Gyrator Question

[Wavebourn]

Originally posted by Hearinspace
Is that what Q3 is in Michael's circuit?

Yes, but they usually have 2 legs, look like diodes, and are called Stabistors. The common name is Zener, but actually Zeners occupy their place in the middle between FET stabistors (lower voltages) and Avalanche diodes (higher voltages). Stabistors usually have lower noises than Zeners do. Michael made one from a widely available MOSFET.


http://search.datasheetcatalog.net/key/STABISTOR

[Sheldon]

Originally posted by Michael Koster
OK, to clarify, the 210 and 1100 are stacked for direct coupling. The third supply (60V) is needed to drive the grid positive for class A2 and could be derived from any source e.g. screen grid supply if you had a pentode... BUT arranging it this way provides a direct return path to the cathode, therefore grid current does not have to be dealt with in the current loop of the driver anode or the power tube cathode. The 60V supply in my breadboard amp (common to both channels) consists of a toroid, a bridge rectifier, and a capacitor ;-) Also of note is the 210 and 60 supplies can be shared in a stereo amp because there are no common return current paths to induce crosstalk. Eventually the MOSFETs should probably be cascode for better PSRR but it's very good as it is.

Cheers,

Michael

PS if you trust the isolation of the MOSFET active loads on the outputs, then all the power supplies and filters can be shared in a multichannel amplifier

Yes, I hadn't really focused on the current loops, but having the current loops separated speaks to the thoroughness of the design.

I'm finding, personally, that even with normal resistive or inductive loads, that crosstalk is not a big issue. Yeah, it's there, but to my ears it doesn't degrade the presentation. Good records sound pretty good to me and you're lucky to get 30dB of separation (even then in probably only a narrow frequency range) with a good cartridge.


Sheldon

[ilimzn]

Originally posted by Wavebourn

An infrared LED can be probably used, but it wants about 1.5V, more than your cathode resistor provides.

A common CQY99 IR LED, for which there are probably dozens uppon dozens of equivalents, has a forward drop of 1.25V at 20mA, 1.15V at 10mA, so it should definitely be usable here...

[Hearinspace]

Michael, Sorry , I'm a little slow at this but I'd really like to understand what you're doing . . . . Looking at the Schematic (still) does the current from the 210V loop go through the 60V supply?

I'll try the MosFet in the cathode of the 417 as well as the IR LED (thanks ilimzn!) but I have a question about how generally to choose the part for this use . I don't see what on the data sheet for the 2N7000 indicates it will be good for 12mA @ 2.5VDC (I just grabbed this Grid Voltage from the 417 sheet, not sure if it's accurate in your circuit or not.) A friend told be it would be IDSS but looking at the sheet it doesn't seem to apply with VGD=0vdc. The only thing that looks like it could be it is the "Ohmic Region" chart at the top of page three of the data sheet . Can somebody tell me what to look for?

Wavebourn,

" There is one more option: to take DC feedback for gyrator from cathode of your output tube coupling tubes directly, but power supply will be more complex. "

If you have the time and inclination can I ask you to show me how you do this?

Thank you!!

[Wavebourn]

Originally posted by Hearinspace

Wavebourn,


If you have the time and inclination can I ask you to show me how you do this?

Thank you!!

A first, I would need some negative voltage supply with tiny current requirements, like 1 mA would be enough, so a parametric regulator with one 120V Zener will work fine. You may use some VR tube here, it looks nice indicating power-on as well.
Next, I would make a voltage divider from anode of the driver to that negative rail, so DC in the point between resistors would be zero, to drive the grid of the output stage. Then I would shunt the upper resistor of the divider by existing coupling cap. Now, I would disconnect that resistor in SVCS that goes to anode and senses anode voltage connecting it to the cathode of the output tube. However, it's value would be recalculated for a different voltage.

PS: thanks Ilimzn for the info!

[Hearinspace]

Hi Wavebourn, Is this what you mean?

[Wavebourn]

Originally posted by Hearinspace
Hi Wavebourn, Is this what you mean?

Exactly!

[Michael Koster]

Originally posted by Hearinspace

I'll try the MosFet in the cathode of the 417 as well as the IR LED (thanks ilimzn!) but I have a question about how generally to choose the part for this use . I don't see what on the data sheet for the 2N7000 indicates it will be good for 12mA @ 2.5VDC (I just grabbed this Grid Voltage from the 417 sheet, not sure if it's accurate in your circuit or not.) A friend told be it would be IDSS but looking at the sheet it doesn't seem to apply with VGD=0vdc. The only thing that looks like it could be it is the "Ohmic Region" chart at the top of page three of the data sheet . Can somebody tell me what to look for?

The "Transfer Characteristics" chart to the right of the Ohmic region chart is a good first approximation. Looing at the graph, one would expect temperature sensitivity and drift but I find the circuit is much more stable than I would have guessed. It is about 2.5 -2.6 volts at 10-12 mA. The dynamic resistance is about 10 ohms at this current, similar to an LED.

"does the current from the 210V loop go through the 60V supply?"

The current through the MOSFET has two branches, one through the 417A anode which returns to ground and the other through the 801A grid which returns to the 801A cathode. The standing anode current of the driver and the driver's share of grid current, which is Ig/(gm*R1), is supplied by both 60V and 210V supplies, and the remainder of the grid current is supplied by only the 60V supply.

Michael

[Hearinspace]

Originally posted by Wavebourn
Exactly!

But this is still cap coupled isn't it?
. . . . .Ahhh, what do I know. I'll still try it.
I'll play around with the present SVCS and the cathode loads for a week or so and then try the direct coupled version. Thanks Wavebourn !

By the way, can I just swap an N channel Mosfet for the PNP in the SVCS?

Originally posted by Michael Koster



The "Transfer Characteristics" chart to the right of the Ohmic region chart is a good first approximation. Looing at the graph, one would expect temperature sensitivity and drift but I find the circuit is much more stable than I would have guessed. It is about 2.5 -2.6 volts at 10-12 mA. The dynamic resistance is about 10 ohms at this current, similar to an LED.

"does the current from the 210V loop go through the 60V supply?"

The current through the MOSFET has two branches, one through the 417A anode which returns to ground and the other through the 801A grid which returns to the 801A cathode. The standing anode current of the driver and the driver's share of grid current, which is Ig/(gm*R1), is supplied by both 60V and 210V supplies, and the remainder of the grid current is supplied by only the 60V supply.

Michael

Thanks Michael, I wondered about the grid current . But as far as the power supplies go, the diagram makes it look like current from the 210V loop might also have to go through the 60V supply and if so I'm wondering how that happens. Am I wrong about that?

[Wavebourn]

Originally posted by Hearinspace

But this is still cap coupled isn't it?

It is DC coupled, but cap shunted.

By the way, can I just swap an N channel Mosfet for the PNP in the SVCS?

You can swap, but you should consider base currents, that means resistors should have smaller values, but the capacitor should have higher value.