Thanks, John.
I'm well aware of both those buffer topologies. Which one of them is better, SQ wise?
Also, I'd love to hear your comments about the rest of the questions in my post from yesterday (cascoding, higher supply voltages and better smoothing of the supply voltages noise).
I'm well aware of both those buffer topologies. Which one of them is better, SQ wise?
Also, I'd love to hear your comments about the rest of the questions in my post from yesterday (cascoding, higher supply voltages and better smoothing of the supply voltages noise).
The complementary design is somewhat better. Lowering the noise with a cap multiplier is good. Cascoding? I'm not sure.
Hi John, I was told that a gain stage consisting of, say, a single JFET will have Miller capacitance. So if this was used at the input of a MM phono stage (with, say, a 60x gain on the first gain stage), the resulting Miller capacitance of, say, 320pF would give too high a capacitance loading on the MM cartridge.
(Miller capacitance is: gain x Crss + Ciss, so 60 x 5 + 20 = 320pF)
Whereas using 2 JFETs in a cascode takes away the Miller capacitance - is this true? (In which case, cascoding would seem to be a good thing! 🙂 )
Thanks,
Andy
Well Joshua, now that you have me thinking about this topic, I might add a few comments:
Both followers work well enough into relatively high impedances >5K, but the complementary design has 1/2 the drive impedance, and twice (or more) the peak current available to drive relatively large capacitances (from a somewhat longer cable, for instance.)
Of course, this extra drive current demands twice as much from the power supply. That is why a pair of cap followers/channel would be better than just Zeners driven by a resistor.
However, with the same sex follower, a 1W Zener should do the job, because the peak current is constrained.
You really should not go much farther than +/-15V with the complementary pair, because the P channel has only a 25V guaranteed breakdown, so even 15V is pushing it, worst case. However the single sex follower can be driven with +/- 18-20V if you really want to, because the N channel breakdown is 40V or more.
Setting the source resistors is relatively easy. Start with 10 ohms, as this will work OK with these hi Gm devices. For some other (low Gm) jfets, a larger resistor might be better.
Both followers work well enough into relatively high impedances >5K, but the complementary design has 1/2 the drive impedance, and twice (or more) the peak current available to drive relatively large capacitances (from a somewhat longer cable, for instance.)
Of course, this extra drive current demands twice as much from the power supply. That is why a pair of cap followers/channel would be better than just Zeners driven by a resistor.
However, with the same sex follower, a 1W Zener should do the job, because the peak current is constrained.
You really should not go much farther than +/-15V with the complementary pair, because the P channel has only a 25V guaranteed breakdown, so even 15V is pushing it, worst case. However the single sex follower can be driven with +/- 18-20V if you really want to, because the N channel breakdown is 40V or more.
Setting the source resistors is relatively easy. Start with 10 ohms, as this will work OK with these hi Gm devices. For some other (low Gm) jfets, a larger resistor might be better.
Attachments
Last edited:
- Status
- Not open for further replies.