I would like to thank everyone for their comments.
To eLarson: Thank you for the point about the transconductance.
To stephanobilliani: My question was regarding the P1.7 as a transconductance amplifier. The Su-Sy issue was not in question. That has been discussed
elsewhere. My comments regarding the removal of parts were too brief. I apologize. I was stipping the P1.7 down to where it resemble a Zen Line Stage (or
low power SOZ) with CCS's. Imagine the P1.7 with only the input snubbing resistors and input impedance setting resistors, no parts on the gain Q drains
other than the CCS's, and no gain setiing resistance between the gain Q sources. Now change the transistors to 240's and 9240's, change the CCS currents
and that's roughly what I'm playing with. Su-Sy is not part of experiment - yet.
Graeme
To eLarson: Thank you for the point about the transconductance.
To stephanobilliani: My question was regarding the P1.7 as a transconductance amplifier. The Su-Sy issue was not in question. That has been discussed
elsewhere. My comments regarding the removal of parts were too brief. I apologize. I was stipping the P1.7 down to where it resemble a Zen Line Stage (or
low power SOZ) with CCS's. Imagine the P1.7 with only the input snubbing resistors and input impedance setting resistors, no parts on the gain Q drains
other than the CCS's, and no gain setiing resistance between the gain Q sources. Now change the transistors to 240's and 9240's, change the CCS currents
and that's roughly what I'm playing with. Su-Sy is not part of experiment - yet.
Graeme
Hi eLarson,
I am not enthusiastic about people who post schematics that aren't built and tested. But that's just my personal view. Here is the current schematic in .pdf form. It is a SOZ with CCS's top and bottom. The CCS design comes from "The Zen Variations - Part 2" fig. 1. The power supply is from "The Zen Variations - Part 5" fig. 8. I omitted the regulators here and substituted CL-60's for the inductors. The transformaer is a Signal 56-12. I started building this thing a few weeks ago because I became intrigued (a dangerous condition) and because I realized that I had most of the parts in my junk box.
The rail voltages are optomistic for the transformer even using the 108V primary tap. It won't certainly won't swing the +- 20V claimed by the F1. So I figure 5 - 8 watts instead of 10. That's enough to drive a single Fostex FE166E - which is the goal.
I used separate CCS's in order to allow adjusting the gain by placing resistance between the gain Q sources. It also spreads the heat out better over my heat sinks and allows room for further experimentation down the road. The upper CCS's are set up to provide aproximately 1.75/2 amps of bias. The F1 dissipates 200W. Subtract 10% for PS losses and you have 180 W or 90W a channel. Figure a 45V to 50V V+ for +-20V and you have about 2A of bias per channel which correlates with the 1.75A given in the spec sheet.
I have a hard time believing that an working transconductance amplifier is this simple, but I'm about to find out.
The amplifier is complete except for some small piece parts and the I/O wiring. I plan to fire it up before the end of the month.
My other reason for not posting a schematic by the way is a desire to avoid taking focus away from the subject of this thread, which is news about FirstWatt. If the moderators wish to move this discussion they have my blessing.
Graeme
I am not enthusiastic about people who post schematics that aren't built and tested. But that's just my personal view. Here is the current schematic in .pdf form. It is a SOZ with CCS's top and bottom. The CCS design comes from "The Zen Variations - Part 2" fig. 1. The power supply is from "The Zen Variations - Part 5" fig. 8. I omitted the regulators here and substituted CL-60's for the inductors. The transformaer is a Signal 56-12. I started building this thing a few weeks ago because I became intrigued (a dangerous condition) and because I realized that I had most of the parts in my junk box.
The rail voltages are optomistic for the transformer even using the 108V primary tap. It won't certainly won't swing the +- 20V claimed by the F1. So I figure 5 - 8 watts instead of 10. That's enough to drive a single Fostex FE166E - which is the goal.
I used separate CCS's in order to allow adjusting the gain by placing resistance between the gain Q sources. It also spreads the heat out better over my heat sinks and allows room for further experimentation down the road. The upper CCS's are set up to provide aproximately 1.75/2 amps of bias. The F1 dissipates 200W. Subtract 10% for PS losses and you have 180 W or 90W a channel. Figure a 45V to 50V V+ for +-20V and you have about 2A of bias per channel which correlates with the 1.75A given in the spec sheet.
I have a hard time believing that an working transconductance amplifier is this simple, but I'm about to find out.
The amplifier is complete except for some small piece parts and the I/O wiring. I plan to fire it up before the end of the month.
My other reason for not posting a schematic by the way is a desire to avoid taking focus away from the subject of this thread, which is news about FirstWatt. If the moderators wish to move this discussion they have my blessing.
Graeme
eLarson,
You stated:
Where did you get this number. Comparing it to the data sheet and the value given in the NP article on mosfets, it seems a bit low. The NP article seems to state that figuring out the transconductance for a given value of Id requires measuring an actual test circuit.
Graeme
You stated:
Where did you get this number. Comparing it to the data sheet and the value given in the NP article on mosfets, it seems a bit low. The NP article seems to state that figuring out the transconductance for a given value of Id requires measuring an actual test circuit.
Graeme
Hi eLarson,
NP states in the F1 manual that he uses no feedback for the purpose of reducing distortion but that he has has DC feedback to set the DC operating point. I think that this has more to do with using a single voltage PS. The SOZ which uses a dual supply and didn't require extra resistors to stabilize the DC operating point. I'll learn more once it's fired up.
You're absolutely right about diaodes and gate resistors. Thank you. The as-built amp has the 221 ohm resistors flying from the perf board to the transistors and the diodes mounted as you describe. I have corrected the schematic.
Graeme
NP states in the F1 manual that he uses no feedback for the purpose of reducing distortion but that he has has DC feedback to set the DC operating point. I think that this has more to do with using a single voltage PS. The SOZ which uses a dual supply and didn't require extra resistors to stabilize the DC operating point. I'll learn more once it's fired up.
You're absolutely right about diaodes and gate resistors. Thank you. The as-built amp has the 221 ohm resistors flying from the perf board to the transistors and the diodes mounted as you describe. I have corrected the schematic.
Graeme
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.