F5 power amplifier

[tinitus]

Pre run with cheaper IR devices
As said, I was lucky to get hold of 8 pair of selected Renesas J162/k1058
I would really hate to see them burn
So, I consider to mount a single pair of cheaper IR devices on first powerup
Just with lowish bias
Just to confirm everything is working ok, before mounting the Renesas
It will be fairly easy as I will do it hardwired

Ehh, might be much easier to mount the Renesas without connecting them, and just drill the exstra holes fore the trial IR devices
Any problem with this, maybe in terms of statics ?

[gaborbela]

#1
lineup
diyAudio Member




Join Date: Dec 2005
Location: the north 10 Watt Single End, JFET input - Lateral N-MOSFET output

--------------------------------------------------------------------------------

hi

Daniel Petrov from Sofia part of Bulgaria
is know at this forum widowmaker - profile
for his Death of Gainclone - DoGC amplifier
DoGC uses a clever and unusual output idea.
See separate topic, with discussion to try figure out how that one works ...


Today, I found another of his creations.
In a Bulgarian forum Daniel started a topic 2007-03-27 ( 27th of March )
Link:
Еднотактен клас А с полеваци

It is a 10 Watt Single End, JFET input - Lateral N-MOSFET output amplifier.
As can be seen in my attached schematic
it features very good, but still not hard-to-find transistors:

- 2SK389BL .... hifi low noise jfet for input - paralleled for 12 mA in first stage

- TL431, BC550C, BC639 for good Output Class A bias control and PSRR (supply insensitivity)

- Two 2SK1529 Lateral MOSFET for Output Stage

- No global feedback from output to input stage

-------------------------------------------------------------------------------------

2SK1529 is a Toshiba Lateral MOS of same series as 2SK1058.
But with a bit higher power:
- 2SK1058, rated 07A, 100Watt
- 2SK1529, rated 10A, 120 Watt
See amplabs.com POWER MOSFET data: Audio Power Mosfets - Buy OnLine @ AmpsLab


Here below is is the attachment of this nice
- True Class A
- Single END
- ZERO global Feedback
- Toshiba Lateral MOSFET output
- 10 Watt
- Easy to build Power Amplifier
- Separated input-output power supply
- CLC filtered supply, a la Nelson Pass


Regards
lineup


------------------------------------------------------------------------

Hello
Thank you , that punch under the belt a bit .
You guys write always to listen to a expert guys , take advise from them. So I did it .It was Lineup who wrote those Toshiba mosfet(s) are laterals .
Even today He say that .
Also I wrote I'm not a engineer or technician , I learn all DIY "hands on", I read the forum or I ask for advise from guys like you !
From the data not always easy (to me) to know if they are laterals or vertical .
It was a long topic in another tread exactly about these Toshiba mosfet(s), ended up with a disagreement . If is so easy to know from the data those are lateral or vertical I think the expert guys would agree on the fact easily .
The reason when I post a mistake I correct it and I'm never ashamed to write sorry about .
Sorry but I have to write there are hundreds of DIY people we depend guys like you JUMA .
We also like to build our amplifier .You know that well .
Any way you help me with the cascode and I greatly appreseate it .
A very big thank you to you for that !!!

Greets

[tinitus]

Ehh, who says the 2SK1529/30 isnt lateral mosfets

[gaborbela]

Is NOT!
These confusion around these mosfet lead me to spent over $600 bucks in vain .
At least in F5 I can use some of them.
Greets

[tinitus]

Originally Posted by gaborbela

Any way you help me with the cascode and I greatly appreseate it .

Start reading
A major part of last posts here have been about cascoding F5
All you need to know is here, not very long back
read first, ask afterwards

And please, try not to pollute the thread with long non relevant quotes like the one you did on lineup, please
Or at least edit out the whats not important
Its difficult enough to keep track of things as is

[tinitus]

Issue with thermistors
Or should I better say, with no thermistors

Say, if someone did a 40watt, with classA bias up to 20watt
Multiple outputs and cascode input
We have heard more than once that above a certain biasing some non IR outputs may get a bit nervous, and run hot, without the thermistors
Or maybe better said, bias gets unstable,as I understand it

Above described 40watt, with lowish bias should not have such problems
Or maybe just until the output exceeds the 20watt classA
What really happens when it goes into classAB above its 20watt classA biasing

Will this amp show problems with bias stability, when driven hard and goes into classAB
This could make me think that it might best to keep the thermistors with such design
I mean, it is a very simple curcuit, and all we have to stabilise bias is the thermistors
Or the neat current limiter
Maybe its time to reconsider it, with proper adjustments
And maybe NOT use higher voltage than makes a full classA drive possible
Yeah, we may have approached this several times before
Just thinking loud to clear my head
But any response would be nice

F5 really seems to be a very finely calculated balance act, and really isnt designed to do what we may try to push it into
Maybe we should be more careful in what we do
It started out with trying higher bias to improve on low impedance drive
Multiple outputs etc
Now its raised supply rails
Yeah, we are "the greedy boyz"
Maybe we will learn it the hard way
With smoke

[bobodioulasso]

You should not brake anything if you first power on either using a variac or a simple lightbulb.
Aren't thermistors useless with laterals?

[tinitus]

Yeah, had a strange night
I think I was on track with the milled and smoothened heatsink and keratherm isolators
Sometimes I forget myself
Appreciate your encouragement

[smyslow]

Hi,

after listening for around 4 month or so to this wunderful amp I think it's time to post some technical details and of course some pictures.

My actual F5 is based on EUVL's balanced F5 posted here:
http://www.diyaudio.com/forums/1542055-post1216.html

But I didn't implement the output protection circuits (Q4, Q5 in the original schematic) and used identical values for the source resistors (R1,2 =10Ohm, R11,12 = 0.235Ohm) and the feedback resistors (R5,6,7,8 = 100Ohm). The output devices are 2SK1530/2SJ201 from Toshiba.

The bias current is set to 1.6A. Unfortunately, the rail voltages differs between the right and left channel (left: 2x12Vac -> +/-14Vdc, right: 2x15Vac -> +/-18Vdc) and the toroids are a bit small for such an class A amp, they are rated only at 250VA. I've ordered new ones rated at 400VA.

It's built up with cviller's boards for the amp channels and power supplies, and boards from Peter Daniel for the rectifier bridges. The enclousure is from Hifi2000 (Pesante Dissipante 05/400).

Here are some pictures:

[smyslow]

Before building my actual balanced F5 I set up an original one, it was something like an "engineering sample" and useful to play around with bias current, supply voltages, etc.

For this version I used the boards from Peter Daniel because they fitted better to the heatsinks I had.

Again some some pictures: