The Hybrid SIT Parafeed

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
REV.3

Ok guys. Here is the rev.3 schematic.

SIT-HYBRID-REV3-MG.jpeg :D
 
nice the parafeed output not to use a too big output cap..
do you prefer a bypassed source resistor than fixed bias with negative supply on the gate ?

I tested in previous version the fixed vs. auto bias and didnt sound
so much different. Maybe the autobias sounds a little more juicy and
fixed more dynamic but this are only minor differences.
I am very happy whit this version. The amp is very stable and very quiet.
Sounds lovely, not too tubey and not too solid. Clean but rich colourful sound.
 
Dear DIYspaceW,
Ok guys. Here is the rev.3 schematic.

earlier on I indicated interest in your schematics and am now glad to get this update! Therefore I would first like to:
1.) congratulate to your achievement
2.) thank you for sharing this.

By looking at the schematics I wonder how the different step-up/downs are orientated. I think to remember, that the first transformer was intended to be a step-down type (ratio?). The interstage seems to be a 4:1 step-down also. Would you mind to explain/confirm on this?
So, at the end what is the remaining overall gain of your amp?
Finally, may I ask for your opinion -as the UTC-LS-33 is not too easy to get hold of- to use one of the more regular 100V Transformers? I think they might be used successfully here. If confirmed, I would be keen to benefit from a local offer on tape-wound core 100V transformers. To my knowledge the smallest (40W) type could have a primary impedance of 250 Ohm via a 4H inductance requiring a larger inductance for the 2sk82.
Therefore, what do you think?
All the best,
Ulrich
 
Hello Ulrich!

The first transformer is an input 1:1 (18K reflected input impedance).
The following one is a step down 4:1 (9,6K:600R), so the gain in first stage
is aprox. 30-32 and the 2sk82 around 3,5.

Please post your schema, I dont know if there is any replacement for UTC LS-33.
Whit my operation points, you are looking for 2-3H primary inductance and 40-60 ohm nominal impedance. If you want a air gapped classical transformer, then 0.5H is sufficient I think....

This weekend I will make the measurements of the circuit.

Dear DIYspaceW,


earlier on I indicated interest in your schematics and am now glad to get this update! Therefore I would first like to:
1.) congratulate to your achievement
2.) thank you for sharing this.

By looking at the schematics I wonder how the different step-up/downs are orientated. I think to remember, that the first transformer was intended to be a step-down type (ratio?). The interstage seems to be a 4:1 step-down also. Would you mind to explain/confirm on this?
So, at the end what is the remaining overall gain of your amp?
Finally, may I ask for your opinion -as the UTC-LS-33 is not too easy to get hold of- to use one of the more regular 100V Transformers? I think they might be used successfully here. If confirmed, I would be keen to benefit from a local offer on tape-wound core 100V transformers. To my knowledge the smallest (40W) type could have a primary impedance of 250 Ohm via a 4H inductance requiring a larger inductance for the 2sk82.
Therefore, what do you think?
All the best,
Ulrich
 
Dear Marko,
thanks a lot for your explanations. Seems a good idea to spent some of the gain to enable a lower impedance drive of the 2sk!

My apologies for not replying sooner: I'm currently a slow-thinker, due to work overload.
The main (obvious) question for me regarding your amp is: what are my real needs and do I want?
Well, in general I would like to see my 2sk82's applied to a schematic that is "convincing" AND will not need too much investment into new parts.
In preparing myself for this adventure I try to understand and learn as much as possible.
But for now I don't have a schematic at hand.
I think I would like to see a balanced input and the following stages also in balanced layout. I could also imagine to use instead of the choke a CCS and a (necessarily) higher B+.
But the OPT question bothers me. This can easily get expansive. Therefore I'm still optimistic regarding the 100V transformers, as a potential alternative but I really do not know their inductance (my earlier reply on this was guessing I have to admit, and the shop has ridiculous opening hours, so just going there and measure these transformers is difficult). Only thing I know for sure is, their turns-ratio (60W 100V Transformer: 167/8 Ohms : U/U= 4,5:1).
The primary impedance might be a little high for 1 2sk82, but in balanced mode they might be looked at in series.
I'll need a little more time to consolidate my thinking.
Many thanks again,
Ulrich
 
Hi Ulrich,

after intensive listening sesions, I concluded that it sound
wonderful. But you need to keep in mind that its more delicate and
needs very good source and software. This amp plays wonderful in genres like
vocals, small bands, accustical material. Where power is needed like orchestral music,
it sound also good but its present the filling that it needs more power!!!

This weekend I took a look into circuit whit an oscilloscope and it looks very well. There are some minor phase shifts, propably due the transformers involved, nothing spectacular. As expected the secound harmonics are dominating.

I think you are on a good path. Test it first and see if it has potential for your needs.
 
Initial idea for discussion

Dear Marko,
after some rough time dictated by work issues, I had a moment to think about my approach towards using the 2SK82 transistor(s).
First I think your initial hint regarding the transconductance seems to "dictate" to prevent to use this transistor as a follower.
Second, I'm greatfull to find a 2SK82 spice model on the web. Not that I would be able to work with spice, but this model enabled me to check other operating points (OP).

Somewhere I read that at increased OP voltage, this transistors becomes more linear. Another spot reported reduced input capacitance at higher OP voltages. Seems to correspond. Thus, I simply used the 2SK82 model available from M. Rothachers homepage assuming that it is sufficiently accurate even at higher OP voltages (that I think it was not tuned for).

Some OP that might be ok could be at 140V/0.2A, enabled by some -29V on the gate, see graph (relying on M. Rothachers SIT model) and having included the (selected) 40W power limit for the SIT.
attachment.php


Third, there was/is this offer from the local shop regarding (hopefully acceptable) 100V transformers. Meanwhile I had the chance to measure the inductance of the transformer being most likely acceptably suited. It's a 100V 60W type with an impedance of 170 Ohm via an inductance of 1.05H and a secondary impedance of 4 Ohms.

This is sufficient apart from your UTC-LS-33, but at the other OP the transistor has an inner resistance of nearly 30 Ohms, if the spice model and my reading is correct. Assuming this being the case an OPT with 170 Ohms seems ok. It's lower (-3dB) frequence will be around 20 to 25 Hz. Which might seem ok, but I don't have experience in judging such data.

Fourth: I'm opening the discussion regarding my intended parafeed pull-push approach relying on a balanced tube input stage most likely driven by 6e6p's (or 6e5p's) with a 2:1 stepdown transformer, or initially -for cost reasons- C-coupled:
attachment.php


I'm expecting to get in the order of 10W out of this with an input voltage swing of +/- 16V assuming a 4-fold amplification via the 2SK82.
It might happen that I don't arrive at sufficient matched 2SK pairs to enable the use of a smaller value auto-biasing R. In that (likely) case I would need an autobias R for each of the 2SK82's.
To date I'm not sure how to connect the two secondaries of the OPT to best use my 8 Ohm speakers, so this is drawn as a dashed line.
So, anybody any idea regarding this? Understandable? Might it work? Something severe I overlooked? Any obvious -or not so obvious- improvement?
Please be patient, it's my first attempt to this & my background is school physics and some reading.
All the best,
Ulrich
 

Attachments

  • OP_2SK82.jpg
    OP_2SK82.jpg
    515.7 KB · Views: 295
  • Basic_Schematic.jpg
    Basic_Schematic.jpg
    581.1 KB · Views: 483
Dear Marko,
after some rough time dictated by work issues, I had a moment to think about my approach towards using the 2SK82 transistor(s).
First I think your initial hint regarding the transconductance seems to "dictate" to prevent to use this transistor as a follower.
Second, I'm greatfull to find a 2SK82 spice model on the web. Not that I would be able to work with spice, but this model enabled me to check other operating points (OP).

Somewhere I read that at increased OP voltage, this transistors becomes more linear. Another spot reported reduced input capacitance at higher OP voltages. Seems to correspond. Thus, I simply used the 2SK82 model available from M. Rothachers homepage assuming that it is sufficiently accurate even at higher OP voltages (that I think it was not tuned for).

Some OP that might be ok could be at 140V/0.2A, enabled by some -29V on the gate, see graph (relying on M. Rothachers SIT model) and having included the (selected) 40W power limit for the SIT.
attachment.php


Third, there was/is this offer from the local shop regarding (hopefully acceptable) 100V transformers. Meanwhile I had the chance to measure the inductance of the transformer being most likely acceptably suited. It's a 100V 60W type with an impedance of 170 Ohm via an inductance of 1.05H and a secondary impedance of 4 Ohms.

This is sufficient apart from your UTC-LS-33, but at the other OP the transistor has an inner resistance of nearly 30 Ohms, if the spice model and my reading is correct. Assuming this being the case an OPT with 170 Ohms seems ok. It's lower (-3dB) frequence will be around 20 to 25 Hz. Which might seem ok, but I don't have experience in judging such data.

Fourth: I'm opening the discussion regarding my intended parafeed pull-push approach relying on a balanced tube input stage most likely driven by 6e6p's (or 6e5p's) with a 2:1 stepdown transformer, or initially -for cost reasons- C-coupled:
attachment.php


I'm expecting to get in the order of 10W out of this with an input voltage swing of +/- 16V assuming a 4-fold amplification via the 2SK82.
It might happen that I don't arrive at sufficient matched 2SK pairs to enable the use of a smaller value auto-biasing R. In that (likely) case I would need an autobias R for each of the 2SK82's.
To date I'm not sure how to connect the two secondaries of the OPT to best use my 8 Ohm speakers, so this is drawn as a dashed line.
So, anybody any idea regarding this? Understandable? Might it work? Something severe I overlooked? Any obvious -or not so obvious- improvement?
Please be patient, it's my first attempt to this & my background is school physics and some reading.
All the best,
Ulrich

Hello Ulrich!
Very interesting idea!
First of all, If you want to do a push-pull you can also consider complimentary arangement whit P-SIT then you can eliminate the parafeed capacitor and use only transformer load.
The OP is an aprox. as such I dont think the lines are correct for high voltages, you should take a look here: diyaudio.com.hr • Pogledaj temu - SIT, u proslosti znan kao VFET

I think you should build it and test it, thats the only way!

I for myself have concluded that my new laudspeakers need more power, so I will retire my 2sk82 SIThybrid project, so if you need UTC LS-33 transformers let me know. They are also perfect for push pull, as they have two primaries. I am selling also some additional stuff left over from it...
 
Dear Marko,

DIYspaceW said:
Hello Ulrich!
Very interesting idea! ....
thanks a lot for your kind mentioning!

DIYspaceW said:
... The OP is an aprox. as such I dont think the lines are correct for high voltages, you should take a look here: diyaudio.com.hr • Pogledaj temu - SIT, u proslosti znan kao VFET

Oh, thanks for this hint! This web-page is completely new to me, (and not too good to read due to the automated translation) but in general it seems to confirm the outstanding linearity of this transistor, especially at higher voltages.

DIYspaceW said:
I think you should build it and test it, thats the only way!

Hm. Again, wright, although my available time for these things is ... the main factor hindering me from starting immediately.

DIYspaceW said:
I for myself have concluded that my new laudspeakers need more power, so I will retire my 2sk82 SIThybrid project, so if you need UTC LS-33 transformers let me know. They are also perfect for push pull, as they have two primaries. I am selling also some additional stuff left over from it...

Holla! This announcement makes me curious to see what's your next idea towards a solution for your situation! But as we're nearly alone here we might turn to PM to checks these things out?
All the best,
Ulrich
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.