Nice curves!
I believe these Tokin SITs are still undervalued even as their prices have increased lately. I used to build single ended directly heated triode amps, and soundwise, these SITs are comparable. However, an equivalent high power tube triode would be several times more expensive and then a very expensive output transformer is required. Instead of $100 for a SIT, you would need to spend at least $1000 per channel (tube and output transformer) for the tube amp.
I believe these Tokin SITs are still undervalued even as their prices have increased lately. I used to build single ended directly heated triode amps, and soundwise, these SITs are comparable. However, an equivalent high power tube triode would be several times more expensive and then a very expensive output transformer is required. Instead of $100 for a SIT, you would need to spend at least $1000 per channel (tube and output transformer) for the tube amp.
Speaking of tube amps... I have learned a thing or two about designing tube amps. Now I wonder how this can be transferred to SITs (or MOSFETs in general). It's not hard to see that there is no need for output transformers, dangerous high voltages, and filament supplies. All this is very attractive!
However, I wonder what it takes to properly drive the gate of a SIT. With tubes, it's become a thing to drive the grid through a FET in source follower configuration to avoid "farting out" when the tube grid starts drawing current, but the grid current would be blocked by the coupling capacitor if there was no FET buffer. What about SITs? How much gate current is there, and should I care? Is the gate current constant, or does it change as a function of the gate voltage (or drain-source voltage), as it is the case with the the grid current of a tube? Any pointers to learn about these things would be great!
However, I wonder what it takes to properly drive the gate of a SIT. With tubes, it's become a thing to drive the grid through a FET in source follower configuration to avoid "farting out" when the tube grid starts drawing current, but the grid current would be blocked by the coupling capacitor if there was no FET buffer. What about SITs? How much gate current is there, and should I care? Is the gate current constant, or does it change as a function of the gate voltage (or drain-source voltage), as it is the case with the the grid current of a tube? Any pointers to learn about these things would be great!
Check this site out for some info on SITs:
AMPLIMOS V-Fet & SIT Amplifiers, One Stage Amplifiers
This is from the site - "[FONT=Times New Roman, Times]As said the SIT has been originally conceived as a solid-state analog of the vacuum-tube triode. The device is normally on, and a reverse bias applied to the gate is used to modulate the drain-source current. In this mode of operation the steadystate current drawn from the gate is negligible, and the SIT can be considered as a voltage-controlled device, like the JFET."[/FONT]
AMPLIMOS V-Fet & SIT Amplifiers, One Stage Amplifiers
This is from the site - "[FONT=Times New Roman, Times]As said the SIT has been originally conceived as a solid-state analog of the vacuum-tube triode. The device is normally on, and a reverse bias applied to the gate is used to modulate the drain-source current. In this mode of operation the steadystate current drawn from the gate is negligible, and the SIT can be considered as a voltage-controlled device, like the JFET."[/FONT]
SITs are robust, but I don't think I would want to keep using it @200W/90°C. My goal is 180W/50°C, and I'm a still a little paranoid about thermal resistance.
All SIT 50W+ Single Ended Amp (beta)
Just for the record, I made a few measurements with my THF51S at different DC bias points ranging between 10...40 V and 0.5...4 A. I used a 1 MΩ gate resistor to read the gate current from the voltage drop across the resistor. The gate current was always between 0.01 and 0.1 μA. The variation was mostly due to temperature changes of the SIT, whereby the current was higher when the SIT was running hot. With a normal size gate resistor I wouldn't expect to see any substantial voltage drop. I also don't expect any dynamic changes of the gate current (or gate resistor voltage drop) due to AC signal variations, as the temperature of the SIT will not fluctuate with the AC signal.
Two years later, how has everyone progressed with operating SITs in high voltage, low current modes?
If i'm reading these curves right (and that's a big if) it appears that these SITs don't need a lot of swing on the gate (i.e they are high mu), but they do need some current to overcome the gate capacitance to preserve HF response. So in my particular case, with a hot source (4.5Vrms DAC that i'm not likely to change soon) i don't think i need a lot of voltage gain in the preceeding stages. Let's say a good operating point would be 300V, 250mA quiescent current through the SIT, single ended, which should be close to -13V bias voltage. I should be able to roughly calculate mu and gm from the curves for this operating point. For example like this. Rp = μ / gm, for a SIT, same as for a triode, right? Then choose a primary impedance of the output transformer that is some value higher than this Rp. Suggestions welcome here...
As for driving the SIT gate however, like with a straight buffer isn't my source still too hot? 4.5Vrms ≈ 12.7Vpp which would bring the gate to around -6.65V on the largest positive peaks... Intuitively it feels like too much swing to me, would probably need a step-down, right?
If i'm reading these curves right (and that's a big if) it appears that these SITs don't need a lot of swing on the gate (i.e they are high mu), but they do need some current to overcome the gate capacitance to preserve HF response. So in my particular case, with a hot source (4.5Vrms DAC that i'm not likely to change soon) i don't think i need a lot of voltage gain in the preceeding stages. Let's say a good operating point would be 300V, 250mA quiescent current through the SIT, single ended, which should be close to -13V bias voltage. I should be able to roughly calculate mu and gm from the curves for this operating point. For example like this. Rp = μ / gm, for a SIT, same as for a triode, right? Then choose a primary impedance of the output transformer that is some value higher than this Rp. Suggestions welcome here...
As for driving the SIT gate however, like with a straight buffer isn't my source still too hot? 4.5Vrms ≈ 12.7Vpp which would bring the gate to around -6.65V on the largest positive peaks... Intuitively it feels like too much swing to me, would probably need a step-down, right?
Yes, probably smart to start smaller. I have some unused Lundahls just sitting, 1.25k primary, gapped for 125mA. I will finish my SissySIT first, i’ll have something to compare with.
Any idea what kind of swing on the gate i should aim for, ballpark figure? Just wire something up and try i guess.
Any idea what kind of swing on the gate i should aim for, ballpark figure? Just wire something up and try i guess.