Article - Sony VFETs part 1

[ianc13]

why not checking in TL431 datasheet , and having some faith in own conclusions .......

Hahaha, well yes that is exactly what I should do I suppose. Us greedy boyz can get just a bit lazy sometimes and ask questions here to shortcut having to do our own thinking 😉

I will pull up the datasheet and do the calculations. Likely it will take less than 5 minutes to do...

 

[Zen Mod]

you always have 2.5V between adj and lower terminal

resistor in between is setting current through "adj. rail"

so , you put upper resistor , current is set by lower one , then voltages across resistors is matter of proportion

sum voltage is ...... just sum

series resistor for entire combo is allowing/setting overal current for TL431 itself , resistor divider , and loading stage/consumer

simple , isn't it

edit : just do not overcook voltage A-K , not overcook max current , not overcook resistors , and you're good

 

[ianc13]

OK, so I did what I should in the first instance and did the calculations. I plugged it into Excel so that I could see what happens if I change R5 (3K3) and R7 (7K5) for different values of the potentiometer P1.

I can get a couple of extra volts bias and stay within the Imin for the TL431 if I drop the supply resistor (R5) and increase R7. 2K and 10K respectively look like it will work.

I will make the changes tonight and check the results.

Thanks Zen Mod!

 

[Nelson Pass]

I am hoping to bias it at ~0.8A, so pretty low. This amp will just feed a tweeter in a 3-way setup so it doesn't need much grunt and I'd rather run it at a lower bias for longevity.

What are the values of your supply rails?

😎

 

[ianc13]

What are the values of your supply rails?

😎

The main PSU is +/-24V or thereabouts (maybe +/-23.5V).

For the bias I am using 12V secondaries which give around 18V rectified DC so the TL431 needs to work with that to develop the bias.

Maybe I should just bias the amp a little higher, it is in a 4U 400mm deep chassis (from Modushop/HiFi-2000) so should easily cope with 1.2A or higher.

As said though, I don't need a lot of power from this amp at all. This will be feeding a 4 ohm planar tweeter crossed at 3100Hz (LR4 crossover). After a bit of reading a highpass driver over ~3000Hz only requires ~15% of total power for white noise - hence why I don't think I'll need a lot of grunt.

Are there other reasons to bias higher, nearer the specified 1.5A?

 

[plasnu]

I'm wondering if mismatched N&P pair would give single-ended-ish sound by carefully controlling the 2nd harmonics.
LTSpice emulation suggests it would be possible. Output capacitor would be needed, though.

 

[Nelson Pass]

I think the answer is yes, and DC offset issues do not necessarily call
for a capacitor.

 

[plasnu]

I appreciate the answer by the master himself. Good to know my current learning process of the Zen amp design is still on track.

 

[Nelson Pass]

I like the low hanging fruit. 😀

 

[plasnu]

🙂🙂

 

[dady]

Dear Nelson, how much offset is a serious compromise ?
Best Regards

 

[Zen Mod]

practically 500mV as transitional , and up to 100mV in temp. equilibrium

at least values acceptable for me

(remember that Papa is FAB (Numero Uno) , don't be surprised if he just erase m from my figures )

edit - for transitional value , it can go even higher , what's important is steepness of it - if DC arrives gentle , it's less critical

 

[dady]

Thanks dude!

 

[Zen Mod]

ya bet .........

it's stressful ...... every time I answer on something directed to Pa , I'm expecting a slap in next 3 hours (each time further diminishing my chances for adoption)

 

[Dennis Hui]

Hey ZM, it's been over 3 hours so you're probably safe from slapping...
this time. 🙂

 

[Nelson Pass]

Has not been slapped yet.

I greatly appreciate having ZM here to help shoulder the load.

🙂

 

[Holzarbeiter]

Adoption

ya bet .........

it's stressful ...... every time I answer on something directed to Pa , I'm expecting a slap in next 3 hours (each time further diminishing my chances for adoption)

ZM, Your chances for adoption were never in doubt. Now proof! H.

 

[Zen Mod]

naah ...... in last few years I started believing that Jack is responsible for eternal ongoing status of my adoption

his place simply can't be in any danger .... 🙂

 

[plasnu]

VFET Accordion Amp

I was reading TubeCad last night, and I thought Accordion Amp topology would be suitable for VFET single ended amp. LTSpice simulation suggests 2x Mu, 2X wattage than regular Sony 2SK82 SE amp, and typical SE THD figure.

I don't know if this amp works good in real world, so I ll appreciate it if someone would give me an advice. The Choke in the schematics is 0.15H 1 ohm, and R3 is for voltage compensation. Maybe a bad idea...

The Tube CAD Journal: The Accordion Amplifier: A new single-ended topology

 

[Zen Mod]

it's already made by Pa ( and numerous times before) , sans output xformer

Pa's is called F6