Hi, I first want to say im greatfull for this forum, Its so awesome to be able to read and discuss things I cant ask any of my friends, who dont know electronics, they know beer and sports tho 😉.
Im using a pnp LTP to drive a single pole VAS 2sc3788 (might change depending on advice), pre driving 2sc4793 and compliment with 2 pairs of 2sc5200 and complement final into 6ohm load.
Was using MJL3281 and compliment and MJL21994 and compliment on other projects, but after seing many use 2sc5200 on this forum I had a look at them and they 1/2 the price of MJL21194 here, and for this project they more than enough.
I want to know how much current I should set my Vas transistor at.
LTP is biased at 1.5mA per tail atm. And Final transistors at 30-40mA per pair.
Supply is 50-0-50 DC unregulated, heading for 100W into 8ohm.
I was thinking of 20mA for Vas transistor, to maximise output.
But im scared that would lead to the Vas running hot.
As I have no experience with this Id really appriciate your advice.
Im using a pnp LTP to drive a single pole VAS 2sc3788 (might change depending on advice), pre driving 2sc4793 and compliment with 2 pairs of 2sc5200 and complement final into 6ohm load.
Was using MJL3281 and compliment and MJL21994 and compliment on other projects, but after seing many use 2sc5200 on this forum I had a look at them and they 1/2 the price of MJL21194 here, and for this project they more than enough.
I want to know how much current I should set my Vas transistor at.
LTP is biased at 1.5mA per tail atm. And Final transistors at 30-40mA per pair.
Supply is 50-0-50 DC unregulated, heading for 100W into 8ohm.
I was thinking of 20mA for Vas transistor, to maximise output.
But im scared that would lead to the Vas running hot.
As I have no experience with this Id really appriciate your advice.
Last edited:
On the face of your proposal, 20mA of VAS current should be sufficient.
Predriver with a gain of 100times and output with a gain of 50times gives a maximum output of 100Apk.
More than enough to drive an 8ohms speaker to 100W and still more than enough to meet a 5times current flow for short term transients that exceed that into an 8r0 resistor by a factor of >>5times.
Is there a driver between the predriver and the output?
But look at it from the other end.
Start with the peak transient output current. Try 15Apk.
What is a typical current gain of your outputs at this current?
What is the base current required to allow 15Apk to be delivered to the load?
What if the device/s you pick performs slightly worse than typical?
Is the minimum specified hFE @ 15Apk a reasonable value to choose, for designing the output stage of the amplifier?
Now progress back through the predriver stage.
What is the peak transient current required by the output stage?
Does any additional current feed other components between the predriver emitter and the output base.
What is the peak emitter current required?
What is the gain of the predriver at that emitter current?
Can you see where I am leading you?
VAS Pq is = +ve supply voltage * Ic. That would be 50 * 20 = 1000mW. That is hot even for a To126. It needs a sink.
Predriver with a gain of 100times and output with a gain of 50times gives a maximum output of 100Apk.
More than enough to drive an 8ohms speaker to 100W and still more than enough to meet a 5times current flow for short term transients that exceed that into an 8r0 resistor by a factor of >>5times.
Is there a driver between the predriver and the output?
But look at it from the other end.
Start with the peak transient output current. Try 15Apk.
What is a typical current gain of your outputs at this current?
What is the base current required to allow 15Apk to be delivered to the load?
What if the device/s you pick performs slightly worse than typical?
Is the minimum specified hFE @ 15Apk a reasonable value to choose, for designing the output stage of the amplifier?
Now progress back through the predriver stage.
What is the peak transient current required by the output stage?
Does any additional current feed other components between the predriver emitter and the output base.
What is the peak emitter current required?
What is the gain of the predriver at that emitter current?
Can you see where I am leading you?
VAS Pq is = +ve supply voltage * Ic. That would be 50 * 20 = 1000mW. That is hot even for a To126. It needs a sink.
Last edited:
Thanks for your advice and also the breakdown, it helps.
I aslo got to 1W on to126 and was worried about temperature,
Ill heatsink it as you advice.
Hehe now to recap on heatsink calculations.
I aslo got to 1W on to126 and was worried about temperature,
Ill heatsink it as you advice.
Hehe now to recap on heatsink calculations.
I think you meant 2SC5200 not 2SC3788. You may have found them cheaper, but beware it is a very commonly faked device.
Another question pls.
2sc3788 vas : VceQ = 50V, with max signal running to close to 0 and 100V.
Do I use derating curve on 50V im guessing average voltage over ce.
or do I use Vce(max) approx 100V.
I could just use 100V to be safe but Im using this as learning experience and
so want to use the proper value and calculate smallest heatsink,
even if i end up using a slightly lower c/w heatsink.
2sc3788 vas : VceQ = 50V, with max signal running to close to 0 and 100V.
Do I use derating curve on 50V im guessing average voltage over ce.
or do I use Vce(max) approx 100V.
I could just use 100V to be safe but Im using this as learning experience and
so want to use the proper value and calculate smallest heatsink,
even if i end up using a slightly lower c/w heatsink.
Attachments
Your LTP transistors will dissipate 75mW each , that s quite high ,
you can reduce the current to 1mA/each without entailing the perfs
providing you use a compensation cap that is reduced by a 1.5 ratio.
As VAS current 8 to 10mA are enough given the high beta
of your drivers + power transistors.
Even with such low currents , it would be cautious to give
the VAS and its eventual current source a little heatspreader.
you can reduce the current to 1mA/each without entailing the perfs
providing you use a compensation cap that is reduced by a 1.5 ratio.
As VAS current 8 to 10mA are enough given the high beta
of your drivers + power transistors.
Even with such low currents , it would be cautious to give
the VAS and its eventual current source a little heatspreader.
- Status
- Not open for further replies.