Veteran, a lot of people would look at your circuit and wonder why you have placed the JFET's on the output.
The input bias current on the LM4562 is about 10nA, which is very good, although still about an order of magnitude higher than GP JFET's (AT ROOM TEMPERATURE)
So, one option here is to place the JFET in front of the LM4562 (to get the 'JFET Sound' and the generally lower input bias current advantage) and then to buffer the LM4562 output with a class A stage - a simple emitter follower loaded with a suitable current source would do the trick. You could also bias the 4562 output into class A so the whole thing was class A. You could also then take feedback from the class A output stage, or for a bit more harmonic content, just around the 4562.
This would allow you to drive loads better, and easily use your pre-amp to drive headphones (with a suitable output stage load current of course).
Just my two pennies worth - good luck with your project.
The input bias current on the LM4562 is about 10nA, which is very good, although still about an order of magnitude higher than GP JFET's (AT ROOM TEMPERATURE)
So, one option here is to place the JFET in front of the LM4562 (to get the 'JFET Sound' and the generally lower input bias current advantage) and then to buffer the LM4562 output with a class A stage - a simple emitter follower loaded with a suitable current source would do the trick. You could also bias the 4562 output into class A so the whole thing was class A. You could also then take feedback from the class A output stage, or for a bit more harmonic content, just around the 4562.
This would allow you to drive loads better, and easily use your pre-amp to drive headphones (with a suitable output stage load current of course).
Just my two pennies worth - good luck with your project.
If I am interpreting what you have posted correctly then I disagree.
The Main Audio Ground (MAG) is where all the various circuits should take their reference voltage from.
That MAG must not be located on the PSU Zero Volts trace/wire/plate.
A single connection must be taken from the PSU Zero Volts to the MAG. This connection can be very short (the thickness of a PCB for example) or longer, much longer.
The Mag can be located on the speaker output terminal, or on the amp PCB, or on the PSU PCB, or floating in mid chassis (mid air) where all the routes to the user circuits are kept shortish and nearly equal (centroid of the user circuits).
So maybe I should have said "a very short conductor that protrudes from the PSU smoothing caps' ground".
It seems like the farther away the star ground is located, from the PSU caps' ground, the more the audio reference grounds' voltages would bounce.
Or maybe you meant that what you call the MAG only has the low-current audio ground returns attached to it? As long as any significant length (actually, any significant inductance or resistance, regardless of length) of the connection back to the PSU ground isn't shared with any large-amplitude or fast-changing currents, then it should be fine.
It seems like the farther away the star ground is located, from the PSU caps' ground, the more the audio reference grounds' voltages would bounce.
Or maybe you meant that what you call the MAG only has the low-current audio ground returns attached to it? As long as any significant length (actually, any significant inductance or resistance, regardless of length) of the connection back to the PSU ground isn't shared with any large-amplitude or fast-changing currents, then it should be fine.
Last edited:
After a few changes I measured the preamp again.
This is what I can squeeze out of my sound card:
Only the FET output stage with dc-servo circuit active:
Complete preamplifier with LM4562 on-board. THE+N is higher because of the gain set to x6:
This is what I can squeeze out of my sound card:
An externally hosted image should be here but it was not working when we last tested it.
Only the FET output stage with dc-servo circuit active:
An externally hosted image should be here but it was not working when we last tested it.
Complete preamplifier with LM4562 on-board. THE+N is higher because of the gain set to x6:
An externally hosted image should be here but it was not working when we last tested it.
I have a news for you
I finished the final version of this preamp. I also added a small PCB with low pass DAC filter and "soft start".PCBs are in production and should reach me in two weeks from now.
This this the final diagram:
http://www.unisonus.com/PDF/unisonus-buffer/FET_beffer_schematic.pdf
http://www.unisonus.com/PDF/unisonus-buffer/FET_beffer_PCB_dimensions.pdf
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
This is how the final design looks like:
An externally hosted image should be here but it was not working when we last tested it.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.