Power amp under development - Page 3 - 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 26th October 2004, 03:57 PM   #21
quasi is offline quasi  Australia
diyAudio Member
 
quasi's Avatar
 
Join Date: Sep 2004
Location: Adelaide, Australia
Default Sorry dinesh didn't see your post.

The 100pf capacitor between the diff pair serves to shunt any high frequency that may be here and amplified through the stages. This may not be a problem all the time but I have fixed troublesome amps this way before. I first got this idea from an Elektor cct I saw years ago. They used a resistor+cap but I found that a small capacitor worked much the same.

The 100 ohm trimpot serves to adjust the output offset caused by tolerance issues in the input stage. The feedback will only drive the offset down so far and I like to get it as close to 0.00 as I can. Once the setting is determined it can be replaced with a resistor network to suit.

The small capacitors on Q7 and Q8 shunt any RF away from their bases. Yes I am paranoid about the FETS oscillating. Your views about Q4 make sense. I'll just use a small capacitor though and omit the resistor.

Thanks for your feedback

Cheers
  Reply With Quote
Old 26th October 2004, 04:12 PM   #22
quasi is offline quasi  Australia
diyAudio Member
 
quasi's Avatar
 
Join Date: Sep 2004
Location: Adelaide, Australia
Umm ...I haven't explained the motivation behind a quasi-complimentary amp.

Thers is one compelling reason; I have literally a bucketful of new and used N channel MOSFETS including IRFP450, IRFP250, IXFH26N50, IRF820 (TO220), philips BUK series and assorted others. Thought I might as well use them.

Cheers
  Reply With Quote
Old 26th October 2004, 05:14 PM   #23
MikeB is offline MikeB  Germany
diyAudio Member
 
MikeB's Avatar
 
Join Date: Apr 2004
Location: GŁtersloh
Hi quasi !

The point is, that the 100pf is on the wrong side of the cascode,
it needs to be above. But a RC is better, like 330pf+330ohm or
something like that. You might use a 1n+330ohm, then make
cdoms smaller. Your feedback-cap (33pf) might be too big, think
more about 15pf.

Mike
  Reply With Quote
Old 26th October 2004, 05:31 PM   #24
Eva is offline Eva  Spain
diyAudio Member
 
Eva's Avatar
 
Join Date: Oct 2003
Location: Near the sea
Send a message via MSN to Eva
Each capacitor you place shunting the signal path to ground [or as a miller capacitor] with no resistor in series causes an additional pole to be introduced in the open loop response. If it's a RC network then it causes a pole-zero

Each transistor adds also a pole or a pole-zero to the open loop response. Each pole produces additional phase lag so phase margin is reduced. I wouldn't add extra poles to the circuit if I was you

In the other hand, those RC networks you've seen are not intended to reduce RF pickup, they are part of the compensation strategy to reduce gain at high frequencies to a suitable value and make the circuit stable [and this pole-zero compensation strategy was common in Elektor circuits]
  Reply With Quote
Old 26th October 2004, 05:49 PM   #25
MikeB is offline MikeB  Germany
diyAudio Member
 
MikeB's Avatar
 
Join Date: Apr 2004
Location: GŁtersloh
I have some very good experience with these RC-networks,
in some circuits they did some "mystical" to the sound, and they
really help stabilizing.

Mike
  Reply With Quote
Old 26th October 2004, 05:57 PM   #26
Eva is offline Eva  Spain
diyAudio Member
 
Eva's Avatar
 
Join Date: Oct 2003
Location: Near the sea
Send a message via MSN to Eva
Maybe not so mystical? they just help in reducing open loop gain above audio band and improving phase margin. That's why I use them, but the wrong component values may actually reduce phase margin
  Reply With Quote
Old 26th October 2004, 06:06 PM   #27
MikeB is offline MikeB  Germany
diyAudio Member
 
MikeB's Avatar
 
Join Date: Apr 2004
Location: GŁtersloh
Okay, mystical to the sound... I was surprised by the improvement
to the sound. As i found out with my last amp, phasecompensating
is extremely important, with simple changes this amp changed from
unpleasant sounding to extreme good sounding.

What i didn't get, is there a formula for the ratio R <-> C ?
My method for adjusting is not very scientific, i find these values
by trying with acsweep in sims...

Mike
  Reply With Quote
Old 27th October 2004, 06:15 AM   #28
quasi is offline quasi  Australia
diyAudio Member
 
quasi's Avatar
 
Join Date: Sep 2004
Location: Adelaide, Australia
Default Thanks Eva and MikeB

Quote:
those RC networks you've seen are not intended to reduce RF pickup, they are part of the compensation strategy to reduce gain at high frequencies to a suitable value and make the circuit stable
Aaaaaah .....thanks.

Quote:
Each capacitor you place shunting the signal path to ground [or as a miller capacitor] with no resistor in series causes an additional pole to be introduced in the open loop response. If it's a RC network then it causes a pole-zero
Eva are you saying that all by capacitors on the base-emmitter junctions should have a resistor? None of the amps I have seen do this, maybe beacuse there is always a resistor feeding the base anyway.

Quote:
The point is, that the 100pf is on the wrong side of the cascode,
I forgot to shift it when I inserted the cascode. Ok the resistor is in.
  Reply With Quote
Old 27th October 2004, 06:49 AM   #29
diyAudio Member
 
Join Date: Aug 2003
Location: Prague,Czech Republic
Capacitors connected in driver stage must be on CB, not on BE.
  Reply With Quote
Old 27th October 2004, 08:49 AM   #30
dinesh is offline dinesh  India
diyAudio Member
 
Join Date: Jun 2004
Location: Bangalore
Quasi,

Loose the caps across the emitters of the cacode. The input impedance at those nodes is ~26 ohms (since it appears you are running 1mA per leg of the diff pair). Any pole introduced by the cap is way up in frequency. There are two reasons for the caps -- as Eva and the rest have mentioned -- the first for compensation. In which case the cap goes across the 1K ohm load resitors on the input diff pair. You *donot* wan't to put a cap alone -- a series RC is better. Why ? Because, with a large step applied at the input (yes I know that the input RC would prevent this) the amp will slew. Slewing is bad because the output is not related to the input.

The other place small caps are placed is across the input of the diff-pair, to avoid RF getting into the inputs and being recrtified by the base-emitter of the inputs (you get to hear AM broadcast radio from the AMP when this happens). This is rarely required in discrete designs (often needed in BJT op-amp designs) because the current levels in the input stage is fairly high (you are at 1mA). This keeps Vbe at fairly large levels and prevents RF rectification.

There is no needed to place these caps across the B-E junctions of the emitter followers Q7 and Q8.

I know I am repeating this.... If you want to keep the pot to zero the offset you make want to place resistors lets say 100ohms from emitter to the wiper point. This is to prevent the unforseen. Lets say the wiper opens for a short instant of time (pots are crummy). Q4 base goes high . Q4 tuns off. Power amp output becomes stuck at the negative rail, loud pop from speaker ....

Lets say you did have an offset and you moved the pot. Notice that you are changing the emitter degeneration resistor. This means that you are changing the open loop gain. Right now you probably have 1mA in the leg (Re = 26/Ie ~ 26ohm). Your gain for the first stage alone is 1000/(50+26) on this stage with pot at center ~ 13 (hope you are not superstitious). With wiper at other extreme it would be 1000/(100+26) ~8, and at the other 1000/26 ~40. If you do wan't to do a zero offset you may be better off injecting a small voltage into the negative input (create a small variable +ve, -ve voltage from supplies and feed it to the top of the 100uF/25V cap).

--
Dinesh
  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 02:56 AM.


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