Discrete Opamp Open Design - Page 198 - 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 3rd November 2012, 04:12 PM   #1971
Jay is offline Jay  Indonesia
diyAudio Member
 
Join Date: Feb 2003
Location: Jakarta
Quote:
Originally Posted by scott wurcer View Post
Open the LTSPICE model files you will notice that the 1N914 and 1N4148 have exactly (to 3 digits) the same values for most parameters which is obviously nonsense.
Hahaha I learned this issue the hard way
  Reply With Quote
Old 3rd November 2012, 05:42 PM   #1972
bcarso is offline bcarso  United States
diyAudio Member
 
Join Date: May 2005
Location: Canoga Park, California
What IS sometimes significant is that we might assume that 4148-style diodes have precisely the same I-V characteristics as diode-connected transistors. They do not, but it many applications the difference is of little importance. But the I-V curve for operation of the diode in the injection region, i.e., forward-biased significantly, will differ from that of a diode-connected transistor. This is discussed in many references* and for empirical data on differences among diodes see the Appendix E table in Pease's book Troubleshooting Analog Circuits, pp. 196-198. There is also a chapter 6, Understanding Diodes and Their Problems, which at least mentions "high-level injection" without showing the modification of the diode equation that it entails. Sadly, Pease omits the diode-connected transistor, and as well doesn't mention most of the manufacturers of the DUTs.

There was an Ideas for Design or some such years ago in which the author linearized a transistor gain stage by use of a diode. When I saw it I thought Ah, why not a diode-connected transistor? But it didn't work --- without declaring such, the circuit worked by virtue of the I-V curve differences, not their similarity

Brad

*including everyone's favorite compendium by Horowitz and Hill, except I can't locate it in the second edition right now.

See 4.4.4 and the rest of this reference where the modifying term eta, high injection, and the effects of series resistance are discussed in greater depth: The p-n diode current . The latter says you don't see much of the high-injection region separate from the series-resistance-dominated region, which is an assertion to be questioned Note that the ideal region is shown as below 550mV at room temperature for a typical small-signal diode, which is rarely where we operate them.
  Reply With Quote
Old 3rd November 2012, 08:14 PM   #1973
diyAudio Member
 
scott wurcer's Avatar
 
Join Date: Jan 2004
Location: cambridge ma
A quick observation. Most folks have a little handheld VOM with a diode check function which works by forcing a current and measuring the drop. Simply test a 100 Ohm resistor and you know the test current (V/R) mine uses 1mA. You can use this to guestimate the saturation current of your batch of diodes/transistors from the diode equation and fine tune your sims.
I just grabbed 3 1N4148's from the junk bin and at 1mA they measured .653, .654, and .656 volts while my batch of 2N4401's measures about .655 volts connected as diodes.So clearly the stock models are off and this explains why Pete B's JE990 sim was so far off on bias, in fact the diodes are pretty close to the transistors.

EDIT - be careful not to heat the devices by holding them
__________________
"The question of who is right and who is wrong has seemed to me always too small to be worth a moment's thought, while the question of what is right and what is wrong has seemed all-important."

Last edited by scott wurcer; 3rd November 2012 at 08:22 PM.
  Reply With Quote
Old 3rd November 2012, 09:07 PM   #1974
diyAudio Member
 
Join Date: Nov 2011
Location: Cooktown, Oz
Default Twitchy SW-OPA?

Quote:
Originally Posted by scott wurcer View Post
I just grabbed 3 1N4148's from the junk bin and at 1mA they measured .653, .654, and .656 volts while my batch of 2N4401's measures about .655 volts connected as diodes.So clearly the stock models are off and this explains why Pete B's JE990 sim was so far off on bias, in fact the diodes are pretty close to the transistors.
I've used 1n4148 in place of diode connected transistors for simple current mirrors in a production amp so have a lot of data on 2 cases.

One was at 0.8mA with BC556c and the other at 4mA with 2sc2362. In both cases, 1n4148 had significantly bigger drop. In the 4mA case, requiring a 180 resistor instead of 200 on the 2sc2362 emitter; ie 80mV

This bears on my investigation into twitchy SW-OPA. ricardo.txt which I uploaded in #1949 has BJT models from several gurus. I have 2 sets of c2240/a970 and also 4401/3.

Scott's c2240 has some 100mV more Vbe than his a970 at the same current.

Cordell's 4401/3 have closer Vbe than fas42's but for some reason results in a lot more THD in some circuits. I've looked at *.log which LTspice produces on .opt analysis cos this gives stuff which I pretend to unnerstan. The only difference which stands out is Ro whatever that is.

My own Jurassic experience with 2240/970 is that the models with closer Vbe are closer to real life.

I've also increased RB to more believable values in preparation for noise analysis but that shouldn't give 100mV errors. I'll take them out & see if that makes a diff.

Guru Wurcer, I'd still like to know if your instruction to adjust for 8mA applies to the output devices or just to the 4R7 resistors.
  Reply With Quote
Old 4th November 2012, 04:11 AM   #1975
RNMarsh is offline RNMarsh  United States
diyAudio Member
 
RNMarsh's Avatar
 
Join Date: May 2012
Location: 2457 Cascade Trail; Cool, CA. 95614
Checking in --- good to see expansions on real world build issues discussed in the Vbe and diode drops/I-V et al. One of the reasons I measure so often rather than relying soley on models ---- Sophisticated programs which use known processes and known results of those processes are not in my domain. I measure at the same voltage/current I will be using the part.... and again for same P/N but different brand... no assuumptions. If you don't, the results can be all over the map. We want repeatability for the DIY'er to make and enjoy the fruits of this labor of love.

Last edited by RNMarsh; 4th November 2012 at 04:13 AM.
  Reply With Quote
Old 4th November 2012, 03:30 PM   #1976
diyAudio Member
 
scott wurcer's Avatar
 
Join Date: Jan 2004
Location: cambridge ma
Quote:
Originally Posted by bcarso View Post
Note that the ideal region is shown as below 550mV at room temperature for a typical small-signal diode, which is rarely where we operate them.
Easy experiment take two that are close and parallel them the transistors show the expected 18mV drop but the diodes don't. Again you can probably predict close to what you want with just a VOM.
__________________
"The question of who is right and who is wrong has seemed to me always too small to be worth a moment's thought, while the question of what is right and what is wrong has seemed all-important."
  Reply With Quote
Old 4th November 2012, 04:40 PM   #1977
bcarso is offline bcarso  United States
diyAudio Member
 
Join Date: May 2005
Location: Canoga Park, California
Quote:
Originally Posted by scott wurcer View Post
Easy experiment take two that are close and parallel them the transistors show the expected 18mV drop but the diodes don't. Again you can probably predict close to what you want with just a VOM.
Somewhere back in ancient times I posted data from a bunch of 4148s, not sure if it was in here. They were off of the same taped batch, not sure of the manufacturer. I taped them to a paper and went down the row touching them only with the Fluke 8062A DVM with the 1mA excitation. The distribution of forward voltages was pretty tight, although a hint of bimodality.
  Reply With Quote
Old 4th November 2012, 07:25 PM   #1978
diyAudio Member
 
Join Date: Nov 2011
Location: Cooktown, Oz
Quote:
Originally Posted by kgrlee View Post
I've looked at *.log which LTspice produces on .opt analysis cos this gives stuff which I pretend to unnerstan. The only difference which stands out is Ro whatever that is.
Anyone know what Ro is?

The 3 online SPICE references don't mention anything at all so this might be only LTspice.
  Reply With Quote
Old 4th November 2012, 07:36 PM   #1979
diyAudio Member
 
scott wurcer's Avatar
 
Join Date: Jan 2004
Location: cambridge ma
Quote:
Originally Posted by bcarso View Post
Somewhere back in ancient times I posted data from a bunch of 4148s, not sure if it was in here. They were off of the same taped batch, not sure of the manufacturer. I taped them to a paper and went down the row touching them only with the Fluke 8062A DVM with the 1mA excitation. The distribution of forward voltages was pretty tight, although a hint of bimodality.
A quick experiment measuring 1, 2, and 3 matched diodes in parallel gives about 35 Ohms series R so they behave "ideally" only up to 10's of uA.
__________________
"The question of who is right and who is wrong has seemed to me always too small to be worth a moment's thought, while the question of what is right and what is wrong has seemed all-important."
  Reply With Quote
Old 4th November 2012, 08:29 PM   #1980
diyAudio Member
 
Join Date: Mar 2002
Quote:
Originally Posted by kgrlee View Post
Anyone know what Ro is?

The 3 online SPICE references don't mention anything at all so this might be only LTspice.
http://people.seas.harvard.edu/~jone...jt_models.html
Attached Images
File Type: jpg t_model_2.jpg (9.6 KB, 253 views)

Last edited by hitsware; 4th November 2012 at 08:32 PM. Reason: add link to picture
  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 05:02 PM
discrete opamp help blackpowderaudio Parts 0 16th December 2009 04:46 PM
THAT transistor headphone amp (250ma discrete opamp) design sanity check. Russ White Headphone Systems 19 13th December 2007 01:52 PM


New To Site? Need Help?

All times are GMT. The time now is 03:30 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