I would like to bias LM4702 into class A with Bias current of 500mA or may be 1A but will have very large heatsinks measuring 300mm x 200mm and fins will be about 50mm length.
I would like to know that the supply voltage be if its driven in class A.
the idea is to put a switch where we can switch the amp in Class A and Class AB
but the question is when the amp is delivering about 100W in 8 ohm with 2sc5200 1943 pair then how much it can deliver in class A 20W?
I would like to know that the supply voltage be if its driven in class A.
the idea is to put a switch where we can switch the amp in Class A and Class AB
but the question is when the amp is delivering about 100W in 8 ohm with 2sc5200 1943 pair then how much it can deliver in class A 20W?
The important thing here is the bias current. With a bias of 500mA, Class A power into 8 ohms is 4 Watts, increasing to 16 Watts with 1 A bias.
Absolutely right Knuth!
The definition of class A is that the output devices are not turning off during the signal cycle.
With 500mA bias, max output peak current in class A is 1A, which is 0.7A RMS (1A * sqrt of 2). Power in 8 ohms is then 0.7^2 * 8 = 4W.
The supply voltage is irrelevant as long as it is high enough for the required class A output level.
1A peak in 8 ohms is 8V, so this 4W class- A can be delivered with a supply voltage just above +/-8V...
jan
The definition of class A is that the output devices are not turning off during the signal cycle.
With 500mA bias, max output peak current in class A is 1A, which is 0.7A RMS (1A * sqrt of 2). Power in 8 ohms is then 0.7^2 * 8 = 4W.
The supply voltage is irrelevant as long as it is high enough for the required class A output level.
1A peak in 8 ohms is 8V, so this 4W class- A can be delivered with a supply voltage just above +/-8V...
jan
100W in 8 ohm requires a bit more than +/- 40V rails.
If your heatsinks are big enough, then a rough guess is that it should be possible to have a bias current between 1 - 2 amp.
2 amps of bias current with no input signal, gives a power dissipation of 80 watt in the NPN and 80 watt i the PNP transistor with +/- 40V rails.
There might be some problems with derating at high Tcase temps:
http://www.st.com/internet/com/TECHNICAL_RESOURCES/TECHNICAL_LITERATURE/DATASHEET/CD00249630.pdf
I assume that you only use one 2SC5200 and one 2SA1943 in your pushpull output stage correct ?
If your heatsinks are big enough, then a rough guess is that it should be possible to have a bias current between 1 - 2 amp.
2 amps of bias current with no input signal, gives a power dissipation of 80 watt in the NPN and 80 watt i the PNP transistor with +/- 40V rails.
There might be some problems with derating at high Tcase temps:
http://www.st.com/internet/com/TECHNICAL_RESOURCES/TECHNICAL_LITERATURE/DATASHEET/CD00249630.pdf
I assume that you only use one 2SC5200 and one 2SA1943 in your pushpull output stage correct ?
Last edited:
I think the OP was looking to compare the classes for his own interest. Without going to high power, Douglas Self's Trimodal amplifier offers A, B, and AB classes at the flick of a switch. A full schematic article and description is carried in his "Audio Power Amplifier Construction Handbook" and "Self on Audio" books, for anyone interested.
This is a worthwhile, finessed project delivering 20-25W/8R in all modes and is more like the full deal for the OP.
http://www.douglas-self.com/ampins/ampins.htm
This is a worthwhile, finessed project delivering 20-25W/8R in all modes and is more like the full deal for the OP.
http://www.douglas-self.com/ampins/ampins.htm
Last edited:
for more fun you need the calcs for realistic loads
loudspeakers aren't pure resistors, even "polite" multiway often have |Zmin|< Rnominal/2
then you can adjust power dissipation for phase angle
add reserve for "pumping" resonace, sudenly reversing phase - the current peak can be >5x the Rnominal calc
loudspeakers aren't pure resistors, even "polite" multiway often have |Zmin|< Rnominal/2
then you can adjust power dissipation for phase angle
add reserve for "pumping" resonace, sudenly reversing phase - the current peak can be >5x the Rnominal calc
Rule of thumb is that the quiescent dissipation for class A is about 4 times the max output power (assuming the power supply voltage is dimensioned for that but not higher).
So that's 400W for 100W class A in 8 ohms.
If you want to keep the temp rise down to say 25 degrees C, which means a heatsink temp of 47 degrees at room temp of 22 C, you need a sink with no more than 25/400 = 0.0625 deg/W.
That will need a lot of (forced) air cooling.
jan
So that's 400W for 100W class A in 8 ohms.
If you want to keep the temp rise down to say 25 degrees C, which means a heatsink temp of 47 degrees at room temp of 22 C, you need a sink with no more than 25/400 = 0.0625 deg/W.
That will need a lot of (forced) air cooling.
jan
If I was to design a 100W into 8ohm class A, I would consider to make it a balanced design. This reduces the rail voltage to half, but off cause doubles the peak current needed. (Amplifiers see 4ohm in BTL)
Rail voltage on output stage can be reduced to a bit above +/- 20volt.
The bias current needed in each Pushpull stage would be 2.5A (20V into 4 ohm=5A or 40V into 8ohm=5A). Power dissipation in each BJT device will be 2.5Ax20V =50W and a total of 200W for the four transistors. (One balanced channel)
I like that both rails of power supply is equally loaded with a balanced BTL output stage.
No nasty rectified signal currents flow in the PSU. This would also be almost hum free, even in a non negative global feedback amplifier. Also the high bias current gives a lower output impedance, compared to a Class AB. But in BTL mode the output imp. is twice compare to one stage.
Cheers
Rail voltage on output stage can be reduced to a bit above +/- 20volt.
The bias current needed in each Pushpull stage would be 2.5A (20V into 4 ohm=5A or 40V into 8ohm=5A). Power dissipation in each BJT device will be 2.5Ax20V =50W and a total of 200W for the four transistors. (One balanced channel)
I like that both rails of power supply is equally loaded with a balanced BTL output stage.
No nasty rectified signal currents flow in the PSU. This would also be almost hum free, even in a non negative global feedback amplifier. Also the high bias current gives a lower output impedance, compared to a Class AB. But in BTL mode the output imp. is twice compare to one stage.
Cheers
yeah i thought it for a pair of output devices..
LME49811 is said to be built especially for class A amplification more for class A designs..
even if im able to get 30W in 8 ohms is still more than required for my application...
I was going through the site
Gain-X2 - LME49811 15-150Watt Gainclone amplifier
what should be the VA ratings of the transformer to get the 20W of class A power?
LME49811 is said to be built especially for class A amplification more for class A designs..
even if im able to get 30W in 8 ohms is still more than required for my application...
I was going through the site
Gain-X2 - LME49811 15-150Watt Gainclone amplifier
what should be the VA ratings of the transformer to get the 20W of class A power?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.