According to simulations, the lower the value of R17, the lower the distortion.
The quality of the pot is determinant off course but it is not worst than it's effect in the input I guess.
As usual I would like to take this build as a lesson because I feel it is not a very good idea to just build things without fully understanding the concept.
Can you please describe the circuit using plain English language ?
What is the purpose of the building block Q6 Q8 Q9 Q10... and why is there a zobel (C2 R39) ?
My first GainWire preamp with a current conveyor as a building block was described here http://www.diyaudio.com/forums/analog-line-level/223835-current-conveyor-voltage-amplifier.html and maybe you could find your answers there, but I think I described a current conveyor principle here too.
C2 and R39 are not a Zobel but shunt compensation.
Regarding the pot if used instead R17 it's passing higher current then if in input, so ordinary pot could deteriorate fast.
I am now studying your shunt and can not see the putpose of the 8.2v zener.
Why is it there and why do we need the diode D5 1N4248 connected to the output mosfet gate ?
Why is it there and why do we need the diode D5 1N4248 connected to the output mosfet gate ?
This is just for mosfet protection in case of Vgs overvoltage. No any action in normal operation.
PS. that was described somewhere in the thread.
I know I should reread the whole thread but that is the problem with the long threads.... there is not an easy index.
Why did you connect Q1 (BC550) emitter to gnd using a diode ?
Why did you connect Q1 (BC550) emitter to gnd using a diode ?
D7 and D8 increase the available voltage for CCS formed with J1(2SK170) and D1 and D2 are for protection.
Perfect
Nice trick to raise the base voltage of the error amp 🙂
Thank you so much.
Nice trick to raise the base voltage of the error amp 🙂
Thank you so much.
Last edited:
Now ordering a suitable TX.
What should be the ideal VAC ?
What is the minimal dropout in the shunts ?
Can I use 18VAC TX to get the needed 15VDC after the shunts ?
What should be the ideal VAC ?
What is the minimal dropout in the shunts ?
Can I use 18VAC TX to get the needed 15VDC after the shunts ?
30 to 50 VA, 2x 15V AC. The shunt consumes about half the current.
18 V AC could be to much, maybe if you use higher heatsinks (bigger).
I will use taller heatsinks V PR32 - 38,1: Special heat sink, 38.1 x35, 0x12, 7mm, 11K - W at reichelt elektronik
V PR32/38,1 :: Special heat sink, 38.1 x35, 0x12, 7mm, 11K / W
IMO mosfets like to have more voltage headroom in order to be more linear... and maybe I will increase current (just a bit).
V PR32/38,1 :: Special heat sink, 38.1 x35, 0x12, 7mm, 11K / W
IMO mosfets like to have more voltage headroom in order to be more linear... and maybe I will increase current (just a bit).
You've chose right. For DN2530 5V is good headroom, but more won't hart, just more power dissipation.
LT1022 was just for simulation, I think that on the schematic was also LF411, this is cheap and good alternative.
Today I measured 100 BC550C and 100BC560C.
Unfortunately I found only
20 x BC550 matched to 680Hfe and
20 x BC560 matched to 640Hfe
Hope this Beta differences do not create difficulties in the mirrors, the input or the output buffer.
Unfortunately I found only
20 x BC550 matched to 680Hfe and
20 x BC560 matched to 640Hfe
Hope this Beta differences do not create difficulties in the mirrors, the input or the output buffer.
Way you say unfortunately, that is good Hfe matching, but more importantly is Vbe matching.
Only one, jfet in the shunt regulator J1 2SK170BL is reversed (wrong shape, but correct pins naming), but even if inserted according to the layout shape it will work as jfet source and drain are interchangeable.
Thank you for your info (IMO in the case of 2SK170, source and drain are not completely interchangeable when using it as an amplifier) so I will correct it 🙂
- Status
- Not open for further replies.