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

The simplistic Salas low voltage shunt regulator
The simplistic Salas low voltage shunt regulator
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 April 2015, 07:00 PM   #6831
Salas is offline Salas  Greece
diyAudio Chief Moderator
 
Salas's Avatar
 
Join Date: Oct 2002
Location: Athens-Greece
The simplistic Salas low voltage shunt regulator
This is a -0.47V pinch off K170BL example I just traced out. Although even straighter later, it looks usable above Vbe=Vds. Its Ohmic region already gives way to the saturation region when under Vbe voltage only in other words.
Attached Images
File Type: png 2015-04-26_202953.png (8.0 KB, 190 views)
File Type: png 2015-04-26_204426.png (33.1 KB, 191 views)
  Reply With Quote
Old 26th April 2015, 07:05 PM   #6832
Salas is offline Salas  Greece
diyAudio Chief Moderator
 
Salas's Avatar
 
Join Date: Oct 2002
Location: Athens-Greece
The simplistic Salas low voltage shunt regulator
Quote:
Originally Posted by bcarso View Post
That seems sound. These short-channel devices are handy for many things but do have a rather high output conductance for current generator use.

The circuit works, so I don't want to get too far into this, and bipolars do have advantages in many places compared to FETs, but I might try replacing this one with an enhancement mode DMOS, as the gate-source voltage will be a few volts and allow the JFET to perform better. Also the impedance at that node will be much higher than a forward-biased base-emitter junction. One would not use a very low threshold voltage DMOS part, but something along the lines of the 2N7000.

Brad
Noise issues concerned? When looking to keep the noise spec the same as before I mean.
  Reply With Quote
Old 26th April 2015, 07:19 PM   #6833
bcarso is offline bcarso  United States
diyAudio Member RIP
 
Join Date: May 2005
Location: Canoga Park, California
Quote:
Originally Posted by Salas View Post
Noise issues concerned? When looking to keep the noise spec the same as before I mean.
Yes, that certainly is a consideration. Unfortunately noise is rarely specified for MOS, especially low-frequency excess noise. I'm amazed by how good some of the DMOS op amps are, but they still have a significant rise at lower frequencies.
  Reply With Quote
Old 26th April 2015, 08:17 PM   #6834
Salas is offline Salas  Greece
diyAudio Chief Moderator
 
Salas's Avatar
 
Join Date: Oct 2002
Location: Athens-Greece
The simplistic Salas low voltage shunt regulator
Quote:
Originally Posted by AndrewT View Post
Try the experiment.

Connect G to S as you would for measuring Idss.
Apply a 10V Vds. Measure the current.
reduce Vds in steps and remeasure the current.
Keep reducing Vds to ~ Vp.
Now plot the Ic vs Vds.

You will find that Ic holds very close to CCS when Vds is >5Vds
From 5Vds down to 2*Vp, the Ic falls as CCS action gets worse.
From 2*Vp down to 1*Vp, the Ic falls so badly that one can no longer call the jFET a CCS. It behaves as a low value resistor.
I tried the experiment for that same -0.47V Vp K170BL. A jumper between G & S. A variable PSU for VDS. The thick red lead in the picture goes to a Fluke 87V in series mA DC mode for better readings than the PSU's at this difficult for its resolution region.

VDS / Id

10V 7.20mA
4.0V 6.93mA
1.0V 6.39mA (at about 2*Vp)
0.7V 6.17mA (at about Vbe)
0.5V 5.83mA (at about 1*Vp)
0.3V 4.84mA

So 9.6% current fall between 2*Vp and 1*Vp at Vgs=0
Attached Images
File Type: jpg IMG_20150426_220340.jpg (282.6 KB, 222 views)
  Reply With Quote
Old 26th April 2015, 09:27 PM   #6835
bcarso is offline bcarso  United States
diyAudio Member RIP
 
Join Date: May 2005
Location: Canoga Park, California
Quote:
Originally Posted by bcarso View Post
Yes, that certainly is a consideration. Unfortunately noise is rarely specified for MOS, especially low-frequency excess noise. I'm amazed by how good some of the DMOS op amps are, but they still have a significant rise at lower frequencies.
I'm curious as to the function of R10---was it needed for stability?

In any event, one concern I'd have is when the output might be inadvertently shorted to common, as C1 will break down the Q7 base-emitter junction, although this will be to some extent reduced by the action of Q9. A reverse-biased diode from the base to common would protect against beta degradation---of course one could simply never short the output.

Back to noise: the circuit response to Q9 and R12 noise is that of a first-order lowpass rolloff commencing at about 0.8Hz, with a low frequency gain of about 37dB (set by R11 and C1). For a ~1nV/sq rt Hz JFET and R12 equal to 523 ohms, this may amount to about 250nV rms at the reg output (very good indeed). The larger contribution will be from Q7 and R10, which have a noise gain of about unity out to fairly high frequencies. At the operating point of Q7 of a few mA Ic, the Q7 voltage noise should be dominated by rbb'---but the 1k of R10 will swamp this, hence e sub n of Q7 can practically be ignored. In a 20kHz noise bandwidth the output noise due to R10 should be about 141 times R10 noise, or about 575nV rms at 300K. Shot and excess noise in Q7 Ib will probably contribute about as much (in a 20kHz noise bandwidth for pure shot noise and an Ib of 9.4uA, about 266nV rms, but base current noise is usually a good deal higher than just that due to shot noise). So the transistor and R10 will dominate the noise, most likely. It could be reduced a good deal by eliminating R10.

As far as DMOS in that spot goes, for my 2N7000 model the output impedance of the circuit is lower at low frequencies but doesn't stay as low as the bipolar version, starting to climb above about 1kHz. As far as noise, I don't know how current parts are doing, so it is hard to predict, although it is probably pretty decent above a kilohertz.
  Reply With Quote
Old 26th April 2015, 09:41 PM   #6836
Salas is offline Salas  Greece
diyAudio Chief Moderator
 
Salas's Avatar
 
Join Date: Oct 2002
Location: Athens-Greece
The simplistic Salas low voltage shunt regulator
-On R10

Yes for stability

For very low noise there was one 1.2 with an RC comp and big C1 filter
Attached Images
File Type: png 2015-04-26_233440.png (6.3 KB, 176 views)
  Reply With Quote
Old 26th April 2015, 09:58 PM   #6837
Salas is offline Salas  Greece
diyAudio Chief Moderator
 
Salas's Avatar
 
Join Date: Oct 2002
Location: Athens-Greece
The simplistic Salas low voltage shunt regulator
Quote:
Originally Posted by RCruz View Post
Ok, I will move VR12 up and also put a led in series with it as you use on the folded.

As for Q9 Vp < 50% Q7 Vbe maybe I could raise this Vbe up by using a diode as emitter for Q7 ?
You probably can but it has a resistance delta and some own noise. I would nest a low side Idss K117GR instead.
  Reply With Quote
Old 26th April 2015, 10:04 PM   #6838
bcarso is offline bcarso  United States
diyAudio Member RIP
 
Join Date: May 2005
Location: Canoga Park, California
Winfield Hill's measurements of 2N7000s a while back sound pretty pessimistic: see Noise performance of small package mosfets | Electronics Forums
The disparity between that and what is being achieved in op amps is striking!
  Reply With Quote
Old 26th April 2015, 10:14 PM   #6839
Salas is offline Salas  Greece
diyAudio Chief Moderator
 
Salas's Avatar
 
Join Date: Oct 2002
Location: Athens-Greece
The simplistic Salas low voltage shunt regulator
Very bad indeed. Wanna see the third party measured noise floor of a V1.1 reg (the one that people make on those black GB boards)?
  Reply With Quote
Old 26th April 2015, 10:21 PM   #6840
Salas is offline Salas  Greece
diyAudio Chief Moderator
 
Salas's Avatar
 
Join Date: Oct 2002
Location: Athens-Greece
The simplistic Salas low voltage shunt regulator
Here it is when using a moderate 220uF across Rref filter. Can take 1000uF with no problem. An Audio Precision analyzer station was used. The orange dotted line is the V1.1

http://linearaudionet.solide-ict.nl/...%20JW%20F7.pdf
  Reply With Quote

Reply


The simplistic Salas low voltage shunt regulatorHide 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


New To Site? Need Help?

All times are GMT. The time now is 03:45 PM.


Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 15.00%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Copyright ©1999-2018 diyAudio
Wiki