SuSy T. PuckFo

Member
Joined 2005
Paid Member
I really like the Fokin, LuFo, and SuSyLu designs. They are all nice for different reasons. I want an amp with a choke load and some Super Symmetry. I want to build an amp with my Tokin SIT pucks. The amp output power should be a bit higher than my good old F5 and Aleph-J. Oh, and I'd like to have some tube glow for voltage gain in front.

I already showed my simplified schematic idea here. Now that I decided to go ahead and make something of my good old SIT pucks, it was time to dream up a more complete circuit. Say hello to SuSy T. PuckFo!

Please take a look at my first draft. The front end is a long tailed pair of tubes for differential input with a voltage gain of approximately 24x. The output stage is a differential pair of THF51S source followers with a balanced choke load for SuSy cross coupling. Extrapolating from the SuSyLu, I'd expect close to 100 W(rms) output.

It will be a while before I'll start building the thing, so don't hold your breath to read about magic smoke or to see fugly photos.

Any thoughs or comments?
 

Attachments

  • SuSy_T_PuckFo.png
    SuSy_T_PuckFo.png
    49.3 KB · Views: 1,214
Interesting points!

(1) What do you mean by proper SuSy? What is not proper here?

(2) A cathode follower or source follower is typically used in tube amps to avoid "farting out" when the power tube starts drawing grid current; this is not an issue with the THF51 (I measured the gate current to be 100 nA or less). Also, a follower may be useful as a buffer if the gate exhibits a high capacitance to the front end. What is the capacitance into the gate of the THF51 in this configuration (source follower)?
Adding a source follower/buffer is an option, but would require a bit more circuitry. Since the follower will have some dropout voltage, it would require a high drain voltage to prevent limiting the AC voltage swing. As far as I can tell, this would require a dedicated power supply for the follower stage.
 
Official Court Jester
Joined 2003
Paid Member
whichever way, all of them are leading to Rome

:)

Tokins are exhibiting some gate current drift in time/temperature domain, ad that's largely neutralized if you have them DC connected to previous stage, with decent low Rout

also, whatever their capacitances are, I feel better when I know that I took care of them in same manner as in sentence before - dunno what is nature of them in voltage domain , but what we know is that these are lowish rail voltages for SIT ........ if anywhere is going to show nonlinearity of important capacitances, that's in lowish rail voltages domain
 
Tokins are exhibiting some gate current drift in time/temperature domain

Yup. I measured the gate current a while ago. It was between 10 nA and 100 nA, depending on temperature. That does not worry me much.

also, whatever their capacitances are, I feel better when I know that I took care of them in same manner as in sentence before - dunno what is nature of them in voltage domain , but what we know is that these are lowish rail voltages for SIT ........ if anywhere is going to show nonlinearity of important capacitances, that's in lowish rail voltages domain

I just looked at the input capacitance of a THF51 operated as source follower (see attached numbers). The values were about 0.8 nF to 1.3 nF, depending on the operating/bias point. I used a pretty simple set up, and I could not take the bias current as high as I intend to run the THF51 in the SuSy T. PuckFo, so these numbers may not tell the whole story... but: the tube LTP front end has an output impedance of about 6 kOhm. Running this into an 1 nF load, there would be a bit of roll off at 10 kHz and above. Not great. I guess a buffer is in order!

Proper susy is exchanged negative feedback from one side to the other...

Isn't that exactly what the transformer load does? I am not a lawyer, and it may not be SuSy sensu strictu as in the patent, but it is SuSy at heart. See this post (and the surrounding discussion) for what I mean by this.
 

Attachments

  • Screenshot from 2021-11-28 11-28-41.png
    Screenshot from 2021-11-28 11-28-41.png
    101.7 KB · Views: 582
Thinking about the buffer a bit, I re-read this thread. While that was aimed at driving a power tube, most of it is probably still meaningful for driving a SIT.

Some thoughts (and questions):
  • Which FET part would be a good choice? The SissySIT R.3 uses an IRF510, which has Crss of about 20 pF and lower for Vds > 10 V. The STF3LN80K5 recommended in the above thread will go down to 0.1 pF, and I already have a bunch of these somewhere. Thoughts?
  • I'd operate the follower buffer on a CCS load. How much current should it be? 1 mA? 10 mA? 100 mA?
  • What happens if the buffer FET fails, pulling the gate of the SIT very high of very low? Will that kill the SIT? Would it work to use a very-low power gate resistor on the SIT as a "fuse" (like a 50 mW part)?
 
Official Court Jester
Joined 2003
Paid Member
if choosing mosfet for "power drive" I believe George did cover it all in post #7 of that very thread

buffer Iq ...... I chose 10mA for buffer in my Singing Bush ...... simply by Thumb of Mighty ZM Rule

if it fails ....... SIT must have protecting G-S zener, foor excessive negative gate voltage

put hefty fuse in positive rail and you are good
 
Okay, maybe I need to explain why I did what I did. My ideas may be off, so please let me know if I am missing something (and what).

Zener connection at the gate:
I put the Zenter directly at the gate (not before the the gate resistors) because I was thinking to use the gate resistor as a "fuse". It the gate voltage goes too much negative, the Zener will start to conduct, and the current will flow through the Zener. If the voltage applied to the gate resistor becomes very negative (aproaching the -70 V), the current will increase and blow up the gate resistor, thereby disconnecting and protecting the gate.

Also, Nelson seems to connect the Zener directly to the gate, not before the gate resistor (example: https://www.firstwatt.com/pdf/art_diy_sony_vfet_pt2.pdf).

+/-70 V for the buffer:
As mentioned in the thread/post mentioned earlier, the buffer FET (STF3LN80K5) has very low input capacitance (1 pF or so) if it has 45 V or more across drain/source. To maintain the low input capacitance even under full AC swing (25 Vpk or so), we need 45 V + 25 V = 70 V.
 
Official Court Jester
Joined 2003
Paid Member
still at first coffee, skipped across fact that you need full blast at SIT gates (follower)

all good

zener - in practice it can be at any side of gate resistor; it just depends what you're thinking you're going to protect

I know for one case where Pa added protection zenner later, along with additional "gate resistor" so anode of zener is sitting in between two gate resistors , cathode to source

:devily: