Go Back   Home > Forums > >
Home Forums Rules Articles diyAudio Store Blogs Gallery Wiki Register Donations FAQ Calendar Search Today's Posts Mark Forums Read

Solid State Talk all about solid state amplification.

Williams Ring of Two Current Source
Williams Ring of Two Current Source
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 13th October 2019, 05:06 PM   #1
chalky is offline chalky
diyAudio Member
 
Join Date: Dec 2003
Location: North of the border
Default Williams Ring of Two Current Source

This current source was first described by Peter Williams in the July 1967 issue of Wireless World and is shown in the first schematic. The most commonly used variant replaces the zener diodes with leds as shown in the second schematic. Williams himself noted that it is theoretically possible that the circuit may not self start, indeed I suspect that most simulations will not self start without a "kicker" resistor. However, in real life, the circuit reliably self starts regardless of ( within sensible limits ) voltage, current, temperature, transistor matching, deliberate transistor mismatching, and the presence or absence of light. I have verified that the circuit self starts at -20C, so if the circuit is kicked into life by leakage current through the transistors it must be very small. In addition to the tests I have used this circuit as a two terminal current source in a myriad of amplifier projects over the last 20 years with no problem.
My questions then are:- (1) Has anyone observed this circuit, or a variant of it, failing to start and (2) Has anyone got a convincing explanation, other than miniscule leakage currents, as to why the Williams ring of two current source always starts so reliably.
Finally, in the third schematic the leds have been replaced with transistors and this circuit too, self starts reliably. I find that two terminal current sources make pcb layout easier and this variant has very good PSRR, potentially very low noise, and fairly low temperature sensitivity. I'll switch to using this variant rather than the led version.
Any comments, particularly as regards shortcomings of this circuit are welcome.
Attached Images
File Type: png WilliansRingOfTwo.png (13.9 KB, 301 views)
File Type: png WilliamsCcsLed.png (5.3 KB, 303 views)
File Type: png WilliamsCcsTr.png (6.8 KB, 299 views)
__________________
If it ain't broke don't fix it
  Reply With Quote
Old 13th October 2019, 10:01 PM   #2
Elvee is offline Elvee  Belgium
diyAudio Member
 
Elvee's Avatar
 
Join Date: Sep 2006
I don't think I have ever observed this exact circuit failing to start (but I am not completely certain: I know that some schemes relying on positive feedback sometimes failed to start naturally, but I do not remember which ones).

Anyway, I generally prefer to be safe, and I always include a starter to ensure peace of mind.
The sim always starts without any problem (but this is not a proof for anything)

This could also interest you:
Improved 2W current sources (II)
__________________
. .Circlophone your life !!!! . . ♫♪ My little cheap Circlophoneİ ♫♪
Electrify your daily life!!! ....⚡⚡ ELEKTRIA ⚡⚡
  Reply With Quote
Old 14th October 2019, 12:37 AM   #3
PaulBysouth is offline PaulBysouth  Australia
diyAudio Member
 
Join Date: Feb 2008
Location: Melbourne
Default 2 terminal current sources

Chalky,
To ensure that the `Ring of 2' starts, you can put a 1megohm resistor between the two bases. This `leakage' current will ensure it starts.

Are you aware of the NSIC2020jbt3, NSIC2030jbt3 and NSIC2050jbt3 current sources? These are 20mA, 30mA and 50mA `ring of two' current sources - SMB package, 120v and cheap. Digikey has thousands of them. I frequently use the NSIC2020's.

For less that 20mA, I normally use a DN2530 (sot89) depletion mode mosfet, plus two resistors. Use about 500ohms for the gate resistor, and the other resistor sets the current. There is also a DN2540 in a To220 for higher currents and voltage. Unlike jfets, the DN2530's have a reasonably consistent Vgs threshold (~ 2.2volts). For better performance you can cascode one with a second DN2530.

Regards,
Paul Bysouth
  Reply With Quote
Old 14th October 2019, 03:00 AM   #4
Mark Tillotson is offline Mark Tillotson
diyAudio Member
 
Join Date: May 2018
Location: Cambridge UK
The self-starting depends on the comparative sizes of the transistor leakage to zener/LED leakage - if the zener/LED leakage is greater it will not fire up. Fortunately the transistor leakage is amplified by its current gain so its highly likely to be the larger.

This argument doesn't apply to variant 3 I think...


I wonder if employing the extra transistor for cascoding is a better use of it, performance-wise? And it allows more freedom in choosing best transistor for each role.
  Reply With Quote
Old 14th October 2019, 10:35 AM   #5
DF96 is offline DF96  England
diyAudio Member
 
Join Date: May 2007
In the (theoretical) absence of leakage is it possible that thermal noise would eventually start it?
  Reply With Quote
Old 14th October 2019, 12:26 PM   #6
anti is offline anti  Slovakia
diyAudio Member
 
Join Date: Jun 2008
Also there is always some B-C capacitance and miller effect ... which may or may not contribute to self-starting scenario.
  Reply With Quote
Old 14th October 2019, 05:10 PM   #7
jackinnj is offline jackinnj  United States
diyAudio Member
 
jackinnj's Avatar
 
Join Date: Apr 2002
Location: Llanddewi Brefi, NJ
Williams Ring of Two Current Source
5mA CCS assuming 2N5087/2N5210 vs simple two transistor 5mA CCS. The Williams is about 36dB better. Eyeballing Walt's "Sources 101 Part 2" AudioXpress article, would appear to be 20dB better than DMOS.
Attached Images
File Type: png Williams_Z.png (16.5 KB, 122 views)
  Reply With Quote
Old 14th October 2019, 05:21 PM   #8
Mark Johnson is online now Mark Johnson  United States
diyAudio Member
 
Mark Johnson's Avatar
 
Join Date: May 2011
Location: Silicon Valley
Williams Ring of Two Current Source
LM334 + J105 cascode requires 5 parts total and always starts up
  • Two 3-legged TO-92s
  • Two resistors
  • One silicon diode

LM334 makes quite a nice 5.5 mA current source, works down to 2.0-2.2 volts

_
Attached Images
File Type: png zowie.png (28.1 KB, 123 views)
  Reply With Quote
Old 14th October 2019, 09:27 PM   #9
chalky is offline chalky
diyAudio Member
 
Join Date: Dec 2003
Location: North of the border
Thanks for all the feedback and info guys, lots to think about here. I don't really think that startup resistors are necessary with the Williams current source but the emitter to emitter resistors suggested by John Rudge as an improvement ( in a subsequent Letter to the Editor of Wireless World ) and implemented by Elvee look to be a worthwhile addition. How did you calculate the value of these resistors Elvee? At the moment I'm trying to find the best solution for a 10mA 200V constant current source as part of a new solid state amplifier design, and the DN2540 suggested by Paul Bysouth looks a good possibility, as well as a higher power version of the Williams current source.
__________________
If it ain't broke don't fix it
  Reply With Quote
Old 14th October 2019, 09:45 PM   #10
jackinnj is offline jackinnj  United States
diyAudio Member
 
jackinnj's Avatar
 
Join Date: Apr 2002
Location: Llanddewi Brefi, NJ
Williams Ring of Two Current Source
Quote:
Originally Posted by chalky View Post
At the moment I'm trying to find the best solution for a 10mA 200V constant current source as part of a new solid state amplifier design, and the DN2540 suggested by Paul Bysouth looks a good possibility, as well as a higher power version of the Williams current source.
Two are better than one! R13 and R18 are the "set" resistors -- DMOS can vary quite a bit so they will have to be adjusted to fit -- this one is set for 5mA. ymmv

Once the PSRR gets better than +120dB stray capacitance, dust, grease etc. can affect measurements.
Attached Images
File Type: png DMOS_CCS_2.png (16.5 KB, 53 views)
  Reply With Quote

Reply


Williams Ring of Two Current SourceHide 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

Forum Jump

Similar Threads
Thread Thread Starter Forum Replies Last Post
Source for attractive metal mesh grille w/dress ring? ro9397 Multi-Way 1 21st March 2019 06:52 PM
faraday ring shape ?(shorting ring) guitar89 Construction Tips 2 9th September 2011 10:38 AM
Aleph-type current source, but source follower this time tschrama Pass Labs 4 29th July 2005 01:55 PM
Source follower electret mike -> use with current source? capslock Solid State 1 28th February 2003 01:54 PM
about the shorted ring (flux stabilizer ring) haggy Multi-Way 3 4th January 2002 03:04 AM


New To Site? Need Help?

All times are GMT. The time now is 01:11 AM.


Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2019 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 14.29%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2019 DragonByte Technologies Ltd.
Copyright ©1999-2019 diyAudio
Wiki