Tried with bootstrap 1.2k+1.5k+100uF
Was not any better - THD 0.092%
So my conclusion is that current source is better than bootstrap.
Was not any better - THD 0.092%
So my conclusion is that current source is better than bootstrap.
I knew its not going to be better, i just wanted to know how much worse it is, the less transistors the better (at least for me)
Yes, that's right.
If we want it simple we should go for bootstrap. That is the benefit.
THD 0.092% is not bad in reality.
I have tweaked the original circuit:
Current source 2.17mA
Output at 6.65V
Bias 0.4A
Result is THD 0.046%
That is the lowest I have manage so far.
If we want it simple we should go for bootstrap. That is the benefit.
THD 0.092% is not bad in reality.
I have tweaked the original circuit:
Current source 2.17mA
Output at 6.65V
Bias 0.4A
Result is THD 0.046%
That is the lowest I have manage so far.
Last edited:
To get lower distortion we can lower the gain.
In my schematic this is to reduce R9 from 220k to 100k.
It is even possibility to get lower. Say R9 to 68k ...
As we use a preamp this is no problem.
See my circuit for this.
Distortion was lowered from THD 0.046% to 0.022%
That way we get closer to hifi.
In my schematic this is to reduce R9 from 220k to 100k.
It is even possibility to get lower. Say R9 to 68k ...
As we use a preamp this is no problem.
See my circuit for this.
Distortion was lowered from THD 0.046% to 0.022%
That way we get closer to hifi.
Last edited:
I think you’re at diminishing returns here. Might get it down a wee bit more at unity gain. Then your sound card won’t even drive it fully running rail to rail.
my idea was to use the transistor to make an "half differential pair" simmilar to the one in the jlh1969 amplifier.
You already designed a better amplifier in this thread
Just change the tips for bd139/140 and the driver tips for the mpsa/other generic transistors.
Just change the tips for bd139/140 and the driver tips for the mpsa/other generic transistors.
Will try this. Attached is LTspice simulation (schema from message #229) which gives bode plot, seems to be stable using standard value 22pF comp cap and this cap can be reduced to 15pF, without comp cap there is some HF instability for sure, like measured from prototype. Uses standard transistor available at LTspice, have no model for 2SA992, which makes all this simulation not so much fancy, but bode plots are useful in any case and simulation is quite near to reality, which is good. Gives also low distortion like original sim made by 'lineup'.
This one also sounds good and works well with simple one transistor + zener regulator + laptop supply. Of course it sounds little bit better if using "discrete feedback reg" using one zener + j-fet + three BJT's. This is not a big surprise, regulator which measures better should also sound a little bit better, no magic or 'snake-oil' here.
This one is simplest setup, but still no tone controls. Volume control can be between preamp and one-watt amp.
This one also sounds good and works well with simple one transistor + zener regulator + laptop supply. Of course it sounds little bit better if using "discrete feedback reg" using one zener + j-fet + three BJT's. This is not a big surprise, regulator which measures better should also sound a little bit better, no magic or 'snake-oil' here.
This one is simplest setup, but still no tone controls. Volume control can be between preamp and one-watt amp.
Measured voltage over R10 is 283mV which makes current around 1.9mA and over the 150ohm in series with 1N4001 have little bit over 300mV, quite near 2mA. Sometimes sims are little bit different than reality, and also all those 1N4001 don't have same Uf, I have measured around 10 pieces to find some which have Uf near 2N3906 Vbe voltage, all this is not so simple as it looks here, but to write this all takes time. There is always some variation between transistors or diodes, those from same wafer may be quite close each other, but who knows this if you buy only some parts for hobby use ?
Amp itself is simple but can be done better if choosing parts before building. Those cheap component tester and some DVM with diode measuments are very useful. I have used also Fluke115 which again is "an overkill" solution, any cheap DVM with diode or transistor check (you name it 🙂 ) will do the job.
Can try also this amp from message #254 when I have time, next is the other preamp with tone controls. To make all these protos takes time... but it's also interesting to listen little bit different solutions, highly recommended also to build and listen something if possible 😃
To be exact, this simple one-Watt amp without any extras sounds good enough and for desktop amp no more power is needed. Also in a small listening room with sensitive speakers one watt is a lot.
To be exact, this simple one-Watt amp without any extras sounds good enough and for desktop amp no more power is needed. Also in a small listening room with sensitive speakers one watt is a lot.
