As discussed with Rod Coleman,
Looking at post 14# of this thread..
Hi,
Well, to start with, select a PMOS FET that you can buy, and that will withstand the HT you have in mind. For small signals I use the ZVP0545, unless the current demand is too high. This part has a Vgs(th) or turn-ON voltage of about 3V. OK Now let's leave R4 at 1K is Wavebourns cct, since it makes the calculation easy in the head.
The trick is this: make the FET gate 5.5V lower than the supply. 3V is lost in the gate, and the other 2.5V appears across R4. 2.5V/1K = 2.5mA, the same current you have now. Drawback: the 3V varies with sample of FET, so some means to adjust is helpful.
But you just made a programmable CCS: use R2:R3 to split the voltage you are aiming for (350-95v)= 255V into 249.5V and 5.5V.
OK, that's the theoretical copy of your resistor circuit, now the improvements!
R5 is just a stopper. 5K will suffice unless it gets in the way. with 2.5mA, it drops 12.5V, so we must account for that. Next, the 6SN7 will work much better with higher voltage. So we keep the same current, and reduce that voltage across the gyrator.
To do this, start by aiming for 25v to 50v across the FET. We choose this because it reduces the FET capacitance. Let's say 40V. First check FET dissipation: 2.5mA x 50V = 125mW [max is 700mW]. Now add 2V for R4, 12,5V for R5. Now the total drop is about 65v - meaning your anode voltage is 285V - a huge improvement on 95V, and the 6SN7 will be more linear.
However - you may be using a following stage that needs 95V in, or some other figure - the procedure is simply to readjust R3 until the voltages work out, as above.
Of course, putting all this in LTSPICE speeds the calculation, but doing it in the head and verifying by building will give you a better grip on the circuit!
If it works out for you, then let's add this to the thread in case others are wondering. I sense that this excellent circuit is not widely understood.
Recalculate R2/3 for these voltages, keeping R2 near to the 330K - 470K needed to maintain the rollof point of ~3Hz with the C3.
-------------------------------------------------------------------------
I found this very useful. However R3 seems to be an issue for me with the large value required to have an anode voltage of 95V ..and the amount of voltage being dropped across R3.
Once again thank's for the help Rod..I know this will help others. I will continue looking at the R2/R3 balance..
Regards
M. Gregg
Looking at post 14# of this thread..
Hi,
Well, to start with, select a PMOS FET that you can buy, and that will withstand the HT you have in mind. For small signals I use the ZVP0545, unless the current demand is too high. This part has a Vgs(th) or turn-ON voltage of about 3V. OK Now let's leave R4 at 1K is Wavebourns cct, since it makes the calculation easy in the head.
The trick is this: make the FET gate 5.5V lower than the supply. 3V is lost in the gate, and the other 2.5V appears across R4. 2.5V/1K = 2.5mA, the same current you have now. Drawback: the 3V varies with sample of FET, so some means to adjust is helpful.
But you just made a programmable CCS: use R2:R3 to split the voltage you are aiming for (350-95v)= 255V into 249.5V and 5.5V.
OK, that's the theoretical copy of your resistor circuit, now the improvements!
R5 is just a stopper. 5K will suffice unless it gets in the way. with 2.5mA, it drops 12.5V, so we must account for that. Next, the 6SN7 will work much better with higher voltage. So we keep the same current, and reduce that voltage across the gyrator.
To do this, start by aiming for 25v to 50v across the FET. We choose this because it reduces the FET capacitance. Let's say 40V. First check FET dissipation: 2.5mA x 50V = 125mW [max is 700mW]. Now add 2V for R4, 12,5V for R5. Now the total drop is about 65v - meaning your anode voltage is 285V - a huge improvement on 95V, and the 6SN7 will be more linear.
However - you may be using a following stage that needs 95V in, or some other figure - the procedure is simply to readjust R3 until the voltages work out, as above.
Of course, putting all this in LTSPICE speeds the calculation, but doing it in the head and verifying by building will give you a better grip on the circuit!
If it works out for you, then let's add this to the thread in case others are wondering. I sense that this excellent circuit is not widely understood.
Recalculate R2/3 for these voltages, keeping R2 near to the 330K - 470K needed to maintain the rollof point of ~3Hz with the C3.
-------------------------------------------------------------------------
I found this very useful. However R3 seems to be an issue for me with the large value required to have an anode voltage of 95V ..and the amount of voltage being dropped across R3.
Once again thank's for the help Rod..I know this will help others. I will continue looking at the R2/R3 balance..
Regards
M. Gregg
Thanks Gregg and Rod, very useful.
Previously it was mentioned that the bipolar gyrator performed better than the MOSFET for currents below 10mA given MOSFET capacitances being higher at lower currents. My question is around sonic experience, has anyone compared how different all these gyrators sounds and found any noticeable difference or preference?
Cheers,
Ale
Previously it was mentioned that the bipolar gyrator performed better than the MOSFET for currents below 10mA given MOSFET capacitances being higher at lower currents. My question is around sonic experience, has anyone compared how different all these gyrators sounds and found any noticeable difference or preference?
Cheers,
Ale
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.