I know from lightly filtered 431 Vref in regulators especially when biased strongly, how you know?
Salas, I built a shunt regulator using a buffered TL431 - it's shown at the beginning of my "All-In-One" preamp thread.
http://www.diyaudio.com/forums/analog-line-level/234596-all-one-preamp-circuit.html
The regulator had reasonably low parts count, but unacceptable hiss when used in conjuction with a cascoded common-source RIAA first stage (no PSRR to speak of). I bodged one of my LED-referenced shunt regulators on to the board in place of the 431 circuit, and the hiss went away.
JG, it's true you can pad down a TL431 with a cap if you're using it as a stand-alone reference, but it'll take a big, big cap, as they have an incremental impedance of about an ohm. It would be interesting to see how quiet you could make an amplified 431 (a 431 with resistive divider slung between cathode, reference, and anode to boost the voltage at the cathode) using an elcap. I haven't tried it yet, as GaAsP LEDs are prettier, they work, and I have tons (and not tons of time to try everything).
http://www.diyaudio.com/forums/analog-line-level/234596-all-one-preamp-circuit.html
The regulator had reasonably low parts count, but unacceptable hiss when used in conjuction with a cascoded common-source RIAA first stage (no PSRR to speak of). I bodged one of my LED-referenced shunt regulators on to the board in place of the 431 circuit, and the hiss went away.
JG, it's true you can pad down a TL431 with a cap if you're using it as a stand-alone reference, but it'll take a big, big cap, as they have an incremental impedance of about an ohm. It would be interesting to see how quiet you could make an amplified 431 (a 431 with resistive divider slung between cathode, reference, and anode to boost the voltage at the cathode) using an elcap. I haven't tried it yet, as GaAsP LEDs are prettier, they work, and I have tons (and not tons of time to try everything).
Last edited:
I would only do the capacitor padding if i was using the 431 just as a reference, not as an active regulation element/error amplifier inside a feedback loop. Driven with a current source, and used just as a voltage reference, it might actually behave with a big capacitor across it. Having said that, I don't feel especially motivated to throw any time at the concept, as I'm getting good results with LEDs. There's a little bit of "once burned, twice shy" as well.
Edit - the attached picture shows what I'm talking about, for the sake of clarity - and that's all the effort I want to spend on this concept for now...
Edit - the attached picture shows what I'm talking about, for the sake of clarity - and that's all the effort I want to spend on this concept for now...
Attachments
Last edited:
Quite true, though what might appear initially as a deficit is useful as a soft start/thump eliminator, esp. for single-ended circuits, just as long as nothing weird happens happens during that long, long trudge up the slope to full Vcc... After all, we wait quite patiently for the filaments on our precious vacuum tubes to warm up - at least, I do.
Edit - the capacitor might not be so huge if you set the -3dB point at 50-100 Hz- but then again, if someone is really hot to trot to try such an experiment, they can do so and report results back here or in a separate thread...
Edit - the capacitor might not be so huge if you set the -3dB point at 50-100 Hz- but then again, if someone is really hot to trot to try such an experiment, they can do so and report results back here or in a separate thread...
Last edited:
Back to the task at hand
Apologies if I've dragged your thread OT Wrenchone. Although I for one have learnt more about CCS's in the last two days than I expected
Its a long weekend here so I might knock up a pcb for the line stage and/or do a bit of matching of parts for a build. Current schematic as to where I'm at is below. The discussion around current sources has now gone somewhat over my head, so if there's an obvious improvement over the attached please do say so. That said depletion fets are out as I can't get them here in oz and I'd like to get on with it
Apologies if I've dragged your thread OT Wrenchone. Although I for one have learnt more about CCS's in the last two days than I expected
Its a long weekend here so I might knock up a pcb for the line stage and/or do a bit of matching of parts for a build. Current schematic as to where I'm at is below. The discussion around current sources has now gone somewhat over my head, so if there's an obvious improvement over the attached please do say so. That said depletion fets are out as I can't get them here in oz and I'd like to get on with it
Attachments
ZM/Choky -
I checked out the NT phono preamp you cited a few posts ago to refresh my memory. I'm using a superficially similar design (in concept at least, 2SK170 single sex folded cascode instead of full complementary) as master studio phono preamps for the radio station where I DJ as a hobby. I used passive instead of active EQ, and included an output follower for each stage. The preamps are still going strong after 10 years, though I'm going to replace them soon with a simpler design.
The NT design would need to be updated with matched pairs of K170/J74 jfets, as the matched monolithic devices used in the original design are well-nigh unobtainium.
I checked out the NT phono preamp you cited a few posts ago to refresh my memory. I'm using a superficially similar design (in concept at least, 2SK170 single sex folded cascode instead of full complementary) as master studio phono preamps for the radio station where I DJ as a hobby. I used passive instead of active EQ, and included an output follower for each stage. The preamps are still going strong after 10 years, though I'm going to replace them soon with a simpler design.
The NT design would need to be updated with matched pairs of K170/J74 jfets, as the matched monolithic devices used in the original design are well-nigh unobtainium.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.