[Congo5]

  I have started learning to build simple amps, before I tackle some of the newer and better designs.
Am trying do drive Fostex T-50's so need about 250ma(should have said 250mw) at 50ohms
 
A couple questions.
todays project is:
split rail 24v power supply ..450ma
basic cmoy but with out a virtual ground
gain is 5x
sezekers 3 channel mosfet follower
24v 50ohms source resistor
7.8v across resistor = 0.156A
using 100ohm dummy load
power supply holds with in .5v during test
Am using a signal generator and  oscilloscope:
1khz sinewave input at 1v max
cmoy shows flattening of negative at 5v output
mosfets start going flat at 1.5v... negative wave only
7v before the top starts to deform

should this have two mosfets to sink current?
otherwise how could it sink more than half the signal?
why does the negative give out first?
 
any pointers would be great.
Thanks

 

[Nagasaki_Kid]

isn't a mosfet buffer cmoy just a tori amp?

 

[Congo5]

I have not tried the Tori.....It looked interesting but could not find any writeups about it.
 
Am using a CMOY to drive this:

The Szekeres VE

 
from:  http://phonoclone.com/diy-sze.html

Tori from http://wiki.faust3d.com/wiki/index.php?title=DIY_MOSFET_Amplifiers_Overview

 
 
BTW do you know what the symbol is on the bottom 0.2A IDC?
 
Thanks

 

[nikongod]

Simple mosfet followers almost never swing symmetrically - it takes a LOT of trial and error or calculating to find the perfect gate voltage to get everything to line up properly. Most people designing stuff with simple mosfet followers just accept the fact that maximum voltage swing does not matter once you have enough (accounting for a decent overhead) and spend more time listening to the beautiful music. 
 
0.2A (not 0.02A...) is the "constant current source" current in the TORI. 
 
On that note, have you ever put 250ma (or whatever it was) into the T50? yea, that's a no. Have you ever put 5v into the t50? yea, that's a no. 
 
Put another way - rather than measuring your amplifier to see if you can meet the standards set up by (who cares? Why do you trust this person? Because he can talk a lot on the internet?) measure your headphones and figure out your own standard. Hook them up to a power amp, and play a 1khz sine through them 20db louder than you normally listen (use an SPL meter). Measure how much voltage you use and calculate the current. 
 
After you measured how much you need to drive your headphones much louder than you need, do you have enough? 

 

[Congo5]

thank you ...for your time and answers...nikongod.
 
No I have not determined max voltage need for t-50's..good idea, will do.
I do know that c weighted 100db is more than I will use...
 
I plays loud now but I could hear distortion so thought I would put it on the scope ...
I have built this six times with changes and do like the sound..keep trying to improve it..
will measure another one and see if its the same.. this one starts clipping at 1.5v..the output stage not cmoy..
 
my question is how can one MOSFET sink what two put out at the same constant current.
why did they not use two on the ground channel?
or maybe it isn't needed I just haven't built one right yet.....
.02amps was mistype..
 
None of this is real important, I just wanted practice before screwing up one of the newer designs.
 
  On my list to build is:
dynalo but would have to make the board...
the O2 which looks simple and easy..and portable if needed
maybe the Amp Camp Amp #1, boards and parts available..
also the krell ksa5 klone looks very inviting....
 
what amp would you vote for?
 
Thank you.
 

 

[Congo5]

Quote:

Simple mosfet followers almost never swing symmetrically

Ok I get that ......just have enough reserve power there..
 
Oh, I'm the guy who made up 250mw....
just in my head .....what I might use x10 = target....

 

[Congo5]

OK today I measured the other amp and headphones..
1k sine wave from BK Precision 3020
 
modded t-50's....3.23v=104db      4.14v = 110db
 
this amp has 27v input >lm317 regulator (adjustable) > sijosae's discrete rail splitter>cmoy>szekers VE
3.23v is clean on O-scope...
 
so if math is correct 3.2v squared divided by 50ohms is 204mw ....0.064amps
 
4.14v would be 342mw
 
If so than a target of  250mw sounds right to me>>>?