Discrete Opamp Open Design - Page 40 - diyAudio
Go Back   Home > Forums > Source & Line > Analog Line Level

Analog Line Level Preamplifiers , Passive Pre-amps, Crossovers, etc.

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 28th August 2012, 01:15 PM   #391
gpapag is offline gpapag  Greece
diyAudio Member
 
gpapag's Avatar
 
Join Date: Nov 2002
Location: Athens-Greece
Quote:
Originally Posted by bcarso View Post
Thanks Jan! Yes, I fished through email and found that it was On Semi that I'd seen first, although the link to the Electronic Products issue has long expired. And I remember that bit about the Sanyo on-chip 220mohm resistors that tended to blow up

EDIT: I also remembered the suggested circuits, which I immediately wanted to do differently

A cool thing (npi) would be to infrared-image the die and use that information. Certainly could be fast response

Brad
__________________
["Second Law is a bitch." - SY] ["The Road To Heaven:Specify the performance & accept the design. The Road To Hell:Specify the design & accept the performance"-Bruno Putzeys]
  Reply With Quote
Old 28th August 2012, 01:16 PM   #392
coluke is offline coluke  Italy
diyAudio Member
 
Join Date: Mar 2009
Quote:
Originally Posted by EUVL View Post
It is highly unstable with the output cascode, but quite easy to tame without.
If I remember well it must work at very high closed loop gain to be stable; moreover its output impedance is fairly high, and both its gain and freq response are strongly affected by loading - nice topology, but imho is pretty useless as a line stage unless some isolation of the gain node is provided by an output buffer, or the output cascodes are operated at a much higher quiescent current.

L.

Last edited by coluke; 28th August 2012 at 01:38 PM.
  Reply With Quote
Old 28th August 2012, 01:40 PM   #393
diyAudio Member
 
dirkwright's Avatar
 
Join Date: May 2011
Location: Virginia
Good Morning everyone. Another day, another schematic.

Hope everyone has a nice day.
Attached Images
File Type: jpg LME49710 diamond buffer.JPG (51.1 KB, 369 views)
  Reply With Quote
Old 28th August 2012, 02:26 PM   #394
gpapag is offline gpapag  Greece
diyAudio Member
 
gpapag's Avatar
 
Join Date: Nov 2002
Location: Athens-Greece
Sorry everyone. Post #391 was a glitch.

Quote:
Originally Posted by Kindhornman View Post
Not being an expert in electronic applications at all I am wondering why a simple thermocouple device could not be used for checking temperature rise? I used these all the time in manufacturing in autoclave applications and also in temperature control circuits. Perhaps the rise time is just to slow to tell you what you are looking for, but it would seem to be more specific than a simple IR handheld device. How could you narrow the specific heat source with one of those devices? You would pick up all the surrounding heat sources and I have never seen a handheld IR that had that narrow of a field of view?
The IR detectors were dropped in here specifically due to the IR image of a die they can provide. See here:
Quote:
Originally Posted by bcarso View Post
A cool thing (npi) would be to infrared-image the die and use that information. Certainly could be fast response

Brad

You can’t realistically have 2/10 d. Celcius resolution temp. mapping of a surface divided into 25x25um areas and following the dynamic variations of the temp at ~1kHz of all these “pixels” with other methods.
All this at a very reasonable price and minimum complexity for a manufacturer or researcher.
But still very expensive for the hobbyist. (That’s why the talking turned to the IR spot thermometer).



Quote:
Originally Posted by bcarso View Post
Yes, emissivity is clearly important to take into account (I'm reminded of the rule-of-thumb at aluminum extrusion houses: assume any piece of extrusion is very hot until proven otherwise, as the emissivity of bare Al is so low!).
This rule of thumb is the basic equation of thermography :

"Emissivity + Reflectivity = 1"

You want to maximize emissivity for to have correct IR temp. measurement. You have to minimize reflectivity to achieve this (shiny aluminum is very reflective).

A trick to achieve a spot with max. emissivity (~1 ) on any material, is to drill a blind hole on it with a depth > 2-3d.
IR temp. measurement into this spot is really very close to the true local temperature.

Quote:
Originally Posted by bcarso View Post
Another confounding variable for looking at a transistor with a short-wavelength detector: they are, to varying degrees, light emitters! In high-field regions especially. When Janesik et al. were developing CCDs for astronomy, there was a persistent excess of apparent dark current in the vicinity of the on-chip output amplifier. It turned out it was light emission from the MOS part, and iirc it got a good deal worse with cooling. They had to turn the amplifier off during long exposures.

Brad
Brad, it sounds to me (especially the counter-effect cooling had on the intensity of the emission) as if it was actually IR emission not adequately filtered in front of the CCD.
Did they verify the wavelength?



Quote:
Originally Posted by Wavebourn View Post
The whole opamp can be heated by precisionally controlled heater, to keep it's temperature stable.
Wavebourn
Raising the temp of the whole opamp, will worsen the dynamic temp. unbalance.
A specific subcomponent when variably heated due to it’s I*R will find it more difficult to get rid of it’s thermal energy through the hotter surrounding mass.
The delta temp. with another subcomponent (less I*R loaded) will thus increase.

George
__________________
["Second Law is a bitch." - SY] ["The Road To Heaven:Specify the performance & accept the design. The Road To Hell:Specify the design & accept the performance"-Bruno Putzeys]
  Reply With Quote
Old 28th August 2012, 02:35 PM   #395
diyAudio Member
 
dirkwright's Avatar
 
Join Date: May 2011
Location: Virginia
Quote:
Originally Posted by gpapag View Post
Wavebourn
Raising the temp of the whole opamp, will worsen the dynamic temp. unbalance.
A specific subcomponent when variably heated due to itís I*R will find it more difficult to get rid of itís thermal energy through the hotter surrounding mass.
The delta temp. with another subcomponent (less I*R loaded) will thus increase.

George
Well, there are always Peltier elements available... How cold do you want it?
  Reply With Quote
Old 28th August 2012, 02:45 PM   #396
diyAudio Member
 
Kindhornman's Avatar
 
Join Date: Aug 2012
Location: Los Angeles, California
gpapag,
Thank you for the explanations George. I actually follow this quite well except my math is more than rusty... I am glad you pointed out the fallacy of increasing the temperature to control temperature as this just seemed completely incorrect to me. By doing that you have just shot yourself in the foot by what appears to be decreasing your chances of dissipating the rising die temperature to a lower stable temperature. Wouldn't you be better off using a direct contact heat exchanger, a finned heat sink than using air as the direct transfer function? I only suggested the thermocouple as they are so cheap and attached directly to the die with a simple hand held meter you would have a very instantaneous reading of temperature rise. You could also drill a hole as you stated and insert the thermocouple into that hole. I had not seen a hand held with that resolution of depth of field. By drilling the hole you would appear to have two affects, one being the properties of the cavity depth to width and the second being a much thinner substrate that has a time lag in response to the actual temperature change internal to the device. Am I following what you are saying properly? Thanks again.

Steven
  Reply With Quote
Old 28th August 2012, 02:54 PM   #397
diyAudio Member
 
jneutron's Avatar
 
Join Date: Jan 2004
Location: away
Quote:
Originally Posted by gpapag View Post
Raising the temp of the whole opamp, will worsen the dynamic temp. unbalance.
A specific subcomponent when variably heated due to itís I*R will find it more difficult to get rid of itís thermal energy through the hotter surrounding mass.
The delta temp. with another subcomponent (less I*R loaded) will thus increase.

George
Actually, that is not the case.***

Interchip dynamics will not change unless the gradient to the outside world is different for each chip. If for example, one die's heat must go past the other to get out.

***the astericks are because the thermal conductivity of silicon is temperature dependent, the eq is:

Kt =286/(T -100) for T =300 to 600K

Cheers, jn
  Reply With Quote
Old 28th August 2012, 02:59 PM   #398
Account disabled at member's request
 
Join Date: Oct 2008
Quote:
Originally Posted by scott wurcer View Post
First off this is for fun and exchanging some ideas. I make no claim to inventing anything here it has all been somewhere before.
Classic op-amps input gm, Vas, and output buffer all end up relying on a linear input voltage to current stage (gm) feeding current to a high impedance node whose voltage is buffered to create a low output impedance. There are lots of embellishments around to increase the DC gain but the AC performance is usually limited by the linearity of the gm, the linearity of the gain node impedance (at AC frequencies the compensation capacitance plus any parasitics), and both the current needed to drive the output buffer and the voltage error across the output buffer. The last is problematic in that it (usually the crossover distortion) is differentiated before it can be expressed as an input current error, hence the rapid increase as frequency increases.
Before I go on I would like to say that a one size fits all op-amp is simply an unnecessary restriction. A simple enough signal path can give the user a couple of foolproof component substitutions that can optimize it for each application.
This being said I want to present a very ordinary signal path with a couple of additions to preserve simplicity and linearity. This circuit is not new but there are some differences to previous versions that I have seen (I donít see everything posted). The signal path on the non-inverting side has been arranged more like a current mirror and on the other a folded cascade. The cascodes on the gain stage have their base current recaptured to keep the collector impedance very high and eliminate non-linear capacitive currents. The bootstrapped triple darling presents a very light load to the gain node and can be independently customized to easily drive low impedances. Though as I have previously stated there is no reason for every op-amp in your signal path to drive something like 50 Ohms.
The circuit here is just an outline and the devices are place holders but even with these fairly ordinary devices Aol is > 100k and more importantly the gain at 20K is 75dB. J1 and J2ís currents are set for bias and to set up the lower half of the input stage to also recapture the base current of the cascodes. Originally I had planned on a completely symmetric circuit with PFETís. That is a fairly easy exercise for the reader.
I would like some suggestions for transistors, as far as I can tell I can get better than -100dB distortions at 20K, etc. with nothing more than this. As I said the degeneration resistors and comp cap would need to be changed for gain of one operation.
Hi Scott
I agree with everything you are saying here.
I think your circuit is the most interesting here so far.
I'll put it in to my simulator and have a closer look at it.
BTW: is this the model you are using for the BF862?

* BF862 SPICE MODEL MARCH 2007 NXP SEMICONDUCTORS
* ENVELOPE SOT23
* JBF862: 1, Drain, 2,Gate, 3,Source

Ld 1 4 L= 1.1nH
Ls 3 6 L= 1.25nH
Lg 2 5 L= 0.78nH
Rg 5 7 R= 0.535 Ohm
Cds 1 3 C= 0.0001pF
Cgs 2 3 C= 1.05pF
Cgd 1 2 C= 0.201pF
Co 4 6 C= 0.35092pF

JBF862 model parameters:

.model JBF862 NJF(Beta=47.800E-3 Betatce=-.5 Rd=.8 Rs=7.5000 Lambda=37.300E-3 Vto=-.57093
+ Vtotc=-2.0000E-3 Is=424.60E-12 Isr=2.995p N=1 Nr=2 Xti=3 Alpha=-1.0000E-3
+ Vk=59.97 Cgd=7.4002E-12 M=.6015 Pb=.5 Fc=.5 Cgs=8.2890E-12 Kf=87.5E-18
+ Af=1)

ENDS BF862

Cheers
Stein
  Reply With Quote
Old 28th August 2012, 03:05 PM   #399
bcarso is offline bcarso  United States
diyAudio Member
 
Join Date: May 2005
Location: Canoga Park, California
Quote:
Originally Posted by janneman View Post
True, but you get back higher gain in the first stage and ultimately lower, or at least not higher, input referred noise.

jan
Yes, in fact I believe that was one of my and Vogel's points in our LTEs.
  Reply With Quote
Old 28th August 2012, 03:09 PM   #400
bcarso is offline bcarso  United States
diyAudio Member
 
Join Date: May 2005
Location: Canoga Park, California
Quote:
Originally Posted by coluke View Post
If I remember well it must work at very high closed loop gain to be stable; moreover its output impedance is fairly high, and both its gain and freq response are strongly affected by loading - nice topology, but imho is pretty useless as a line stage unless some isolation of the gain node is provided by an output buffer, or the output cascodes are operated at a much higher quiescent current.

L.
Agree. I wouldn't be so enthusiastic about losing the cascodes, as besides higher unloaded gain they reduce the effect of thermals in the preceding stages. But a buffer is a good idea, then the compensation required won't be so load-dependent. Of course this is basic stuff.
  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
Discrete OPAMP audio-gd Vendor's Bazaar 27 20th September 2012 04:02 PM
discrete opamp help blackpowderaudio Parts 0 16th December 2009 03:46 PM
THAT transistor headphone amp (250ma discrete opamp) design sanity check. Russ White Headphone Systems 19 13th December 2007 12:52 PM


New To Site? Need Help?

All times are GMT. The time now is 03:03 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