F5 with 2SK2013/2SJ313

[crt]

Originally Posted by Nelson Pass

I think you want some capacitance across the 100 ohm feedback resistor.
Start with 1 nF.

I'll try it tomorrow
Thank's

Originally Posted by juma

Input looks fairly OK if the picture is taken under same conditions - func. gen. loaded by amp's input.

Ups, sorry.
That signal generator to scope direct without loaded by amp input

this one loaded with amp input

Originally Posted by juma

Try first the nr. 3 (add 330R in series with 240R at the gates of MOSFETs) - just to eliminate the last thing first...

Well this one is "heavy" task, because i had to pull the amp pcb off from heatsink
i'll try it after add some capacitance at feedback first

[juma]

Tempest in a teapot
Your amp did just what it's supposed to do - overshot in, overshot out...

[crt]

Originally Posted by juma

Tempest in a teapot
Your amp did just what it's supposed to do - overshot in, overshot out...

Yes, that slight overshoot but not much compare than output

Here's the report :
1. Remove one of 150R (75R sum) at feedback ==> no luck here
2. Change 243R gate stopper to 681R ==> same, no luck
3. Put capacitance as Nelson Pass says at 150R (50R sum) ==> compress the overshoot a bit (try to lower it to 200pF and then 100pF, nothing change)
4. Add a bypass caps at PCB Amp (10uF with 100nF) ==> compress overshoot a bit

Here the result (combination no 3 & 4)

Thank's (Juma and Nelson Pass) for all support

[The Space Egg Corp]

Hi Juma

I like your amp design !

Just got to the end of the thread...
...with the world's s l o w e s t internet...
...Phew !

1 question...
...The thermistor.
I see where it is in the schematic...
...I see people mount it 'hovering near' an OP device ( not a metal-tab one, on sink ).

I dunno if I'm being thick here ( probably )...
...but, What does it do ?

You obviously wouldn't have it in your schematic if you thought it wasn't needed...
...one builder didn't fit it...
...I guess they just fitted a 'suitable' resistor, or re-calced the parallel 2.2k etc.

Any explanation of this function would be most helpfull.

People seem to have enjoyed this amp very much...
...great work !

Cheers.

Si.

[juma]

Originally Posted by The Space Egg Corp

...
I dunno if I'm being thick here ( probably )...
...but, What does it do ?...

Not thick - just plain lazy...
Read the F5 manual/article at the firstwatt.com

NTCs are there to thermally compensate for the vertical MOSFETs' tendency to increase channel current with temperature.
With biiiiig heatsinks and enough time at hand to wait for thermal equilibrium to be established (an hour or so), you might decide that you can live without NTCs in this amp...

[The Space Egg Corp]

Hi Juma

Not at all...
...cheeky !
( lazy, I mean...def. thick ! )

I tried downloading the F5 manual THREE times at 1stwhat?...
...it ain't happening on this computer, is all I can say.

I see...
...thanks...
...short & sweet ( just like the amp, by the sound of things ! )
I figured it was for something like this of course.
It just seemed to me, with 'loose coupling' ie. no metal-tab...
...to be more of a 'randomizer' than a 'stabilizer' ! ( probably just me though ).

I see the F5 similarity.
But don't really see this as anything like an F5.
Loads of bits taken out...
...& regulated supply put in.

Reminds me more of some kind of Hitachi or Motorola application notes, I vaugely remember from the 70s' or 80s'.

biiiiig is better !

Cheers.

Si.

[The Space Egg Corp]

Hi Juma

I'm still thinking your amp looks pretty groovy.

So not wishing to appear TOO lazy...
...made a start today...by erasing all the text from the schematic.

( no need to be alarmed...we do that on all our new projects here )

WOW ! it's just so mind-blowingly MINIMAL !

C O O L !

Cheers.


Si.

Here it is after the erasing work.

[juma]

Actually, we can shave this version of F5 a bit more and still make it work nicely.
If we accept lower power (10W RMS @ 8 Ohms) and use the lower voltage in power supply (+/-15V) we can bias the MOSFETs near the zero tempco point (about 0.6A) which will get us rid of NTCs and with carefully chosen JFETs' drain resistors we can avoid trimming pots.
For lower power 2 pairs of output MOSFETs will be OK.

[The Space Egg Corp]

" Jumpin' junction count Comissioner Gordon !...
...That joker Juma will have the Gotham City semi-conductor industry on it's knees by January ! "

" Exactly Bat-Ears !...
...You and the Boy-Wonder are the Gotham high-end audio industries last hope...
...we can't allow this kind of musical minimalism on the streets ! "

" Robin !...To the Toshiba-mobile ! "

" Yes Bat-Ears ! "

[The Space Egg Corp]

Hi Juma

I like your schematic by the way !

If I had 22-bit digits ( instead of 20 ), I could count ALL the parts in 1 Byte ( including the extra thumbs ! ), and NOT loose count...
...N I C E !

Could be possible to loose 1 of those feedback resistors, I guess; and replace with 1 BIGGIE !
I know people get all 'steamed up' about 'mythical inductance' here...
...any objections to 1 B I G chunky 25R wire-wound ?
Mmmm...

I like the parts minimalism Juma...
...but I rec. you 'had it nailed' the 1st time, with the 6 little Toshey's you used.
( I wanted 8, when I first looked; but hey, dunno why really )

Your original schematic looks 'good to go' to me...
...especially since The Space Egg Corp. parts-stash, already has most of the bits & bobs for the amp !!!
The transformer we have BTW is a 225 Watt 18-0-18 donut...
...could get a second one of these as well; I know they like 'em BIG round these parts.
( seems like enough Watt-not's to me though; bout 100 needed )

Loads of other parts here as well.
Main missing items being the IRF's & a couple of 2SJ74's ( got some 2SK170's ).

As for the whole NTC thermistor buisiness...
...We have some quite unusual heatsinks here...
...been in 'cold storage' for a few years; waiting for 'the perfect amp' !
4 of 'em; strange-shape, actualy intended for forced-air cooling...
...although they might do the job with just convection ( prefered ).
( we have an in-audible Akasa 120mm ball-bearing 'balanced' PC fan here, as well, N I C E ! )

The data I have for the heatsinks, is all for forced-air apps...
...so unsure of the convection specs.
They do look a 'bit' small for you boys round here...
...but I rec. I might JUST get away with it; if the case is designed around them, in a good way.

The other interesting part I have, is a whole bunch of 220R wire-wound 70s' instrumentation pots ( N I C E ! they don't make 'em like THAT any more ! ).
I know pots send shivers down the spines of yer average 'audiophool'...
...but hey, these dudes is q u a l i t y.
Perhaps a few K's across 'em for a bit of 'ultimate stability'.

Anyhow, all good.
Can't really live with 10 Watts...
...I dunno though...
...Could we ?...
...Mmmm...?

That 'IDEAL' operating point you mentioned, does sound interesting though...
...we aint got a 15-0-15 though.

I guess you originaly went for what you did; cos of your 'lowish' Ohm speakers.

Cheers.

Si.