Might be relevant to your post but my SE's have 0R1 , not the 0R2 you have so I've set the emitter resistors bias current to 2mV (20mA)
Douglas Self proposes value of bias in voltage units: 9-12 mV, if I remember correctly the exact value, but I'm sure it is in mV, not mA.
It would be the best to observe changes with osciloscope.
Yes, thanks, I get that. The problem is that these the units seem to have different emitter resistances so the if set to 9-12mV it has a very significant impact on the bias current. I'm suggesting that a one size may not fit all scenarios.
Also, I do not see the artifacts on the output waveform that are shown the the link above. Seems very clean but I am still still experimenting with higher voltage supply rails.
Like you suggested I reduced the gain 🙂
🙂
Also, I do not see the artifacts on the output waveform that are shown the the link above. Seems very clean but I am still still experimenting with higher voltage supply rails.
Like you suggested I reduced the gain 🙂
🙂
Last edited:
I have verified: Douglass Self suggests: 10mV across emitter resistors when amplifier is at working temperature. Otherwise, you can determine this value by using a distortion analyser.
This isn't rocket science: use 10 mV as a rule of thumb.
This isn't rocket science: use 10 mV as a rule of thumb.
Correction: but not for this amplifier.
I'm working simultaneously with four different amplifiers now and have confused this with something else. Correct value for this amplifier is:
I have a question about this modification. Why use a potentiometer to adjust the quiescent current value of the output transistors instead of 4k7 and not 9K1?
We are getting mV and mA mixed up here, and on top of that, some of these amplifiers are equipped with emitter resistors with a value of 0.22R and a value of 0.1R. If we are talking about the quiescent current value, the value of the emitter resistor does not matter. But if we are talking about the voltage measured on the emitter resistor, the value of the resistor matters a lot. 10mV on a 0.22R resistor means a quiescent current value of 45mA. But 10mV on a 0.1R resistor means 100mA quiescent current for one transistor, and that is a big difference. If Berluscony recommends a value of 26mV and I see in his picture of the board on • #360 the emitter resistors with a value of 0.22R, then he recommends a quiescent current value of 118mA per transistor.
I would be very interested in what the original value of the quiescent current is for the board in its original state, i.e. with a resistance value of 4k7 and in the second position 9k1 or a value of 8k2 - there are two versions, as well as with the values of the emitter resistors.
I did not notice this information anywhere, only some write the values that they set by changing the values of the resistors in these positions.
We are getting mV and mA mixed up here, and on top of that, some of these amplifiers are equipped with emitter resistors with a value of 0.22R and a value of 0.1R. If we are talking about the quiescent current value, the value of the emitter resistor does not matter. But if we are talking about the voltage measured on the emitter resistor, the value of the resistor matters a lot. 10mV on a 0.22R resistor means a quiescent current value of 45mA. But 10mV on a 0.1R resistor means 100mA quiescent current for one transistor, and that is a big difference. If Berluscony recommends a value of 26mV and I see in his picture of the board on • #360 the emitter resistors with a value of 0.22R, then he recommends a quiescent current value of 118mA per transistor.
I would be very interested in what the original value of the quiescent current is for the board in its original state, i.e. with a resistance value of 4k7 and in the second position 9k1 or a value of 8k2 - there are two versions, as well as with the values of the emitter resistors.
I did not notice this information anywhere, only some write the values that they set by changing the values of the resistors in these positions.
I set the bias on mine by measuring IMD and IIRC I ended up with around 120mA per transistor. It was still not as clean as I wanted but it was low enough for the amp to sound good at normal levels.
Transistors aren't exact copies of each other, rail voltages impact bias current/voltages, temperature too. Besides, bias may change significantly in the neighborhood of optimal settings. All together render impossible to use the same pre-calculated resistor values. It is therefore inevitable to use measurements for each set of variables that influence bias. I was hoping it wasn't the case but I have decided to use potentiometers. Use protection for at least preliminary testing to gain some feeling how currents change. Sometimes changes may be sudden and violent. Be prepared to disconnect power supply in case of emergence. Be careful especially with L20 V9.2 and V10. They have incredibly poor thermal design. These are simply insufficient to sustain higher rail voltages and output power. I have solution how to make them attain higher power output - safely.
This is exactly proper way to set proper bias value. You can further improve this amplifier by increasing feedback ie. reducing gain. I have found that noise occurs asymetrically and rides on positive side of signal output. This makes sound dull but the noise is hidden at lower output levels. This makes it difficult to identify the real source of the problem.
Last edited:
Voltage should be observed. Cordell makes this clear in his book which I have cited in #365. Cordell is very clear and uses precise language. Unambiguously. Voltage.
To berlusconi,
You didn't answer the question asked, maybe my English is too bad. So once again and better - why put the potentiometer in the position of the 4k7 resistor for setting the quiescent currents? I'm asking why not put the adjusting potentiometer instead of the 9K1 resistor. Why should it be instead of 4k7?
You didn't answer the question asked, maybe my English is too bad. So once again and better - why put the potentiometer in the position of the 4k7 resistor for setting the quiescent currents? I'm asking why not put the adjusting potentiometer instead of the 9K1 resistor. Why should it be instead of 4k7?
My circuit consultant gave me the opposite answer, that in the event of a potentiometer failure, it would be dangerous at position 4k7.
It is about the current flowing at rest. We only talk about voltage because we have Ohm's law here, which gives the relationship between current, voltage and resistance. The current flowing directly can only be measured by disconnecting the circuit and inserting an ammeter into the circuit. In this case, we can measure the voltage drop across a resistance of known value without disconnecting the circuit and simply calculate the current I=U/R
I asked another expert about amplifier circuits, and he gave me the opposite answer - that there is a danger when placing a potentiometer instead of 9k1. So it seems that Berlusconi's recommendation is correct. Put the potentiometer for adjusting the quiescent period in the 4k7 resistance position
@PANDINUS 👍
Thank you very much for this information.
Safety is important when working with electrical devices and it can be almost entirely in our hands, except when something out of our control goes wrong. That's the life and we have to accept the risks if we want to improve.
I have considered this dilema and have arrived to the following conclusion: With proper soldering techniques and components of adequate quality it is virtually all the same whether we decide to adjust 9K1 or 4K7 resistor. Important is that the ratio of their values determines the bias.
As mentioned before, there is an interesting conversation about this question in another related thread >>HERE<<
Thank you very much for this information.
Safety is important when working with electrical devices and it can be almost entirely in our hands, except when something out of our control goes wrong. That's the life and we have to accept the risks if we want to improve.
I have considered this dilema and have arrived to the following conclusion: With proper soldering techniques and components of adequate quality it is virtually all the same whether we decide to adjust 9K1 or 4K7 resistor. Important is that the ratio of their values determines the bias.
As mentioned before, there is an interesting conversation about this question in another related thread >>HERE<<