What To Do With Those 2SJ28's

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
If the extra voltage references don't bother you , then yes, or the V-FETRON with all the polarities flipped.

I've heard V-FETRON with 2sj28 many times and we all love it so much, that is one reason I fall in love with your schm(6) at first sight.
The other reasons, I dream one day I could understand and feel free to use auto bias methol for Vfet and I saw you used auto bias in schm(6). :)
 
Last edited:
Pass DIY Apprentice
Joined 2001
Paid Member
That diode will drop about a volt(ish) at 1.5-2A.

If you want to move forward on your own, here's what you could do: build the circuit with 22-25 volt supplies, your lowest Vgs pair, using 4-5 25W ohm rheostats and start by balancing the current loops at 1.5A and low DC offset. I wouldn't bring the diodes into play until you've tried that. If you find you can't get the current below 2A, then you can pursue the diodes.

Don't forget to have fun :)
 
That diode will drop about a volt(ish) at 1.5-2A.

If you want to move forward on your own, here's what you could do: build the circuit with 22-25 volt supplies, your lowest Vgs pair, using 4-5 25W ohm rheostats and start by balancing the current loops at 1.5A and low DC offset. I wouldn't bring the diodes into play until you've tried that. If you find you can't get the current below 2A, then you can pursue the diodes.

Don't forget to have fun :)

Thank you, Mr. MR! It is kind of you to answers all my silly questions :eek:

One more question:
Now I don't have low value wirewound potentionmeter Rheostat, but I find some 25ohm wirewound variable Rheostat which could be easy divided into two parallel 12.5ohm rheostat by cut resitance wire(about 50W). Could we use them to adjust bias and DC offset in schm(6)?
If not, I find 4-5 25W ohm rheostats.

Best Regards
Jamahamvui.
 

Attachments

  • IMG_0787.jpg
    IMG_0787.jpg
    471.5 KB · Views: 540
View attachment 613528

4 parts and a power supply constitutes a feed forward error correcting (SUSY-ish) power amp with very low levels of distortion. 2 X voltage gain.
It has a very good damping factor, due to the feed forward EC and the NFB.

If the 2SJ28 needs a gate resistor then the amp needs 5 components per channel plus a power supply.

L1 and L2 are two windings (bifilar) on an aircored choke. Since they are very close magnetically coupled together they will force the two outputs to closely follow each other in anti-phase (hence the SUSY-"ish" feed forward error correction).

Cheers
Johannes

Johannes amp are interesting too.
Hope he find one pair of 2SJ28 on the flea market for experimental builds, innovate creations.
 
View attachment 613946

View attachment 613947

A simple test with a battery biased SRPP dynamic current source and a 2SJ28 as a source-follower.

10+10 volt P-P input gives 7+7 volt P-P output. The 2SJ28 does not have a lot of transconductance. I am used to the IRFP7430 and IXTH140P05T so the 2SJ28 seems very weak and "soft".

Find more transconductance in Tokin 2SK180 if you like...well that other thread story :)
 
The one and only
Joined 2001
Paid Member
Something worth contemplating when you design with SITs:

The transconductance is a function of current and voltage, so that
it increases both with current Ids and voltage Vds. This is true
of all gain devices, but in SITs (and Triodes) the variation due to
voltage is much greater. We would say that they have a lower Drain
(or Plate) impedance.

In amplifying we see that the voltage Vds decreases as current Ids
increases and vice versa, so there is the opportunity for some
cancellation, giving a more constant transconductance figure, lowering
distortion.

At the "sweet spot" they cancel 2nd harmonic, giving the lowest
distortion numbers. Off the sweet spot, there is a 2nd harmonic,
variable in amplitude and phase.

There is a curve of such "spots" depending on the load, and what we
see is that the idle dissipation of an SIT in or near the sweet spot tends
to remain the same.

As a single-ended Class A example, the SIT-1 transistor does this
around 25 watts. Into an 8 ohm load minimal distortion occurs
around 15 volts and 1.7 amps. The output for this is around 25 watts
peak, and the dissipation is about 25 watts.

For a 16 ohm load, the voltage goes up to 21 V and the current down
to 1.2 A. Same dissipation and peak output.

A similar proportional relationship applies to push-pull operation, as well
as bias through mu followers. I leave you to figure the numbers out for
these cases (hint: they alter the apparent load).

Once you figure out what this number is, you can estimate the probable
ratio between supply voltage and bias current for a given SIT topology,
and if you work with these it may save you some time and effort.

:wiz:
 
Something worth contemplating when you design with SITs:

The transconductance is a function of current and voltage, so that
it increases both with current Ids and voltage Vds. This is true
of all gain devices, but in SITs (and Triodes) the variation due to
voltage is much greater. We would say that they have a lower Drain
(or Plate) impedance.

In amplifying we see that the voltage Vds decreases as current Ids
increases and vice versa, so there is the opportunity for some
cancellation, giving a more constant transconductance figure, lowering
distortion.

At the "sweet spot" they cancel 2nd harmonic, giving the lowest
distortion numbers. Off the sweet spot, there is a 2nd harmonic,
variable in amplitude and phase.

There is a curve of such "spots" depending on the load, and what we
see is that the idle dissipation of an SIT in or near the sweet spot tends
to remain the same.

As a single-ended Class A example, the SIT-1 transistor does this
around 25 watts. Into an 8 ohm load minimal distortion occurs
around 15 volts and 1.7 amps. The output for this is around 25 watts
peak, and the dissipation is about 25 watts.

For a 16 ohm load, the voltage goes up to 21 V and the current down
to 1.2 A. Same dissipation and peak output.

A similar proportional relationship applies to push-pull operation, as well
as bias through mu followers. I leave you to figure the numbers out for
these cases (hint: they alter the apparent load).

Once you figure out what this number is, you can estimate the probable
ratio between supply voltage and bias current for a given SIT topology,
and if you work with these it may save you some time and effort.

:wiz:

I make Mr. Pass post screen capture to add into personal collection of " The Very Best from Papa
s0820.gif
DiyAudio " :magnet::superman:
 
Pass DIY Apprentice
Joined 2001
Paid Member
KD-33, KE-33, and KF-33 are the same. They all end in "3" and share the same range of Vp.

In circuit #5, the complements should ideally be matched for for low Vgs values at 1.5-2A. This may be difficult to accomplish in actual practice given the scarcity of parts. The alternative is to implement fixed (adjustable) bias supplies. :)
 

Attachments

  • 2BC138D3-9B48-42A9-8245-246A58A83547.JPG
    2BC138D3-9B48-42A9-8245-246A58A83547.JPG
    71.4 KB · Views: 547
  • 69A37404-7C7B-42C7-9A9E-B566411D789A.JPG
    69A37404-7C7B-42C7-9A9E-B566411D789A.JPG
    294.7 KB · Views: 527
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.