Maybe we can brodean the usable SITs this way?
Please let me know why or why not this would work...I am a dumb builder..
The F in DEF stands for Follower, and your circuit has voltage gain.
Maybe nothing is free, what does it lose in return..... damping?
Of course I do not understand at all, why DEFiSIT is a follower and my circuit is not.......
Last edited:
go to FW site , under articles , look for pdf titled art_sony_vfet_pt1
edit: here it is : http://www.firstwatt.com/pdf/art_sony_vfet_pt1.pdf
read from start , but especially observe second circuit iteration in article
edit: here it is : http://www.firstwatt.com/pdf/art_sony_vfet_pt1.pdf
read from start , but especially observe second circuit iteration in article
Last edited:
Wow it's the same as CSX-2, replacing 2SJ28 with IRF9240. 
"They may look like Common Drain connections (followers) but they are not, since the drive voltages provided by the transformer secondaries are floating – they are still Common Source amplifiers."
"In particular, it doesn't have much of a damping factor, and so don't expect a lot of control out of it."
...So already one of the three merits of DEFiSIT is gone ..
higher damping (X)
single ended harmonics (?)
power noise reduction (?)
"They may look like Common Drain connections (followers) but they are not, since the drive voltages provided by the transformer secondaries are floating – they are still Common Source amplifiers."
"In particular, it doesn't have much of a damping factor, and so don't expect a lot of control out of it."
...So already one of the three merits of DEFiSIT is gone ..
higher damping (X)
single ended harmonics (?)
power noise reduction (?)
Your finding gave rise to an improvement idea. The first iteration is not so sensitive because the 2.5 V + 400 mV/A signal from Hall chip is attenuated to < 48 % by a resistor network. There is an option that increase effective signal to > 86% by DC shifting the opto using a voltage reference chip like the LM385-1.2 as attached. However, usability with the pucks still need further investigation.
Attachments
yeah .... gut feeling
per sims , comparison with (so called
) final Babelfish M25 sim , elevating optooupler for 1V25 , then increasing both Hall output resistors and gate-rail resistors ...... result is few times worst Iq stability ....... both in temp domain and rails change domain
so , gut ...... told me that Babelfish XA25 will stay with current mirrors in bias circuit
Hall chips - they simply demand additional gain stage , between them and opto(s)
in any case ..... idea was worth checking ; we are not in state Mu , in terms of scientific process (few steps back , few steps forth , one step back , few steps forth )
per sims , comparison with (so called
) final Babelfish M25 sim , elevating optooupler for 1V25 , then increasing both Hall output resistors and gate-rail resistors ...... result is few times worst Iq stability ....... both in temp domain and rails change domainso , gut ...... told me that Babelfish XA25 will stay with current mirrors in bias circuit
Hall chips - they simply demand additional gain stage , between them and opto(s)
in any case ..... idea was worth checking ; we are not in state Mu , in terms of scientific process (few steps back , few steps forth , one step back , few steps forth )
Even though we are no expert like Papa, we can always use the proven plug and pray method in LtSpice and avoid too many blown up parts.
speaking of ......... I really can't say where I induced more gray poof smoke - in sims or in real life
Attachments
Yes, wrong approach. AC component became prominent while stable Iq needs the DC average.
Or some kind of LPF to attenuate the AC component.
Attachments
Yes, and I realized that I am struggling against not-so-sensitive speakers for this fun game, hence using the ridiculous aluminum tower.
So an Altec speaker is being traded with a friend of mine and many SITs are available for play. 2SK60 2SK82 2SK180 THF-51S.........
Cheers