F5 power amplifier

Hey westend,
Thanks for the reply. The palm sander must have been dirty work. I hand wet sanded which was dust free. But like I said, it was along process. I was very pleased with the results though. I mayr try your technique this time. What's the highest grit you used on the palm sander?

Steve
Hey Steve,
The polishing process is as follows: Wet sand with 1/4 sheet palm sander-220,320,400,600,800,1000,1500,2000. Buff with rotary 6" car buffer (low speed) No. 7 rubbing compound, Turtle Wax`Polishing Compound for final buff.
Even with an aluminum panel with deep mill marks this process is not that daunting. I only sanded with each grit for a couple of minutes, using different directions. There are better metal polishing compounds but I used what I had on hand.
 

Attachments

  • IMG_0006.JPG
    IMG_0006.JPG
    79 KB · Views: 730
  • IMG_0001 (4).JPG
    IMG_0001 (4).JPG
    78.3 KB · Views: 709
  • IMG_0008.JPG
    IMG_0008.JPG
    76.3 KB · Views: 685
I am using the same voltage but have increased the bias current for higher class A output into 4 Ohms.
I did test it though up to +/- 30 Volt rails with no issues what so ever. In fact it seemed to perform slightly better (not significantly) at higher voltage.

Hello Thanh
I find your amplifier , if I see it well U use 3pair IR power mosfet not Renesans or Toshiba devices .
If you can drive the 3pair IR devices so can I my Toshiba's .
Also I took your advise and made another layout(before I found your amp) I was a bit worried to drive 3pair mosfet with those litle JETs .
Now I have a question if I use cascode (only I ask these because your sim. tests much better if cascoded ) what would be the resistor value at 24V power supp.
Juma made the drawing in cascode mode at 40V rail and his resistor advise was 4.7k . I think 4.7k to high for 24V rail .
I wan to order the parts all of them the same time from Parts Connexion with one shipping cost .
Do you have any idea ?
Please let me know .
Thank you very much
Greets
 

Attachments

  • IMG_6516_edited-1.jpg
    IMG_6516_edited-1.jpg
    171.3 KB · Views: 618
  • F5-casc.gif
    F5-casc.gif
    4.8 KB · Views: 612
the pair of 4k7 divide the supply volts in half. Well not quite half due to the small base current going to/from the cascode BJT.

The impedance seen at the emitter of the cascode is little affected by small changes in the base voltage dividing resistors.

You may want to consider using V grade jFETs to give more current capability to drive the load capacitance and the lower Gm of the Laterals. This may require the cascode voltage to be altered to bring the Vgrade temp back down to acceptable levels.
 
Last edited:
....You may want to consider using V grade jFETs to give more current capability to drive the load capacitance and the lower Gm of the Laterals. This may require the cascode voltage to be altered to bring the Vgrade temp back down to acceptable levels.

Exactly! V grade JFETS are better solution here and resistors in voltage divider should be something like 4k7 from base to rail and 2k7 from base to ground.
 
Thanks Andrew
The 4k7 was calculated if I use 40V rail . My question if I decrease the rail voltage to 24VDC what should be the right resistor value .
I think in that case the 4k7 it would be a bit to high .
Can you help with these please . Somewhere around 2k4-2k7 or so would be the right value but that need to be calculated I think .
How you wrote need a small current (which adjusted by the right value resistor )
I have a nice 1.5KVA 18-0-0-18VAC transformer , I would like to use that and build
a XF5 .
Thanks
Greets
 
Hello
I hate to give up!!!! Especially when I have most of the parts (power MOSFETs , transformer , heat sink etc.
Juma you suggest to use for cascode JET not BJT?
Please explain it me a bit .
I know these amp how I design to build it it would work with out cascode . But if I cascode better , easier to drive the 3 pat Toshiba mosfet .
Can you draw it done it would be easier to understand !
Please if you have time . Million thanks for it .
Thanks
Greets
 
Hello
Now I understand what did you advise Andrew ! I did look up what a heck is that V Fet .
I would like to stick with the 2SK170 / 2SJ74 JETs orig design I just want to cascode it how you advised & and made your drawing JUMA .
There are working amp 1pair JET drive even 3 pair IR mosfet , but if I cascode it that would help the JET .
Even the sim test better if is cascoded .Of course I would like to cascode for safety purpose .
Otherwise I will have to build it with out cascode , but I can not give up .
Greets
 
Last edited:
Ex-Moderator R.I.P.
Joined 2005
Yeah, sure thing
On other hand, we learn along the way
Maybe not not much, but still small details that could be the step to the next
Im certain I can hardwire it all, which I wouldnt dare to before
And Im confident I can hardwire other amps as well, or at least output stage

I have decided to do it and have no fear of burning it
If done hardwire it should be fairly easy to do changes here and there
If I loose my mosfets its not the end of the world
Just rebuild it with cheap IR devices, and it will work without problems
I will be happy to at least have given it a try
Good luck to you, you need it

But if you have hints about this or that here and there, please tell us about it
No strings attached
 
Please Andrew , it a big waste to have all the matched mosfet and not to use it and not to talk about the 25kg heat sink and over sized power transformer .
If no one help I will still go on and build it (with out cascode)
Why so hard to let me know the correct value resistor instead we just talking about nothing .(Sorry about but that is the reality )
Juma you already draw the cascode mode for the 40V rail power supp .
Please just let me know the right value resistors for the 24V rail power supp .
Even if the X for some reason not working I can used in bridge mode!
Greets
 
Ex-Moderator R.I.P.
Joined 2005
Garbobela, I think its good that you question things like this, making us aware we may need to investigate further
But its not fair to ask for a ready to go solution
But sure, some hints would be nice

Andrew is entitled to take that position, and perfectly understandable
Tho I like Nelsons approach much better, small hints here and there, to make us think for ourselves, and always positive about it
Reminds me of Miles Davis
He just played a single well chosen tune, and the band could play whatever they liked
His genious was that this special single tune was so dominant that it steered the band in the direction he expected, just one single tune was enough
Im rambling, sorry

Garbobela, I will advise you to make your board with copper side up
That way it will be easy to mod the curcuit without much trouble, as in hardwired ;)
 
Last edited: