Power Supply Resevoir Size - Page 125 - 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 17th September 2012, 08:18 PM   #1241
Krisfr is offline Krisfr  United States
diyAudio Member
 
Krisfr's Avatar
 
Join Date: Nov 2007
Location: near the House of the Mouse
Quote:
Originally Posted by OnAudio View Post
Thanks Krisfr, hopefully Gootee will provide the formulas
It is easy, don't think I am that smart
__________________
Thanks
Never give a Systems Analyst a screwdriver or hammer
  Reply With Quote
Old 17th September 2012, 09:48 PM   #1242
diyAudio Member
 
danielwritesbac's Avatar
 
Join Date: Nov 2007
Quote:
Originally Posted by Krisfr View Post
I can do a Android app, just give me the inputs and what outputs you want based on formulas. The oscope and DA are out of my scope, and they can be done, but not by me...
See the interesting file that Tom attached to post 1219.
__________________
Tools, Models & Software for DIYClipNipper boostLM1875 TurboPowerful TDA7293 kitTDA7294 pt2pt ♦ My post has opinion.
  Reply With Quote
Old 17th September 2012, 09:57 PM   #1243
Krisfr is offline Krisfr  United States
diyAudio Member
 
Krisfr's Avatar
 
Join Date: Nov 2007
Location: near the House of the Mouse
I looked at the spread sheet, but I could not directly see the formulas.

My display made some of the fields tiny.

When Tom "Plainly" states what they are I will put it on an app.

I follow most of what has been detailed in this thread, but I wish someone would state the max max of 110 at 60 htz with 15 or 20 amps or 220 at 50. There should be some magic power output and rail voltage combo to be derived a great sounding amp.

Maybe I just want to build the biggest one I can based on 110, then it will loaf along never being stressed, never hitting the rails, over kill my way into the sweet spot.
__________________
Thanks
Never give a Systems Analyst a screwdriver or hammer
  Reply With Quote
Old 17th September 2012, 10:21 PM   #1244
diyAudio Member
 
danielwritesbac's Avatar
 
Join Date: Nov 2007
Quote:
Originally Posted by Krisfr View Post
Maybe I just want to build the biggest one I can based on 110, then it will loaf along never being stressed, never hitting the rails, over kill my way into the sweet spot.
Is your mains voltage really that droopy? We both live in the USA, but my outlets do 120v@60hz.
P.S.
Well, the 60hz is wishful thinking--there's a lot of bonus noise content, but the higher pitched signals don't have as much strength as the 60hz; yet the bonus noise content is enough to drive an ordinary transformer, rectifier, caps, setup to a few volts higher unloaded and that makes more rail bounce.
__________________
Tools, Models & Software for DIYClipNipper boostLM1875 TurboPowerful TDA7293 kitTDA7294 pt2pt ♦ My post has opinion.

Last edited by danielwritesbac; 17th September 2012 at 10:25 PM.
  Reply With Quote
Old 17th September 2012, 11:06 PM   #1245
Krisfr is offline Krisfr  United States
diyAudio Member
 
Krisfr's Avatar
 
Join Date: Nov 2007
Location: near the House of the Mouse
I am saying 110 for the sake of talking, I have not measured any voltage on my home.

Maybe I need to do a SMPS, no chance of droop
__________________
Thanks
Never give a Systems Analyst a screwdriver or hammer
  Reply With Quote
Old 18th September 2012, 06:47 AM   #1246
gootee is offline gootee  United States
diyAudio Member
 
Join Date: Nov 2006
Location: Indiana
Blog Entries: 1
Quote:
Originally Posted by DF96 View Post
I thought I had already done that! At least for Vrms, which is by far the strongest influence on Cres. I seem to keep repeating myself in this thread. Am I just wasting my time? I thought you had produced a formula too?

In the early days of radio people didn't know how antennas actually worked, so they just played around with different lengths and shapes and established rules of thumb. Then one day someone had a good theory and said "Just put a quarter-wave length of wire straight up into the sky". At first they carried on bumbling around, believing that it couldn't be that simple. Eventually they realised that he was right.

Years later a new generation came along who for reasons we don't understand did not trust theory (or didn't understand algebra - we are not sure which it is). They started trying different antennas. An old-school engineer kept saying "Quarter-wave!", but they ignored him. Curiously, at the same time some of them began asking if anyone had a formula for the right length. How long before they 'rediscover' the right formula? (This story is not entirely historically accurate - it is told to make a point).

As I keep saying, to get the minimum cap value you need to do a ripple calculation (this involves Vrms, Vpk, Rload, Vdrop). You will find that the Cres depends critically on (1.414xVrms -Vpk - Vdrop=Vripple-max). Then double it. This will be close enough. There is little advantage in going bigger, and there may be disadvantages in going much bigger.
DF96,

I understand your point but I think that most people here know how to calculate the reservoir capacitance using the standard approximate formulas, based on an RMS or DC load current. (For example, see Unregulated Power Supply Design, or, for another method, my post #1061 at Power Supply Resevoir Size , but with the + sign changed to a minus sign in step 5. (However, note that in the example at the end, the resulting capacitance was too LOW, by a factor of just over 2. The actual answer should have been 6650 uF instead of 3000 uF. But also note that the calculated value was not "doubled" arbitrarily. The method calculates the C of both rail capacitances as one capacitance, but since they are in series as far as the calculation is concerned, the value has to be doubled to get the two individual rail capacitances.)

I think that I was originally wondering about the possibility of those types of standard methods not accounting for non-constant load currents and the worst-case timing of load-current peaks and charging pulses, while I was investigating just the very first step of finding a general rule for "What C value could and should we use?", with that first step being to see if I could find a way to know the "true" lower bound of the C value, for sinusoidal load currents at the rated maximum output power. I have had limited time and that first step has seemed quite belabored.

Anyway, I think that the part where you say "Then double it." would need to be justified. I'm sorry if you already did that. I know there were a couple of posts of yours, with equations, that I didn't get to study and that I meant to go back to, but haven't yet.

At any rate, my original plan was to try to use simulations to find the ACTUAL lower bound for the C value, for different cases, and then compare that to what the standard equations give, and then try to move forward from there. I think that I am just now finally able to compare the simulation results to the standard equations' results, which I attempt to do, below or in another post.

I do realize that you pointed out, correctly, that it will also depend on the amplifier topology, since the actual absolute-minimum C depends on the minimum voltage that can exist across the amplifier, i.e. between the rail and the high side of the load, with the minimum C being the one that allows the rail voltage to have a worst-case drop that doesn't quite reach Vload_max_peak + Vamp_minimum, for the worst-case timing of a load current peak and the first quarter of a charging pulse. But I can either leave the minimum Vamp as a variable, or, just add a few volts (for a minimum Vce plus a small series resistor), for now, and be close enough for this purpose.

(Yes, it appears that a rail-voltage incursion into the output voltage only occurs during the upward-curving portion of the beginning of a current pulse from the transformer, between zero and the first inflection point. Made me wonder if the second derivative of the transformer current was important.)

On to the standard equations: (I HOPE that someone will check my algebra and my arithmetic!)

From Unregulated Power Supply Design :

Ripple Voltage (peak-to-peak) = Vr = i_load_rms / (2 ∙ f ∙ C)

Peak Rail Voltage = Vc_pk = (Vrms_secondary x 1.414) - 1.4

Average rail voltage = Vc_avg = Vc_pk - (Vr / 2) = Vc_pk - ( i_load_rms / (4 ∙ f ∙ C) )


For the absolute minimum C, we would need:

Vc_avg - (Vr / 2) ≥ Vout_pk + Vamp_min

in order to not have any rail voltage dips gouging into the output voltage.


But Vc_avg - (Vr / 2) = Vc_pk - Vr = Vc_pk - ( i_load_rms / (2 ∙ f ∙ C)) .

So we can re-express what we need as:

Vc_pk - ( i_load / (2 ∙ f ∙ C)) ≥ Vout_pk + Vamp_min


Solving for C, we need to have:

C ≥ i_load / (2 ∙ f ∙ (Vc_pk - Vout_pk - Vamp_min))


That's actually a little better than the "standard" formula, since it also accounts for the minimum voltage across the amplifier. This should give slightly higher capacitance values than the usual standard formulas.


We already know:

i_load_rms_max = √(Power_rms_max / R_load)

Vout_rms_max = √(Power_rms_max ∙ R_load)

Vout_pk = (√2)(√(Power_rms_max ∙ R_load))

Vc_pk = (Vrms_secondary x 1.414) - 1.4

Vamp_min = 3 (Min Vce plus worst-case voltage across the 0.22-Ohm R)

--------------------------------------------

To compare to the four cases for which I recently posted simulation results,

Power_rms_max would be one of 25, 50, 75, and 100 Watts RMS,

giving i_load max rms values of 1.77, 2.5, 3.06, and 3.535 Amps RMS,

and Vout_pk values of 20.00, 28.28, 34.04, and 40.00 Volts peak.

With the 44-0-44 transformer output, we get

Vc_pk = 60.83 Volts peak.

Plugging in those values and turning the crank yields C_minimums of

C = 390 uF, 705 uF, 1100 uF, and 1650 uF, for

Power = 25 W, 50 W, 75 W, and 100 W.

But, with a 240 VA secondary for each rail (a 480 VA transformer), simulations found the "actual" C_minimums to be:

C = 330 uF, 770 uF, 1540 uF, and 6650 uF

------------------------------------------

It appears that the standard equations are assuming an ideal transformer, and that they also don't account for the occasional badly-timed load current peak relative to a charging pulse.

So the standard equations do well when the transformer is ridiculously over-sized compared to the output power spec.

HOWEVER, as I've been saying repeatedly, everything in my simulations depends on how good the transformer models are. And Terry Given just mentioned that it looks a little wacky.

Does anyone have some "typical" values for such a model? (The transformer ratings don't have to be the same as my original ones.)

Hey Terry, I have an LCR meter that I did not have, back when I was doing the original transformer measurements. Could that help to check them? I think that I still have the original 120 VA 25 Vrms transformer that was measured.

Cheers,

Tom

Last edited by gootee; 18th September 2012 at 06:59 AM.
  Reply With Quote
Old 18th September 2012, 07:13 AM   #1247
gootee is offline gootee  United States
diyAudio Member
 
Join Date: Nov 2006
Location: Indiana
Blog Entries: 1
Quote:
Originally Posted by AndrewT View Post
I am sad to have to admit I have given up reading this Thread, mainly because the posts are so long and it was just taking me too long to read and follow all the technicalities.

So I just pop in now and again to see the latest.

The above has me confused.
I use a 35+35Vac transformer for a 100W into 8r0 amplifier.
Why the reference to 44-0-44 Volts transformer?

Is that a peak of the AC waveform? or the AC voltage after applying the transformer regulation, or taken literally as a 44Vac-0-44Vac transformer giving supply rails when powering a quiescent output stage of very approximately +-60Vdc.

I am lost and this may be due to not following the Thread in detail.
Andrew,

I temporarily used a 44-0-44 transformer model in order to try to remove the effects of having a rail voltage that was already relatively close to the maximum peak output voltage (with a possibly-too-weak transformer).

I just wanted to see how the numbers came out, for the highest C that would still make Vrail droop until it corrupted the output.

And I also wanted to see if any trends came into focus better, when there was not so much of a problem with the headroom.

But I'm starting to suspect that the particular transformer model parameters I am using might not be giving very good load regulation.

However, regarding the load regulation, we should also keep in mind that I am simulating with square waves that have the same peak amplitude as would a sine wave that was producing the maximum rated output power. i.e. The square waves produce 2X the maximum rated output power. (But still, the transformer seems like it should handle it way better than it is.)

Regards,

Tom
  Reply With Quote
Old 18th September 2012, 07:19 AM   #1248
gootee is offline gootee  United States
diyAudio Member
 
Join Date: Nov 2006
Location: Indiana
Blog Entries: 1
Quote:
Originally Posted by AndrewT View Post

<snipped>

I'll go and re-measure a 35+35Vac transformer. Volts and ohms only, I have not worked out yet how to do the complicated stuff.
Andrew,

If you have a VARIAC, and a low-value power resistor, that's all you need besides a multimeter.

The complete (and very simple) instructions are printed right in the diagram, at Power Supply Resevoir Size .

You might want to read it to see if my safety warning was adequate, for people playing around with exposed mains connections and no safety devices whatsoever, except their breaker box.

Regards,

Tom
  Reply With Quote
Old 18th September 2012, 10:25 AM   #1249
AndrewT is offline AndrewT  Scotland
diyAudio Member
 
Join Date: Jul 2004
Location: Scottish Borders
I'll try to fit in the whole test today.
__________________
regards Andrew T.
  Reply With Quote
Old 18th September 2012, 10:52 AM   #1250
DF96 is offline DF96  England
diyAudio Member
 
Join Date: May 2007
Quote:
Originally Posted by gootee
I think that I was originally wondering about the possibility of those types of standard methods not accounting for non-constant load currents and the worst-case timing of load-current peaks and charging pulses, while I was investigating just the very first step of finding a general rule for "What C value could and should we use?", with that first step being to see if I could find a way to know the "true" lower bound of the C value, for sinusoidal load currents at the rated maximum output power.
I think you have demonstrated that the 'standard methods' are completely adequate, when used properly i.e. circuit theory actually works OK.

Quote:
Originally Posted by gootee
Anyway, I think that the part where you say "Then double it." would need to be justified.
The justification is simple: normal engineering caution, combined with broad electrolytic value tolerances (can be 20%). The calculation gives the minimum. There seems to be no harm in increasing this (within reason) so increase it. There is no point, and possibly some harm, in increasing it too much. Hence double it. x3 or x4 would be just as good, although maybe unnecessary.

The calculations you have done show good agreement with simulations for lower powers. The discrepancy for higher power mainly comes from Vmin being a bit too small. This has little effect at low power (i.e. high ripple voltage) but is critical at high power (low ripple voltage). As you have included the 0.22ohm drop in Vmin (not a good idea, in my opinion) your Vmin should be bigger for greater power. Try repeating your calcs for Vmin=4V instead of 3V.

Including transformer resistance in the calculations gets messy, unless the resistance is sufficiently low that the cap can recharge more or less fully on each mains half-cycle. As I said in an earlier post, this gets worse for really big caps which may be a reason to avoid really big caps.

So we are in a position where we can calculate, with some confidence, the minimum cap value. We can also estimate, roughly, the maximum. Isn't that 'job done'?
  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
Valve power supply - How to size transformer? SanderW Power Supplies 25 4th January 2013 05:12 PM
How do you calculate choke size in a power supply? Original Burnedfingers Tubes / Valves 25 5th January 2012 01:23 AM
power supply bypass cap size BigE Power Supplies 11 5th July 2011 03:59 PM
Power Supply Case Size diymixer Power Supplies 1 10th October 2010 06:47 AM
What size power supply should I get for repair work? spooney Car Audio 3 7th December 2007 12:50 AM


New To Site? Need Help?

All times are GMT. The time now is 08:27 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