rails of +- 20VDC. The estimated PO was about 12~14 WPC.

I did extensive experimentation/mods and finally ended up with a pair of nice monoblocks built on vero board. Later I gave it away to a relative who probably still has it. It was the most valve like solid state amp I ever heard. Period.

Sorry for going OT but the reason I mention it that the JLH amp was HUGE in terms of physical size, power consumption and heat generation. The total parts count was less than 20 qty each channel but still each monoblock weighed almost 15kg. I had used large heatsinks salvaged from an industrial panel box but they was still insufficient. I was forced to use a table fan aimed directly at the heatsinks to cool it down a bit.

As per books, a Class A stage stays ON throughout the 360deg of the waveform, continuously drawing a steady current. For an output stage this results in tremendous heat. The efficiency factor is very low around 15~20% in true Class A mode. Of course one can bias a device to be in shallow or deep Class A territory. Nevertheless, they remain inefficient, draw massive electric energy and waste most of it in form of heat. From my own (in)experience an 18WPC Class A amp will need heatsinks the size of the cabinet you are using !

I am finding the subject amplifier very interesting.

Warm regards

AKS

 

[aks07]

I feel after getting warmed up and after the quiscent current is stabilized it becomes class A. Before the warm-up it works in class B as the current is yet to build-up in the circuits.... just like a Tube amplifier.

No Sir. A true Class A stage will remain that right from the word go although it could be a "hot" or "cool" Class A depending on the state of bias.

In most cases the quiescent current is stabilized within a few seconds of full rail voltage availability. Over the next minutes or hours of operation it may drift a little; specially in case of a thermionic tube.

However, it is believed that the warm-up enhancement is most likely due to the electrons getting their acts in a harmony once the parts have warmed up. I too have felt it and make any changes on amp (bias etc) once it has warmed up.

Hi AKS did you mean efficient in terms of the conversion to output power by the amp ?. From the little I know the pure class A efficiency is less than 15%.. Hence input power should be at least 120 w consumption..am I right ?

Yes Sir, you are absolutely correct.
Input draw will of-course depend on output rating, efficiency, etc.

 

[omishra]

Hi Hari, nice build. Only one thing I am curious - if this has very small quotient current 120mA bias then it must be class AB and not class A yet till you increase bias upto 1A or so. Are you planning to increase bias? Is there anything I am missing?

 

[Hiten]

For sustained dynamic music it will clip and will not be good for speakers.

Sorry Hari, my mistake. The load speakers puts on an amp and its output devices capabilities will decide the clipping. My bad.

If I remember reading correctly the bias setting decides when will the amp leave class A operation and to class A/B. So if it not generating heat as most class A amp do it is leaving class A operation early. I think with proper bias setting and more heatsink you can increase output wattage in class A. You can very well turn the heatsink upside down with PCB so instead of bottom it faces top and heat desipation will be more. I think Pearraux amps do the same thing.
Regards

 

[Hari Iyer]

Hi hari, great to see. From what I know the power expended as heat depends on the efficiency of the speaker...the more the speaker draws, the less dissipated as power.

Also class A gets to optimal sound only after a warmup of 20-30 mins..was the case with the Sugden I owned as well..


Sent from my iPad using Tapatalk

I have connected a TL speaker to the class A amp which is more efficient than a vented enclosure. Also the impedance of the speaker is flat over the entire FR even at resonance. Hence the amplifier has a constant load to drive which does not vary with the music signal. This could be one of the reason the amp not getting too hot.

 

[Hari Iyer]

Sorry Hari, my mistake. The load speakers puts on an amp and its output devices capabilities will decide the clipping. My bad.

If I remember reading correctly the bias setting decides when will the amp leave class A operation and to class A/B. So if it not generating heat as most class A amp do it is leaving class A operation early. I think with proper bias setting and more heatsink you can increase output wattage in class A. You can very well turn the heatsink upside down with PCB so instead of bottom it faces top and heat desipation will be more. I think Pearraux amps do the same thing.
Regards

The maximum recommended bias for this amp is around 200mA. Currently i have set this to around `120mA. Will plan to increase this to around 150mA after some listen. I had initially planned the heatsinks as per your suggestion but am finding challenges mounting them properly without spacers. Once i am able to locate the correct spacers will turn them around.

 

[Hari Iyer]

Hi Hari, nice build. Only one thing I am curious - if this has very small quotient current 120mA bias then it must be class AB and not class A yet till you increase bias upto 1A or so. Are you planning to increase bias? Is there anything I am missing?

The max recommended bias is around 200mA for the class A operation but is just a reference point. I can be increased if there are enough cooling options available like an fan to cool the amplifier and a larger heatsink for the additional heat generated. Pls. check the link in the first post for details.

 

[Hari Iyer]

Dear Shri Iyer

The heatsink in your build seem appropriate for standing current of 120mA. And I am convinced it should be sounding darn good as well, although the mentioned current seems too low for Class A operation of a solid state power device irrespective of the voltage.

I dont have much idea on these but the last solid state Class A amplifier I built was based on a very popular John Linsley Hood design. It was 15 years ago and I vaguely remember it used a pair of 2N3055 output BJT, a standing current of approx 2.5A and rails of +- 20VDC. The estimated PO was about 12~14 WPC.

I did extensive experimentation/mods and finally ended up with a pair of nice monoblocks built on vero board. Later I gave it away to a relative who probably still has it. It was the most valve like solid state amp I ever heard. Period.

Sorry for going OT but the reason I mention it that the JLH amp was HUGE in terms of physical size, power consumption and heat generation. The total parts count was less than 20 qty each channel but still each monoblock weighed almost 15kg. I had used large heatsinks salvaged from an industrial panel box but they was still insufficient. I was forced to use a table fan aimed directly at the heatsinks to cool it down a bit.

As per books, a Class A stage stays ON throughout the 360deg of the waveform, continuously drawing a steady current. For an output stage this results in tremendous heat. The efficiency factor is very low around 15~20% in true Class A mode. Of course one can bias a device to be in shallow or deep Class A territory. Nevertheless, they remain inefficient, draw massive electric energy and waste most of it in form of heat. From my own (in)experience an 18WPC Class A amp will need heatsinks the size of the cabinet you are using !

I am finding the subject amplifier very interesting.

Warm regards

AKS

The current requirement for 18W rms with +/- 30 volts is around 2A for each channel. The heat sink i have used is meant for 30A current load used in one of a salvaged industrial power supply. Hence it should be quite sufficient for this amplifier i persume. Since the amp is currently under test am not taking any risk with them to increase the bias now. Will drive them to deeper class A after the initial test are okay and after i feel that they can take more aggressive current settings.