20000 uF , it seems to me much. A large capacitor take some time to charge and discharge, with big value, the sound become lax. how did you make the decoupling ?
I saw the PCB uses 2N3906/2SD669/2SD1047. The output transistor must be in synergy with the other transistor, especially with such a fast D1047 (20MHz). So when you compared those transistors (1047/3055/15003/359), did you use D669 for all? Any bias adjustment?
Yes, 2sd669 and 2n 3904 are very good transistor for this amp. I made a final bias adjustement with R2 = 910 ohm with 1047. Bias current is 1.2 A (the best sound) at V = 33v. 15 w power output in 8 ohms.
R5 is fixed on this kit, i made some additional holes in the pcb an put a 150k precision trimmer instead the 100k resistor to adjust V/2.
What is important also and what is not on the original plan is to add a resistance of 0.47 ohm between V + and the collector of the transistor TR2. The sound becomes more articulated and less tightened.
I forgot : TR3 and TR4 should not be too fast if not there will be oscillation problems which arrive quickly with this assembly because input and output are in phase.
Last edited:
about 5 to 10 times the RC time constant.
For a 8r0 dummy load (with the speaker switched out) that would be around 5*0.02F*8r0 = 0.8s to 10*0.02F88r0 = 1.6s
You could switch the speaker in after ~2seconds.
one requires that the coupling capacitor passes the audio signal.
If one wants to pass the 20Hz to 20kHz audio band then one usually looks for F-3dB turn over frequencies around a decade away from the required passband, i.e. 2Hz to 200kHz.
8ohms and 20mF has an F-3dB of ~1Hz.
That allows this output capacitor to be a comfortable octave outside the input filter which is the passband limiter.
The input filter is the 4u7F into 100k||100k||base input impedance of the first 3906. Maybe 40k & 4.7uF giving F-3dB ~0.85Hz.
OOPs, the input filter is set so low that the output filter is now acting as the extreme LF filter and that will create a AC voltage drop across the coupling capacitor and maybe introduce avoidable distortion. It is too small for the 4u7F at the input.
Can someone estimate the input impedance of the first transistor? Maybe it's a lot lower than 200kohms.
Last edited:
the DC input resistance is 100k||100k.
That should read 50k, if the resistors are reasonably accurate.
The AC input impedance must be lower, not higher.
That should read 50k, if the resistors are reasonably accurate.
The AC input impedance must be lower, not higher.
My trimming resistor is 150k not 100k because I can not adjust V / 2 correctly with 100k. It is normal to measure 55 k in this case , no ??
See post 3224
Then 100k||150k||660k gives your 55kohms.
That 660kohms seems very high for the transistors input impedance. But maybe that is a good value for this circuit and whatever hFE that the 3906 has.
I am surprised you have to offset the input so far from 50% of Vcc
150 as part of (100+150) is 60% !!!!
Ah ! there's another 39k in there.
the DC at the input is 150/(150+100+39) = 51.9% of Vcc
But the 39k is decoupled at AC, so does not affect the AC input impedance
That 660kohms seems very high for the transistors input impedance. But maybe that is a good value for this circuit and whatever hFE that the 3906 has.
I am surprised you have to offset the input so far from 50% of Vcc
150 as part of (100+150) is 60% !!!!
Ah ! there's another 39k in there.
the DC at the input is 150/(150+100+39) = 51.9% of Vcc
But the 39k is decoupled at AC, so does not affect the AC input impedance
Last edited:
The trimmer is not set at 150 but about somewhere between 120 and 130 k. 3906 hfe is about 300 if I remember correctly.
I was also surprised to have to increase the resistance, it was the first time in ten jlh I mounted.
But this is not a problem for the proper functioning of the amp, I am very happy with it !
I was also surprised to have to increase the resistance, it was the first time in ten jlh I mounted.
But this is not a problem for the proper functioning of the amp, I am very happy with it !
Last edited:
because it's a trimmer, not a resistor and I did not specify what value was set.
Last edited:
Hello,
I am thinking of building the 1969 version of the amp, and I have nice 20V AC transformers available. And I'm thinking what kind of PSU to build basic or regulated. If it's regulated then the voltage provided will be lower.
So I am thinking now what difference in power into 8ohms will there be comparing 24V and 28V power supply voltages?
I am thinking of building the 1969 version of the amp, and I have nice 20V AC transformers available. And I'm thinking what kind of PSU to build basic or regulated. If it's regulated then the voltage provided will be lower.
So I am thinking now what difference in power into 8ohms will there be comparing 24V and 28V power supply voltages?
Neglecting the device or supply voltage drops, the maximum power is proportional to the square of the voltage.
The ratio is (28x28) /(24x24) = 1.36. So the higher voltage will give around a third more power, or +2.7dB.
Last edited:
Well, I understand Ohms law, but I am confused with this amplifier. It is titled as a 10 W amplifier. 10 W into 8 ohm is ~9 V 1.125A. 9Vrms has ~12.5 V peaks. To achieve that output I need ~12.5*2=25V power rail. So 25V PSU is needed for 10W. Is there any sense in my thoughts?
Yes, 10W = Vrms * Vrms / 8 Ohms, so Vrms = 8.94Vrms
Vp-p = 2 x 1.414 x 8.94Vrms = 25.3Vp-p
Add to this the device drops, and the supply regulation drop, at full power.
Last edited:
Ok, so 27 V PSU suggested in the original amplifier was to accomodate for device drops.
Thank you, rayma
Hi
You can just get 30V unregulated from your 20V transformers, but there may be some ripple on the power supply lines. So you should be able to get 10W, but hum reduction may need some careful filtering. Ideally I would suggest a 30V transformer, rectified=50V, then the series pass transistors in a regulator would drop 20V.
Do your 20V transformers have a centre tap. You might put 20V+10V together, but they would need to be able to provide the current needed for both channels in stereo.
You can just get 30V unregulated from your 20V transformers, but there may be some ripple on the power supply lines. So you should be able to get 10W, but hum reduction may need some careful filtering. Ideally I would suggest a 30V transformer, rectified=50V, then the series pass transistors in a regulator would drop 20V.
Do your 20V transformers have a centre tap. You might put 20V+10V together, but they would need to be able to provide the current needed for both channels in stereo.
