♫♪ My little cheap Circlophoneİ ♫♪ - Page 99 - diyAudio
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 1st June 2012, 04:59 PM   #981
Elvee is offline Elvee  Belgium
diyAudio Member
 
Elvee's Avatar
 
Join Date: Sep 2006
The compensation amounts to a dominant pole scheme, but unlike the classical Miller of a "normal" amplifier, it doesn't have a negative impact on the slew rate, which is good.
It does have one serious drawback though: the corner frequency is pushed much lower in frequency, and as a result the loop gain*bandwidth product is only ~130KHz.

This means that the distortion is much higher.

Now we just have to find a scheme combining the best of both worlds...
__________________
. .Circlophone your life !!!! . .
♫♪ My little cheap Circlophoneİ ♫♪
  Reply With Quote
Old 1st June 2012, 05:38 PM   #982
diyAudio Member
 
keantoken's Avatar
 
Join Date: Aug 2006
Location: Texas
Blog Entries: 2
I don't know why you say the GBWP is so low. In simulation it is 2.56MHz, and distortion at 1KHz is below .001% at 12V into 6R.

Remember I lowered R31 and R33, causing the input stage RC comp to become about 15 times bigger, but this is just an impedance thing, the loading hasn't changed. This scheme makes the amp as fast as it will go without significant overshoot.
  Reply With Quote
Old 1st June 2012, 11:17 PM   #983
diyAudio Member
 
keantoken's Avatar
 
Join Date: Aug 2006
Location: Texas
Blog Entries: 2
All my sims so far suggest CFP is only useful below 10KHz and it hurts BW and makes distortion a little worse at HF. So I'm going without it for the time being.

There are several ways to make this amp stable without significant overshoot, but this one has worked the best so far. My current simulation has a GBWP of 7.3MHz and a slew rate of 110V/uS. 1KHz distortion hasn't changed, but it is a whole lot faster. This causes stipulations.

At this point the input would become a GDO and would need to be shorted to ground with several hundred pF, and even that may not be enough. An input buffer would trivialize this, but that adds a transistor to the circuit and may complicate DC offset. The Jfet buffered schematic I've seen would probably work well for this. Ground plane may also help.

THD at 1KHz is still about .0034% at 12V into 6R. BUT, at 20KHz it has been brought down from .02% to .006%.

There is a problem with the speed. When slewing the VAS tends to go into class B, and this causes a jump in output current bias while switching, up to several amps. Thus I've fitted the Schottkeys across the VAS. This limits the slew rate to 60V/uS, except it's the diodes that are saturating, not the VAS or the output stage. Small signal behavior is the same.
Attached Images
File Type: png Circlophone_keancomp_schem.png (124.8 KB, 223 views)

Last edited by keantoken; 1st June 2012 at 11:32 PM.
  Reply With Quote
Old 3rd June 2012, 07:31 AM   #984
Elvee is offline Elvee  Belgium
diyAudio Member
 
Elvee's Avatar
 
Join Date: Sep 2006
Quote:
Originally Posted by keantoken View Post
Remember I lowered R31 and R33, causing the input stage RC comp to become about 15 times bigger, but this is just an impedance thing, the loading hasn't changed.
OK, I had missed the resistors change, now I am a little under 1MHz, but that's probably down to the different types of transistors you used.
A problem with low value resistors is the bias current, which is only moderately improved compared to a non-CFP stage.
__________________
. .Circlophone your life !!!! . .
♫♪ My little cheap Circlophoneİ ♫♪
  Reply With Quote
Old 3rd June 2012, 08:17 AM   #985
diyAudio Member
 
keantoken's Avatar
 
Join Date: Aug 2006
Location: Texas
Blog Entries: 2
In my simulations distortion increases by 20 times if you have a 25k source impedance, which will occur if you use a 50k volume pot.

CFP helps with this, but I think an input buffer may be better for this specific problem. We could buffer the entire LTP with the same number of transistors. With this schematic distortion only increases by 2.5% when 25k input impedance is used.
Attached Images
File Type: png Circlophone_Buffer_Kcomp.png (111.6 KB, 177 views)
  Reply With Quote
Old 3rd June 2012, 01:32 PM   #986
forr is offline forr  France
diyAudio Member
 
forr's Avatar
 
Join Date: Dec 2004
Location: Next door
if you have a 25k source impedance, which will occur if you use a 50k volume pot.

It should 25 kOhm maximum for a 100 kOhm pot with a preceding voltage source, 50 kOhm from cursor to ground in parallel with 50 kOhm from cursor to source.
With modern audio components, one should never use more than 10 kohm for volume control pots.

Last edited by forr; 3rd June 2012 at 01:37 PM.
  Reply With Quote
Old 3rd June 2012, 02:04 PM   #987
diyAudio Member
 
Join Date: Mar 2012
Location: ISANDO GAUTENG
Good day!
I tried to follow you clever guys on this amp and now I know why there is a strong movement back to tubes. Like the old Mullard 5/10 or 5/20 circuits!!!!!!!!!!!!!! Can it still be called the circlophone? JUst wondering!!! I still have the book published by Mullard and remembered them 5/10's feeding the p.a. systems of Johannesburg station. I know I serviced them in 1968. When I was in the college we had them and you could hear 10 watts mono for about three kilometres on a farm . Oh for the quiet life!!! I saw an ad for the latest from Korea and only the digital part was transistorised.
  Reply With Quote
Old 3rd June 2012, 02:14 PM   #988
diyAudio Member
 
keantoken's Avatar
 
Join Date: Aug 2006
Location: Texas
Blog Entries: 2
I stand corrected.

I have a 100k stepped attenuator I'm using, my amp has an input impedance of 150k (buffered, so source impedance is hardly an issue). Before this, changing from a 10k pot to a 50k pot made a big improvement. I suspect it is soundcard issues since I'm using an on-board card (stac9227). Lately I've decrease the soundcard volume and increased my amp volume, this too seems to help.
  Reply With Quote
Old 3rd June 2012, 03:48 PM   #989
Elvee is offline Elvee  Belgium
diyAudio Member
 
Elvee's Avatar
 
Join Date: Sep 2006
Quote:
Originally Posted by keantoken View Post
In my simulations distortion increases by 20 times if you have a 25k source impedance, which will occur if you use a 50k volume pot.

CFP helps with this, but I think an input buffer may be better for this specific problem. We could buffer the entire LTP with the same number of transistors. With this schematic distortion only increases by 2.5% when 25k input impedance is used.
The peculiarity of the input impedance has been signaled from the beginning:
Quote:
Originally Posted by Elvee View Post
Here are some additional characteristics and information:
.....
The input resistance is ~10K, but the source impedance is preferably kept below 1K, because of the base current modulation caused by the bias servo.
It is an annoyance, but a relatively minor one in a semiconductor environment, where impedances are best kept low: a 4.7K pot brings practically no degradation

On an other subject, I was aware that a "decompensated" version could be made stable in sim, but in practice I was never able to reach stability without the additional components.

However, my tests were made with closed loop gains ranging from 10 to ~20.
It might be stable at higher gains, I didn't test it.
__________________
. .Circlophone your life !!!! . .
♫♪ My little cheap Circlophoneİ ♫♪
  Reply With Quote
Old 3rd June 2012, 04:18 PM   #990
diyAudio Member
 
Join Date: Nov 2007
Location: Dallas
The more you clamp down on a marginally stable open loop,
the worse Bode seems to peak right there at the end, where
phase gets all shifty and stuff...

In Class B, you got a reset that quenches some modes of
parasitic oscillation and ringing each crossing. But if there
is a huge peak in Bode, even such a reset might come too
little too late.

The differential mode open loop gain may be much lower in
crossing, because modulation from the common mode closed
loop works to smother it. Less open gain is more stable. But
in this case, is it real? Does it apply? Are we hiding a closed
loop peak that we would see if we look away from crossing?

Can we allow common closed within our Bode of differential
open. Or must we analyze all four cases of open open, open
closed, closed open, closed closed?

Because open differential closed common gain is higher off-
center, and higher for large signals that do not dwell in the
B crossing, is it still stable with a small DC offset to inhibit
the crossing reset? Or would that just be the same Bode as
both loops open?

Last edited by kenpeter; 3rd June 2012 at 04:41 PM.
  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


Similar Threads
Thread Thread Starter Forum Replies Last Post
Need help with OB design - cheap, cheap, cheap djn Multi-Way 38 30th May 2010 06:26 AM
Scosche SPL meter: Cheap find with potential, or cheap junk? theAnonymous1 Everything Else 5 11th October 2006 03:40 AM
Alpine cda-9831 vs. 9833 vs. 9835 cbecker33 Car Audio 3 10th December 2004 12:47 PM


New To Site? Need Help?

All times are GMT. The time now is 12:31 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright İ1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2