New Doug Self pre-amp design... - Page 28 - 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 17th May 2012, 04:41 PM   #271
diyAudio Member
 
DouglasSelf's Avatar
 
Join Date: Mar 2007
Quote:
Originally Posted by Bob Cordell View Post
Hi Doug,

Fair enough, but the answer "it worked in my lab" is not always sufficiently conservative.

Bob
True, but my real answer is "It has worked in production for 16 years." which I think carries a little more weight.

The advantage is lower distortion with less compensation.
__________________
Douglas Self
  Reply With Quote
Old 17th May 2012, 05:51 PM   #272
FdW is offline FdW  Netherlands
diyAudio Member
 
FdW's Avatar
 
Join Date: Feb 2009
Location: Both feet on the ground (near Gouda)
Quote:
Originally Posted by DouglasSelf View Post
True, but my real answer is "It has worked in production for 16 years." which I think carries a little more weight.

The advantage is lower distortion with less compensation.
There is no advice (at the least in the TI datasheet) for the 22pf capacitor; the datasheet shows graphs for 0pf to 22pf (and at some points up to 47pf). This is what datasheets do, they show what can and may (or even be useful) to do, not what must be done. In ‘your’ application you can do whatever you think is right, my opinion is, less and smaller capacitors is better (as opposed to more and bigger) (anyway most of the time).

The exact text in the TI datasheet is:

These operational amplifiers are compensated internally for a gain equal to or greater than three. Optimization of the frequency response for various applications can be obtained by use of an external compensation capacitor between COMP and COMP/BAL. The devices feature input-protection diodes, output short-circuit protection, and offset-voltage nulling capability with use of the BALANCE and COMP/BAL pins (see the application circuit diagram).
  Reply With Quote
Old 18th May 2012, 12:17 AM   #273
owdeo is offline owdeo  Australia
diyAudio Member
 
Join Date: Jun 2003
Location: Sydney
Quote:
Originally Posted by ChristianThomas View Post
I would say no films for inductance reasons, and because we have been gifted C0Gs which fulfil our every desire.
My desire for 100nF or larger COG does not seem likely to be fulfilled anytime soon - they don't seem to be available from any supplier I know of in small qtys...
Quote:
Originally Posted by ChristianThomas View Post
I would probably set the RC at around 120Hz, working from the datasheet which gives you an R of about 5.6 ohms. Since that's miles away from the 100+ohms needed to stop the amp working you have plenty of scope for playing around. Take it up to 27 ohms which is the equivalent of a 120 + 47uF that I once used years ago and which worked reasonably well.
Interesting. I was just looking over some old opamp based preamp designs that I recall as sounding good from the long-defunct Electronics Australia magazine. I see in the power supply to the main opamps they have used 68R + 100uF, about the same corner freq.... I wonder if this is coincidence and how deliberate the choice was on their part...

Here are the schematics and specs for their flagship no-frills preamp from 1990, designed by Rob Evans. There are a couple of other interesting things about this design of relevance to this discussion that prompted me to post them:

1) The strong emphasis on "Low Impedance Design" to achieve best noise performance - 22 years ago. Note the feedback Rs, right at the limits and probably lower than they should be so causing extra distortion?

2) Value of compensation caps and feedback caps around the 5534s - the feedback caps are quite large, presumably to define the upper rolloff. Is this perhaps a bad idea?

Would be interested to hear what all you very knowledgable chaps think...
Attached Images
File Type: jpg EAProSer2Pre1.jpg (506.8 KB, 539 views)
File Type: jpg EAProSer2Pre2.jpg (638.1 KB, 514 views)
File Type: jpg EAProSer2Pre3.jpg (665.4 KB, 488 views)
  Reply With Quote
Old 19th May 2012, 12:10 AM   #274
owdeo is offline owdeo  Australia
diyAudio Member
 
Join Date: Jun 2003
Location: Sydney
Thinking about this more, is putting large Rs in series with supplies to opamps (which I assume all have class B output stages) really a good idea?

Apart from the LP filtering, isn't it going to introduce a varying voltage drop, meaning the supply voltage is modulated by the current draw of the opamp's output stage? Assuming the load on the opamp output is heavy enough, just when there is a loud passage and the additional current draw is at its highest, the supply rails sag....surely not a good thing and potentially audible? Can any of the experts offer their thoughts on this?
  Reply With Quote
Old 19th May 2012, 02:52 AM   #275
diyAudio Member
 
Join Date: Feb 2006
Location: On a hill, in a wooden shack, next to the woods, in Somerset.
Quote:
Originally Posted by owdeo View Post
Thinking about this more, is putting large Rs in series with supplies to opamps (which I assume all have class B output stages) really a good idea?

Apart from the LP filtering, isn't it going to introduce a varying voltage drop, meaning the supply voltage is modulated by the current draw of the opamp's output stage? Assuming the load on the opamp output is heavy enough, just when there is a loud passage and the additional current draw is at its highest, the supply rails sag....surely not a good thing and potentially audible? Can any of the experts offer their thoughts on this?
D'you know, I was having just that same thought either this morning or last night? Simple thoughts at first, because of course you want keep output impedance low and the supply stiff. And then more quantitive thoughts about how much they would move, and how quickly, and whether or not there would be a optimum between the two.

But, while a good way to easily model it hasn't yet popped into my head, it would seem to me to boil down to such and such a dV/dI, at whatever bandwidth. And that looks suspiciously like tailoring the output impedance. And from that point of view, short of putting the huge capacitor on the output alone, minimizing the effect of the inductance would imply a resistor. And since that would not do the whole job, a capacitor as well.

It could be well worth looking into, though the question still remains of whether you want the rejection to be getting poorer above120Hz (as it does) just so you could get a smooth output impedance. And, of course, the loads really aren't that low usually, though they are effectively just about in parallel for each op amp (ie, four or five of them will have their output voltages change by approximately the same amount.) Just some hand calculations could help here (though I haven't even done those). Oh, and you have to look at where the current into the 317 is coming from. Now, over to the experts.
  Reply With Quote
Old 19th May 2012, 09:53 AM   #276
AndrewT is online now AndrewT  Scotland
diyAudio Member
 
Join Date: Jul 2004
Location: Scottish Borders
To help explain look at Gootee's posts on decoupling and currents and transients.

I'll present a short version of Gootee's thoughts/experiments.
Attach an opamp to a regulated supply with a few short wires and PCB traces.
Attach a value for resistance and if you want extra complication capacitance and inductance. Now attach that tiny decoupling cap to the opamp power PINs.
The resistance/impedance prevent the power PINs seeing the regulated supply at VHF. The power PINs only see the decoupling. That is where the opamp gets it's current from to deliver to the load when a fast transient demands that current.

At the slower scale of bass notes, where small impedances have little effect then the power PINs do see way back along the supply lines. Then the current comes from a combination of all the decoupling and the regulator and if fitted any smoothing after the regulator.

These are all RC or LC filters as far as the power PINs are concerned.
One simply selects the different R&C to suit the opamp demands.
__________________
regards Andrew T.
  Reply With Quote
Old 19th May 2012, 11:59 PM   #277
owdeo is offline owdeo  Australia
diyAudio Member
 
Join Date: Jun 2003
Location: Sydney
Thanks Andrew but I haven't been able to find those posts using the search function - could you post a link?

As I understand it at VHF the opamp shouldn't be doing anything unless it's oscillating. The small decoupling cap(s) are there primarily to prevent this by the mechanism you've just described, and Christian has stated their self-inductance is important which makes sense here. I fail to see how on their own they are very significant at audio frequencies though.

My pondering was on what happens when the opamp is amplifying audio frequencies and there is sufficient loading that the class B output stage draws significant currents in addition to its quiescent draw.

With a power amp, the (class B) output stage causes nasty half-wave rectified signals to be superimposed on the supplies, hence there is usually some RC filtering to the earlier stages to stop it getting in. But with an opamp you don't have this option, and it seems as though putting large Rs in series with the (common to all internal stages including class B output) supply could make things worse.

Would Mr Self and/or Mr Cordell be able to help steer us back on the straight and narrow here?
  Reply With Quote
Old 20th May 2012, 12:56 AM   #278
jcx is offline jcx  United States
diyAudio Member
 
Join Date: Feb 2003
Location: ..
you need to think impedance, frequency, compare regulator, bypass, PS trace RLC and op amp PSRR

DIY: solid state is good on regulators

below 1 kHz the common LM317 can have output Z well below 100 mOhm, somewhere in the upper audio depending on bypass C, possible damping Zobel the Z may peak to a few Ohms

so a heavy op amp current like 10 mA only gives <100 mV on the PS rail at the worst case frequency, this can be kept to 10 mV if you design the output bypass properly

op amp PSRR is at worst equal to the loop gain (for the rail the internal compensation cap is attached to)

a 20 MHz GBW op amp has 60 dB gain at 20 kHz

so for unity gain, today's faster op amps, a supply with ~1 Ohm max Z from proper
bypass/damping you only have 10 uV of PS loading PSRR/"noise"

a common assertion is that music has ~3 kHz power bandwidth - it is possible to trim the LM317 circuit for less than 100 mOhm Z below 3kHz
with those assumptions you only typically have ~ 1 uV PSRR error without "extreme" measures
  Reply With Quote
Old 20th May 2012, 01:13 AM   #279
Bonsai is offline Bonsai  Taiwan
diyAudio Member
 
Bonsai's Avatar
 
Join Date: Jul 2003
Location: Taipei, Taiwan
Some nice points raised here. I agree with you jcx and AndrewT - the PSRR of modern op amps really can reduce some of these problems to insignificant levels if a few simple rules are followed.

That said, I do not believe 'stiff' power rails are needed and prefer to decouple each supply pin with a 22 Ohm resistor and a 100uF to a dedicated supply bypass ground rail. This results in only low frequency PSU currents flowing in the supply lines and here the op amp supply rejection is at its best (well above 100dB in many devices). Doing this, you also kill the wide band noise of the LM3xx devices, which after decoupling of the Vref pin can still be around 45uV on 15V rails.

It also helps to bias the op amp output stages into class A (nice bootstrapping trick from you btw jcx!) since this also results in very clean signals on the supply rails.

So, in summary, decouple locally, run op amps in class A, use a dedicated decoupling ground return and stick with LM3xx. No need for super regs, shunt's etc.
__________________
bonsai
Amplifier Design and Construction for MUSIC! http://hifisonix.com/

Last edited by Bonsai; 20th May 2012 at 01:15 AM.
  Reply With Quote
Old 20th May 2012, 03:02 AM   #280
diyAudio Member
 
Join Date: Sep 2006
Quote:
Originally Posted by owdeo View Post
Thanks Andrew but I haven't been able to find those posts using the search function - could you post a link?

As I understand it at VHF the opamp shouldn't be doing anything unless it's oscillating. The small decoupling cap(s) are there primarily to prevent this by the mechanism you've just described, and Christian has stated their self-inductance is important which makes sense here. I fail to see how on their own they are very significant at audio frequencies though.

My pondering was on what happens when the opamp is amplifying audio frequencies and there is sufficient loading that the class B output stage draws significant currents in addition to its quiescent draw.

With a power amp, the (class B) output stage causes nasty half-wave rectified signals to be superimposed on the supplies, hence there is usually some RC filtering to the earlier stages to stop it getting in. But with an opamp you don't have this option, and it seems as though putting large Rs in series with the (common to all internal stages including class B output) supply could make things worse.

Would Mr Self and/or Mr Cordell be able to help steer us back on the straight and narrow here?
Hi owdeo,

I just bypass every op amp's + and - rails to ground with a 0.1uF ceramic and usually sprinkle around some 10uF to 100uF electrolytics. The ESR of the electrolytics acts to damp any resonances formed by the low-ESR ceramics and rail inductances. The observation about half-wave-rectified currents from the class-B output stages of the op amp is astute. However, bear in mind that if these cause significant noise/distortion on the rails that is not obviated by PSRR, then measurable distortion will result. If you achieve very low measured distortion, this phenomenon is probably not at work. My experience is that this phenomenon does not result in measurable distortion. However, I must also say that I do not generally have op amps driving really low-impedance loads. I used this basic bypassing approach in my THD analyzer, where of course extremely low distortion was important and was achieved.

I do not subscribe to the idea of just having a single bypass capacitor from rail-to-rail on each op amp. This will tie the rails together at high frequencies, but may allow the rails to flop around with respect to the local ground. At high frequencies, I like to have it all nailed together locally.

Cheers,
Bob
  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
Doug Selfs NE5532 Power Amp. Thoughts anyone ! Mooly Chip Amps 87 15th June 2012 08:08 PM
blameless amp and doug self streetera Solid State 17 16th July 2007 10:06 PM
4th Ed. of Audio Power Amplifier Design Handbook (Doug Self) nickds1 Solid State 4 9th November 2006 12:02 PM
Doug Self Advanced Pre Amplifier PCBs macka Swap Meet 0 9th October 2006 07:41 AM
Help me design pre-amp PS hardcore Tubes / Valves 7 10th November 2004 06:32 AM


New To Site? Need Help?

All times are GMT. The time now is 12:45 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