Full wave rectifier with center-tap transformer and output stage of Filter and Regulator

MrAl

Joined Jun 17, 2014
11,389
I simply disagree. I appreciate theory just fine, and use it daily. I just don't design power supplies down to the last millivolt unless I'm doing a reference voltage.

Based on an old QC datasheet which I still have, in 1975, I designed a 28 volt linear supply that was accurate within .0002 volts from 0 to 4 amps of load and power line voltage from 105 VAC RMS to 125 VAC RMS, but that's not competitive in the retail market and that's not what I teach to beginners.;) If you think .000714% accuracy across line and load is showing no appreciation for theory we will just have to agree to disagree.

Hi again,

What??? I thought you didnt need that level of perfection :)

I dont think either of you guys get the point yet. 0.0007 percent accuracy is not what this is about, it's about 0.000000000000000000000 percent accuracy. In fact, it is about no error at all. Why would we need such a thing? So we know when we have 0.0007 percent accuracy or 0.0008 percent accuracy, and all by calculation, which does not require a direct measurement of any kind.
Granted this is not easy to get sometimes, but when it is it is like pure magic.

If you can somehow find a way to reject this too, then you must reject all scientific work since year 1. That would be very unreasonable.
 

#12

Joined Nov 30, 2010
18,224
What??? I thought you didn't need that level of perfection :)
I don't. I was just wasting time at work to pad my paycheck during a slow period in the sales department.:D
I just wanted to see how well I could do that while I had a factory full of power supplies and a fully equipped test bench, including a differential volt meter accurate to 5 digits.:D
we know when we have 0.0007 percent accuracy or 0.0008 percent accuracy, and all by calculation, which does not require a direct measurement of any kind.
I did it with the direct measurement way, but I defy you to figure out how I did it by adding exactly one resistor to the basic analog design, and with you using nothing but math.:rolleyes: In fact, I would guess that you can prove it can't be done, using nothing but math.:p
Granted this is not easy to get sometimes, but when it is it is like pure magic.
That's my point. You can calculate the bejesus out of a design problem and reality will usually intrude in some foul way.:(
If you can somehow find a way to reject this
I don't know what it is you think I'm rejecting that would invalidate all my beliefs for all of history. I'm just rejecting the idea that a millivolt quality of design rigor causes a millivolt quality result in the real world, unless you're doing something like precision instrumentation. Have you ever seen a ttl chip that required 5.000 volts for its power supply? And have you ever seen a power supply that delivered 5.000 volts to every chip on a board full of digital chips? It isn't necessary. Nobody does it in the real world. Why should a student even be taught to expect that of himself?
 

MrAl

Joined Jun 17, 2014
11,389
Hi again,

I like what you have to say about this, but unfortunately we are talking about two different aspects of how to solve a problem and why we would want to use different methods.

It's funny because i just ran into this issue recently working with inductance calculations. When i worked with a small book recently purchased, i was getting two different results from two different formulas in the SAME book, so i started to wonder what was going on. Which formula should i use myself for some basic calculations. One of these involved a very simple straight wire, well, possibly two straight wires and their mutual inductance. As it turns out, you can not measure the inductance of a short straight wire, so there was no way to prove or disprove either formula. It turns out though that we CAN calculate the inductance perfectly, based on magnetic theory, and then we have a comparison to the formulas presented in that book. As it turns out, various authors from the past (like up to maybe 1950) came up with various approximations for the inductance, and they all used a different approximation, and these formulas made it into a lot of books. Lucky they work for many different situations, but for some they dont, and that's when we have to turn to pure ideal theory.

But the real question here i think is, why would a professor take the time to calculate the exact solution to a full wave rectifier (no cap ESR) in the first place? What would motivate him/her to do such a thing if approximationas are always good enough?
I would like to hear how you answer this question.
I can probably find the link to the equations which span a full page.

Here's a excerpt from what he wrote:
Full-wave rectifiers with capacitor filters are, without question, the workhorse — and the
unsung hero — of the modern electronic world and for the gadgets we have come to enjoy. That
being said, it is surprising — and sadly so — that a symbolic solution set describing
steady-state circuit performance has been absent.

In 1943, Schade gave a graphical solution.[1] In 1972, Millman outlined a few steps, stating
that “no analytic expression can be given for….; it must be found graphically”.[2] In 1982,
Williams presented a more analytical, albeit incomplete, attempt.[3] In 1991, Kassakian gave, in
Example 3.1, a description for a half-wave rectifier[4] and two years later, Tarter showed
several mathematical steps.[5] In 2004, Shepherd offer yet another result, although it was
questionable.[6] There have been numerous other trials too long to list, but none have come
close to providing a satisfactory result.
Link:
http://www.powerelectronics.com/power-management/analyzing-full-wave-rectifiers-capacitor-filters

That was followed by a full page of equations.

So in your own words, what made him do the work, and talk about this?

LATER:
I almost forgot, if you show your circuit i can take a look at what resistor you were talking about and see what might be calculated. Usually we can find an optimum value but how much time we want to put into it may also become a factor. This does sound interesting though.
 
Last edited:

#12

Joined Nov 30, 2010
18,224
why would a professor take the time to calculate the exact solution to a full wave rectifier (no cap ESR) in the first place?
First, that calls for mind reading, which I am really awful at.:D
Some people call it elegant. I call it, "mental masturbation", but that's what Math Majors (and other of the Intellectual Elites) do.:D
There is good purpose in decimal places for something like aiming a space craft at Jupiter. For finding the last millivolt of the voltage drop of a rectifier in a circuit with a zener diode? Not so much.:(
In this particular Thread, I don't think the professor called for an exquisite rendition of the math. I only fear the student became overly concerned, or overwhelmed, with the plethora of information thrown at him. Being too inexperienced to know which formulas were critical to the function of the circuit, he tried to perfect all of them, when perfection is, in fact, a myth for this application. Or, "There ain't no, "perfect" as long as zeners have equivalent series resistance across a current range of 100 to 1." The true solution to this design problem is a great, long equation which includes everything from the power line voltage to the knee of the zener. The practical solution is much sloppier.
if you show your circuit i can take a look at what resistor you were talking about
The only circuit I can remember talking about here was presented in this Thread, post #1, a resistor in series with the zener diode.
That resistance is dependent on the capacitor voltage, which is dependent on the power line voltage, the transfer function of the transformer, the voltage loss in the rectifiers, the amount of capacitance, the equivalent series resistance and inductance of the capacitor, the temperature of the diodes, and, if you believe in cold, hard theory, the inductance and resistance of the wires or circuit board traces used to connect the components. You want to masturbate that to a milliohm? Fine with me. I already built it and have a circuit running, within specs.:)
 

JoeJester

Joined Apr 26, 2005
4,390
If the professor wanted you to go the three significant digits, they should have told you. If they didn't specify, you could report it any way you want. Did the professor set the stage by requiring some accuracy at any time in the course? If they did, that is what you should give them.
 

MrAl

Joined Jun 17, 2014
11,389
Paraphrased question:
"Why do you think the professor went thought all that trouble to come up with that full page solution of the full wave rectifier circuit?"
First, that calls for mind reading, which I am really awful at.:D
Some people call it elegant.
Hi,

I call it elegant, a lot of people call it elegant.

So you have no idea whatsoever why an engineer at Lockheed Martin would want to solve the full wave rectifier circuit in an exact analytical form? How about why he wrote a whole book on solutions to power electronic circuits?

I am not a mind reader either, but i can come up with plenty of reasons. One of them is especially interesting. But you cant even guess at one reason why.
 

#12

Joined Nov 30, 2010
18,224
I am not a mind reader either, but i can come up with plenty of reasons. One of them is especially interesting. But you cant even guess at one reason why.
I'm not here to guess for your entertainment.
You are going to have to do what you do best without me.
 
Last edited:

MrAl

Joined Jun 17, 2014
11,389
I'm not here to guess for your entertainment.
You are going to have to do what you do best without me.
Hi,

I was already doing that :)
Yes, just for my entertainment, yes, that's it, and that's the only reason for ever asking anyone a question, just for their personal entertainment.
 
Top