thanks Marcel!
Seems that i've made a mistake with drawing the schematic!
I did check my pcb's R20 is 1 kohm instead of 62 ohm. I couldn't believe the amp would play with the 62 ohm value for R20.
So the current through the diode string and the resistor will be something about 11,4 mA.
best regards,
HB.
Seems that i've made a mistake with drawing the schematic!
I did check my pcb's R20 is 1 kohm instead of 62 ohm. I couldn't believe the amp would play with the 62 ohm value for R20.
So the current through the diode string and the resistor will be something about 11,4 mA.
best regards,
HB.
The temperature coefficient of the base-emitter voltage of a transistor or the forward voltage of a semiconductor diode can be calculated as the difference between the base-emitter or forward voltage and the extrapolated bandgap voltage, divided by the absolute temperature ((VForward-VBG,extrapolated)/T).
For a silicon transistor or diode, VBG,extrapolated~=1.23V at room temperature. With 0.63V forward voltage at 300 kelvin, you find a temperature coefficient of -2mV/K. It will be at bit less (I mean closer to zero) at high current densities (high Vforward) and a bit more negative at low current densities (low VForward), but it is always in the order of -2mV/K for practical currents.
The bias voltage between the bases of T8 and T7 should have such a temperature coefficient that it compensates for the temperature dependence of T7, T8, T11, T12, T9 (?) and T10. You can achieve that either by putting four diodes between the bases of T8 and T7, or by using less than four diodes and giving the current through R28 the right temperature coefficient. The problem with using four diodes is that you may not be able to achieve optimal class AB biasing of the output stage for any potmeter setting. The total voltage drop across the four diodes and the resistor may be higher than optimal for all potmeter settings.
By the way, the temperature compensating diodes should have as close a thermal contact with T7...T12 as possible; at least you should somehow mount them on the same heatsink.
Another thing: with 12V or more across R20, you probably reduce the possible positive signal swing at the output significantly. You can achieve higher positive peak output voltages if you reduce the voltage across R20 and use a lower valued resistor.
I hope this helps you,
Marcel
For a silicon transistor or diode, VBG,extrapolated~=1.23V at room temperature. With 0.63V forward voltage at 300 kelvin, you find a temperature coefficient of -2mV/K. It will be at bit less (I mean closer to zero) at high current densities (high Vforward) and a bit more negative at low current densities (low VForward), but it is always in the order of -2mV/K for practical currents.
The bias voltage between the bases of T8 and T7 should have such a temperature coefficient that it compensates for the temperature dependence of T7, T8, T11, T12, T9 (?) and T10. You can achieve that either by putting four diodes between the bases of T8 and T7, or by using less than four diodes and giving the current through R28 the right temperature coefficient. The problem with using four diodes is that you may not be able to achieve optimal class AB biasing of the output stage for any potmeter setting. The total voltage drop across the four diodes and the resistor may be higher than optimal for all potmeter settings.
By the way, the temperature compensating diodes should have as close a thermal contact with T7...T12 as possible; at least you should somehow mount them on the same heatsink.
Another thing: with 12V or more across R20, you probably reduce the possible positive signal swing at the output significantly. You can achieve higher positive peak output voltages if you reduce the voltage across R20 and use a lower valued resistor.
I hope this helps you,
Marcel
mikek said:i would suggest losing the rudimentary short-circuit protection pretty sharpish....i can almost gaurantee its audible in normal use,....and doesn't do a good job of protecting the transistors either...
replace with a single slope variant...at the very least...
I don't know if it's audible or not, I finally have a scope and a generator and I will check it out soon!
What do you mean with single slope variant??
I want to make a full protected output, because the amps will be used on parties etc... anybody have other good ideas/circuits??
best regards,
HB.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.