Initially I received a schematic from one of our colleagues here, with a choke as load for the THF51 SIT transistor, capacitor coupled with the speaker. Since my trade is audio xfmrs I modified the schematic replacing the choke with an output transformer. I've added also an 1:3 input xfmr. Any comments are welcome! Thanks!
Last edited:
Nice design. It's good to see transformers and chokes. It should sound good.
I built a transformer coupled Tokin SIT amp: https://www.diyaudio.com/community/...ansformer-coupled-vfet-sit-amp-design.372261/. It's also a follower but with a transformer coupled SIT voltage amplifier as well. Output is about 18 watts into 8 ohms at 1% THD.
SITs do behave much like triodes and but unfortunately are becoming harder to find. Still, they are available and cheaper than high power tube triodes. Plus, they last much longer than tubes.
I built a transformer coupled Tokin SIT amp: https://www.diyaudio.com/community/...ansformer-coupled-vfet-sit-amp-design.372261/. It's also a follower but with a transformer coupled SIT voltage amplifier as well. Output is about 18 watts into 8 ohms at 1% THD.
SITs do behave much like triodes and but unfortunately are becoming harder to find. Still, they are available and cheaper than high power tube triodes. Plus, they last much longer than tubes.
@Sadface - I know the xfmrs can be made, I'm waiting for some volunteers to send me a couple of SIT transistors - I usually work with tubes, solid state was somehow out of my area of interest, but I realize there is a huge interest in these triode-like transistors. Having them buffered input-output with xfmrs seems to me a healthy solution 🙂
I don't see the point of this. A large, expensive, gapped for 3A transformer instead of a cap? Which apart from all other transfomer ills kills the damping factor?
The output impedance of a SIT follower is a bit marginal for normal speakers, i think the Df typically is about 15. The output transformer here cuts it at least by half, still good compared to a typical SET, but not great. The stepup input transformer is a similarly poor idea - SITs gates like being driven from a low impedance and here the driving impedance of the preamp is multiplied by 9 + dc resistance...
Not sure how much the triode characteristics benefit a follower anyway, perhaps not much.
It may be interesting to compare this to a common source topology where the transformer is in the drain and has a step down ratio. It may lose the novelty of being a spud though.
The output impedance of a SIT follower is a bit marginal for normal speakers, i think the Df typically is about 15. The output transformer here cuts it at least by half, still good compared to a typical SET, but not great. The stepup input transformer is a similarly poor idea - SITs gates like being driven from a low impedance and here the driving impedance of the preamp is multiplied by 9 + dc resistance...
Not sure how much the triode characteristics benefit a follower anyway, perhaps not much.
It may be interesting to compare this to a common source topology where the transformer is in the drain and has a step down ratio. It may lose the novelty of being a spud though.
@Dorin Bodea You'll want a buffer to drive the SIT gate properly at high frequency. Also, next time please quote the source of the curve data you showed in post #2.
Yes, I've noticed an attenuation above 12kHz, which dissapears if the input xfmr is 1:1 (it's all about playing with available impedances, where SIT has a lower input impedance than a native triode). The main reason I've inserted an input xfmr 1:3 there was for adapting the aprox. 2Vrms to what's needed to drive this device.
Also, I received the graph from one of our colleagues here, I don't know from where it is, but it is anyway more or less meaningless since the SIT acts as follower, as @analog_sa noticed 🙂
Romania, Europe 🙂
see some of my xfmr work here: https://www.facebook.com/profile.php?id=100057288107123
Last edited:
I knew what's on SS market maybe 30 years ago, but since then I've played with tubes exclusively. So feel free to suggest a list of actual SS devices in production today, which you consider less waste than a SIT... which was suggested to me on the initial schematic described at the beginning of this thread.
You could try some power JFETs. For example, the Lovoltech LU1014D and the UnitedSiC UJ3N065080K3S.
You don't need the "triode-like" curves for a source follower. Any power FET will work fine as a source follower. Use the SITs where they matter!
When the output is taken from the drain, so you actually make use of the triodey gain characteristics. Nelson Pass has a few such designs, and I am sure there are others.
Probably for a scheme like this the designer's choice to use the high voltage and low current VFET is the best choice (among other things the VFETs in these working conditions are more linear)
N. Pass is using SITs in his SIT-5 (common drain) amplifier. SITs have also been compared in common source and common drain configurations and have been preferred as followers by some people who prefer SITs over FETs.
This only to say some people with experience with them would not agree they are being wasted as followers.
Last edited: