The Well Regulated Power Supply - Page 3 - diyAudio
Go Back   Home > Forums > Amplifiers > Power Supplies

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 31st May 2012, 03:52 PM   #21
diyAudio Member
 
andrea_mori's Avatar
 
Join Date: Jan 2005
Location: Italy
Revised schematics: some values changed, ballast resistor R17 added.
Attached Images
File Type: jpg TWRPS.JPG (99.7 KB, 941 views)
  Reply With Quote
Old 31st May 2012, 06:17 PM   #22
diyAudio Member
 
jan.didden's Avatar
 
Join Date: May 2002
Location: Great City of Turnhout, Belgium
Blog Entries: 7
If you use a series zener or even just an LED on the output of the opamp you can supply the opamp from the final regulated voltage. That improves all regulator performance areas.

jan didden
__________________
If you don't change your beliefs, your life will be like this forever. Is that good news? - W. S. Maugham
Check out Linear Audio!

Last edited by jan.didden; 31st May 2012 at 06:21 PM.
  Reply With Quote
Old 31st May 2012, 08:30 PM   #23
AndrewT is offline AndrewT  Scotland
diyAudio Member
 
Join Date: Jul 2004
Location: Scottish Borders
You will have to "sort out" all the traces that connect to Ground and similarly all the traces that connect to +ve.
__________________
regards Andrew T.
  Reply With Quote
Old 31st May 2012, 11:13 PM   #24
diyAudio Member
 
andrea_mori's Avatar
 
Join Date: Jan 2005
Location: Italy
Quote:
Originally Posted by janneman View Post
If you use a series zener or even just an LED on the output of the opamp you can supply the opamp from the final regulated voltage. That improves all regulator performance areas.

jan didden
I have already simulated the configuration that you suggest, like Jung super regulator, but I think that cannot work because I need a voltage on the base of D44H11 greater than emitter voltage, so if I supply LT1028 from the final regulated voltage I cannot reach a positive vbe. At least I should replace NPN transistor with a PNP type.
Jung super regulator operate in a different mode, pass transistor is biased from a constat current source and the op amp modulate this current with error signal.
BTW the decoupling network R5-T2-C2 improves significatively PSSR of the op amp, more than 30 db in my simulation, not much different, I believe, than supply op amp directly from the final regulated voltage.

Andrea
  Reply With Quote
Old 31st May 2012, 11:26 PM   #25
diyAudio Member
 
andrea_mori's Avatar
 
Join Date: Jan 2005
Location: Italy
Quote:
Originally Posted by AndrewT View Post
You will have to "sort out" all the traces that connect to Ground and similarly all the traces that connect to +ve.
Of course, when I'll design PCB I'll take maximum care to create the right path for ground and ve.
Current sensing resistors R14-R16, per example, need to connect with traces as short as possible to the output of the regulator, that's the reason I'll use a 4 terminal resistor.

Andrea
  Reply With Quote
Old 3rd June 2012, 05:12 PM   #26
diyAudio Member
 
andrea_mori's Avatar
 
Join Date: Jan 2005
Location: Italy
Default Ready to prototype

Follows the final schematics of two versions of regulator: HG version is a 6Vdc 700mA with hi-grade components, SG is a 5Vdc 700 mA with standard components.
I attach some analisys of HG simulated circuit.
I'm going to order components to build two prototypes.

Andrea
Attached Images
File Type: jpg TWRPS-HG.JPG (98.3 KB, 845 views)
File Type: jpg TWRPS_SG.JPG (96.4 KB, 796 views)
File Type: jpg Osc.jpg (56.7 KB, 768 views)
File Type: jpg AC_Transfer_Char.jpg (58.2 KB, 502 views)
File Type: jpg Output_Noise.jpg (49.9 KB, 131 views)
File Type: jpg Signal_to_Noise.jpg (51.4 KB, 121 views)
File Type: jpg Transient_Pulse_6V_1ms.jpg (43.8 KB, 116 views)
File Type: jpg Transient_Square_10mV_120Hz.jpg (40.4 KB, 130 views)
  Reply With Quote
Old 3rd June 2012, 06:15 PM   #27
diyAudio Member
 
jan.didden's Avatar
 
Join Date: May 2002
Location: Great City of Turnhout, Belgium
Blog Entries: 7
Quote:
Originally Posted by andrea_mori View Post
I think that cannot work because I need a voltage on the base of D44H11 greater than emitter voltage, so if I supply LT1028 from the final regulated voltage I cannot reach a positive vbe. At least I should replace NPN transistor with a PNP type.
Jung super regulator operate in a different mode, pass transistor is biased from a constat current source and the op amp modulate this current with error signal.
Andrea
You are right, I missed that. It only works when you use a PNP pass.
'
jan
__________________
If you don't change your beliefs, your life will be like this forever. Is that good news? - W. S. Maugham
Check out Linear Audio!
  Reply With Quote
Old 3rd June 2012, 08:46 PM   #28
gootee is offline gootee  United States
diyAudio Member
 
Join Date: Nov 2006
Location: Indiana
Blog Entries: 1
This might be slightly off-topic, in a way, but is probably worth thinking about.

What about simulating with digital loads (e.g. pulsed current draw), and including power and ground rails' impedances, and looking at the voltage across the load? Then everything will depend on the decoupling capacitors at the load.

Typically, it seems that people expend much effort on a power supply but not enough effort on the effects of the impedance of its connections to the load, and the decoupling capacitor networks at the load.

What matters most is what happens at the load, not at the power supply output.

Even with a truly-perfect power supply, any load drawing transient currents through inductive power rail conductors (or planes) will produce voltage disturbances on power/gnd rails (or planes), at the load, with amplitudes of voltage disturbances dependent almost solely on decoupling capacitors and load's transient-current demand's amplitudes and timing.

Decoupling caps at each load are like small point-of load power supplies, for transient current demands. (With a known maximum change in current that might be needed, di, and a known minimum time during which that might have to occur, dt, and a desired maximum voltage disturbance on the power rail (at the load) due to the transient current, dv, which we get to choose, the needed minimum decoupling capacitance can be estimated with C = di dt / dv. There is more to consider, and different ways to look at it, but that, or double that to be safe, is better than just guessing. [You do also have to calculate an estimate of the inductance that can be tolerated in the connections of the caps to the load, to determine their maximum acceptable length and whether or not multiple parallel caps need to be used instead. Simplified way would be L = dv dt / di.])

What is the actual goal? Isn't it for the power pins of the active devices to receive (or give) the current they demand, exactly as needed (amplitude vs time), while the voltage is maintained at a constant value? (Or is the constant voltage value even always necessary, as long as the device pin's current's amplitude vs time is exactly correct?)

I guess low-noise steady-state stuff is relatively easy and well-known. But transient current demands can ruin your best design if you haven't paid very close attention to the rail impedance effects and used optimal decoupling networks very near the points of load. It's still just about as important when using full power and ground planes.

Without power and ground planes: Best way would probably be to have zero-length power rails and a fantastic PSU. Second-best might be to use multiple parallel copies of each voltage and ground rail (lowering overall impedance by paralleling paths, i.e. by having no mutual self-inductance), each path with its own caps at both load and PSU ends and paths all connecting only at a single point at each end. Third best is probably the usual way but to also carefully design the decoupling capacitor networks. You can simulate all of those. Probably nothing else is really worth trying, if performance is critical.

Even when using power and ground planes, with many dispersed decoupling caps around chips, you would probably also want to disperse the power and ground connections to the planes (which is sometimes forgotten), with fully-parallel connections all the way back to the actual main supply and ground points. You could probably simulate a best-case scenario for that, by assuming all decoupling cap currents take separate paths (so no mutual inductance makes inductance reduce fully due to parallel paths), and calculating the overall impedance of the parallel paths.

Sorry to have blathered-on about all of that, for so long.

Cheers,

Tom
  Reply With Quote
Old 4th June 2012, 01:53 AM   #29
gootee is offline gootee  United States
diyAudio Member
 
Join Date: Nov 2006
Location: Indiana
Blog Entries: 1
In your simulations, you should probably also try including estimates of the parasitics, of both the components and of the conductors. It will give a much-more-realistic simulation (although slower-running), and could also help you to find out which, if any, conductors' lengths might be critical to keep short.

For a first approximation, you can usually use .001 Ohm of series resistance per inch of PCB trace and 30 nH of series self-inductance per inch. (Obviously you might have to guess at, or estimate, the PCB trace lengths, for now.)

If you don't know the parasitic series inductance of a capacitor, then at least include 30 nH per inch of lead spacing. And many resistors have roughly 0.2 pF or so of parallel parasitic capacitance.

The parasitic series resistance (ESR, or Equivalent Series Resistance) of electrolytic capacitors changes quite-significantly, with frequency. If ESR is not given at the frequency you need for simulation, you can usually convert it, using the assumption that the dissipation factor (tan(delta)) varies by much less, versus frequency, usually less than 100% overall. They are related by

tan(δ) = 2 ∙ π ∙ f ∙ C ∙ ESR

which can be used to get tan(δ) if you know only the ESR at a particular frequency.

Then you can solve for ESR and re-calculate the ESR for different frequencies, assuming that tan(δ) doesn't change by much:

ESR = tan(δ) / (2 ∙ π ∙ f ∙ C)

There is usually also a leakage current spec, such as "I = .01 C V or 3 uA, whichever is greater", which should be used to calculate the parasitic parallel resistance, i.e. V/I = R = 1 / (.01 C) .

Although they might be too cumbersome, and are probably unnecessary to use for this project, I should mention that there are ways to make spice models of electrolytic capacitors with ESR modeled such that the effects of any frequency are included. For example, see Cornell Dubilier Plug-In Thermal / Life Calculator , which also nicely displays how various electrolytic capacitor parameters change with temperature and frequency (and will automatically provide a frequency-dependent spice model for any of their electrolytic caps).
  Reply With Quote
Old 4th June 2012, 07:13 AM   #30
diyAudio Member
 
SyncTronX's Avatar
 
Join Date: Oct 2008
Location: Tejas, where blues are king
Blog Entries: 3
Quote:
Originally Posted by andrea_mori View Post
Based on a Maxim application note, it can cancel noise up to 1.3 mV. Best choise for op amp should be OPA350, but I'm note sure it can tolerate 6V power supply; data sheet indicates 7V as absolute maximum supply voltage and 5.5V max operating voltage.
So why don't you just use that TI part that the Spaniard discussed.
Use the LM833? It goes from +- 5v to +-18v. I ugess you
wouldn't have to worry about low voltage spikes. ; )

Sync
__________________
-- Sync
  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
Regulated power supply meanman1964 Everything Else 13 9th April 2007 03:50 AM
Regulated power supply akunec Chip Amps 3 5th November 2006 11:27 AM
Regulated power supply croccodillo Tubes / Valves 2 2nd November 2006 11:47 AM
regulated power supply adason Swap Meet 1 11th December 2005 07:32 PM
regulated power supply yugaaa Solid State 17 28th January 2003 03:39 PM


New To Site? Need Help?

All times are GMT. The time now is 03:14 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