Go Back   Home > Forums > Amplifiers > Solid State

Solid State Talk all about solid state amplification.

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 19th January 2009, 08:17 PM   #691
diyAudio Member
 
keantoken's Avatar
 
Join Date: Aug 2006
Location: Texas
Blog Entries: 2
Quote:
Originally posted by AndrewT
Hi Kean,
could you substitute a red LED for the twin diodes?

I hear that a red LED has a similar temp co to a BJT.

I would like to see both the ripple and the temp variations.
The ripple is indeed less.

This is the tempco simulation. This doesn't show in the graph, but the current through the LED CCS actually rises as it gets warmer, while the others get lower. I hope I did this right. The LED is biased at around 10mA.

I used a red LED model I found via Google, not sure how accurate it is.

- keantoken
Attached Images
File Type: gif betterccs_tempco2.gif (22.8 KB, 345 views)
  Reply With Quote
Old 19th January 2009, 08:20 PM   #692
diyAudio Member
 
keantoken's Avatar
 
Join Date: Aug 2006
Location: Texas
Blog Entries: 2
And here is the ripple.

- keantoken
Attached Images
File Type: gif betterccs_tempco3.gif (16.4 KB, 326 views)
  Reply With Quote
Old 19th January 2009, 08:32 PM   #693
diyAudio Member
 
keantoken's Avatar
 
Join Date: Aug 2006
Location: Texas
Blog Entries: 2
I'll assume I've got the right to double post now since I'm the man with the sim.

I was googling and found this:

Quote:
For protecting mosfets (zener as a shunt across gate and source) I have used with good success a reverse biased base-emitter junction of generic BC550 or BC560. Capacitance is low and the zener voltage is quite the same across different devices.

Cheers,

Gerhard
Has anyone tried this? I don't think a simulator will do this right. I think I've read this in an application note or something also.

Perhaps mr. Christer or someone can set us up some good measurements.

- keantoken
  Reply With Quote
Old 19th January 2009, 08:42 PM   #694
diyAudio Member
 
ostripper's Avatar
 
Join Date: May 2008
Location: Albany , NY (smallbany)
I did the diode vs. LED test and the LED wins..
(see attachment)
PSRR wise
In an actual circuit with the cap (like the FA2 I'm listening to)
the LED beats the diodes by 10 more db(most likely the cap)
and has almost no thermal drift. BUT "strait up' with no
decoupling, the led beats the diodes by just 1 DB.

Next post I will use the decoupling techniques from the nakamichi
620 which compensates the current source differently.
OS
Attached Images
File Type: gif 2diodes.gif (20.4 KB, 323 views)
  Reply With Quote
Old 19th January 2009, 09:18 PM   #695
diyAudio Member
 
keantoken's Avatar
 
Join Date: Aug 2006
Location: Texas
Blog Entries: 2
Which LED model are you using?

- keantoken
  Reply With Quote
Old 19th January 2009, 09:52 PM   #696
diyAudio Member
 
ostripper's Avatar
 
Join Date: May 2008
Location: Albany , NY (smallbany)
Quote:
Which LED model are you using?
The QTLP690 fairchild 1.8V red led.

In the whole scheme of things
the PSRR/ripple factor of the diamond CCS's really only
help to stabilize the
biasing of the OP stage (vbias). The whole diamond buffer
is corrected by NFB so the current sources are just overkill.
In the Krill, they are using a dual voltage supply even with
current sources ,but the Krill does not wrap the current stage
with the NFB loop.

Attached below is the (I think) close to final plots that I
really like. (perfect gain to 100k and 100db phase margin).
The diamond seems to extend the bandwidth of any
standard amp by some degree and allows for a unloaded, low
current VAS as a bonus..
OS
Attached Images
File Type: gif fa3lg.gif (22.3 KB, 307 views)
  Reply With Quote
Old 20th January 2009, 09:55 AM   #697
diyAudio Member
 
nelsonvandal's Avatar
 
Join Date: Nov 2006
Quote:
Originally posted by ostripper



Attached below is the (I think) close to final plots that I
really like. (perfect gain to 100k and 100db phase margin).
The diamond seems to extend the bandwidth of any
standard amp by some degree and allows for a unloaded, low
current VAS as a bonus..
OS
DB's do sound good. I find them better than simple EF's. If you listen to an open loop DB you'll find they sound very transparent.
  Reply With Quote
Old 20th January 2009, 06:02 PM   #698
diyAudio Member
 
Steve Dunlap's Avatar
 
Join Date: Sep 2008
Location: Here
Hi ostripper. I'm trying to catch up here after a couple of days without Internet service.

Run open loop, as I do, my amps have a very low output impedance and excellent speaker control. I prefer the sound without feedback so that is how I build. Distortion specs are also very good run open loop and there is no chance for the stage to become unstable or oscillate. As an added bonus, if the output to the speaker gets shorted you don't loose the feedback and go OLG.

I never really worried about what would happen if the compensation (diode string in my case) failed. If it shorted it would simply reduce bias. This would affect the sound some, but do no real damage. If it opened it will blow the line fuse. With less than 0.1mA of current the diodes should never fail. It is always possible for a connection to break however, so I will add a resistor in parallel with the whole string to set a maximum bias in this event.

Thanks for pointing out this potential problem.

I am attaching a simulation result showing the output stage only, run open loop with a 20V input. As you can see it shows a 3dB down point at only 3 MegHz. A test on a real unit only gave about 1.5 MegHz. Still, I think that is good enough for audio.
Attached Files
File Type: pdf graph1.pdf (10.6 KB, 40 views)
  Reply With Quote
Old 20th January 2009, 07:29 PM   #699
diyAudio Member
 
ostripper's Avatar
 
Join Date: May 2008
Location: Albany , NY (smallbany)
Quote:
Thanks for pointing out this potential problem
I didn't mean to point out a problem , but I don't want to be
a contributer to burning amps .


I've tried both including or omitting the diamond current
stage in the NFB loop with little effect other than ,with the
isolated current stage, no change with load. (As you pointed out.)

Another interesting note is that by degenerating the darlington
(470r - 2.2k) like a standard EF type 2, "tames" a little
bump at 5MHz (I think that is where this thing "likes" to oscillate)
and sets the phase margin.

I already did the carlos thing and no fire , so I know it works. I tested what would happen if my vbias died and it
is the safest I could imagine.. just 300ma (severely overbiased).
Idiot proof, much better than a typical amp. A total short at
the output would just burn the basestopper resistors.

It seems DIYaudio has gone "krill crazy" these days, even
carlos (no trashbin or buckshot) likes it, 3 seperate diamond
topology threads... you started it all ,steve, thanks!
BTW..., yours(krill) is most like the Nakamitchi 620, except they use an EF'ed diamond which is the thermal compensation .
(attached)
OS
Attached Images
File Type: gif nk620gif.gif (81.0 KB, 283 views)
  Reply With Quote
Old 20th January 2009, 07:44 PM   #700
diyAudio Member
 
ostripper's Avatar
 
Join Date: May 2008
Location: Albany , NY (smallbany)
This is what I have so far, I can't seem to find any problems
with this. (comments welcome)
OS
Attached Images
File Type: gif fa4.gif (28.8 KB, 290 views)
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off



New To Site? Need Help?

All times are GMT. The time now is 02:18 PM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2