RMS approximation of Square/PWM signal

I feel so good that you need to address me by my real name and not the alias we use here. So let me return the favor and do the same for you..

Phil/Phillis = Mr/Miss. Asshole!. Depends on what day of the week.

That is all his mother could come up with before she through him/her out with the bath water..

In this case, A child that not even a mother could love!

"

"JMini"

** A moment ago you thought a small lamp did not respond fast enough ?

Got a clue how slow that R plus NTC combo would have been ?

BTW- what sort of mad design is this ?

A tiny 30 x 6mm cavity for a complete 20 amp PWM regulator ????

But the lamp is right next to it too.

...... Phil

No wait John, I'm being blasted by yours truly for suggesting the same thing.

wait your turn, I'm not done getting blasted yet! :)

"

Jamie" = Maynard A. Philbrook KA1LPA

A f****it, asshole radio ham and bad code scribbler by trade.

Radio hams don't even know the significance of rms values.

You are an absolute FUCKWIT Philbrook.

FOAD you retarded, autistic POS cretin.

...... Phil

Yawn! Same boring shit from demented Phill/Phillis, different day.

I knew he/she didn't have anything else to offer..

"

But he's controlling it... and he doesn't know it?

John

"John Larkin"

** You are the one to say how he knows it.

Cos that is your claim.

...... Phil

If he's generating it digitally, like from a uP PWM controller, he knows. If the system is all analog, it's still not hard to measure the duty cycle, but the math gets hard to do.

John

Squaring is dead easy and you don't need multipliers. Just double integrate the sensed voltage (it's roughly a constant) then average the sampled value at the turning off time. Obviously reset the integrators at the turning on point. With enough luck (I didn't check) and the right values this might be as simple as a pair of switches and a pair of LPFs.

--
Thanks,
Fred.

Cute, Fred!

a boost PFC using V^2 rather than V is a linear control system. no small-signal perturbation required. this is, IIRC, called "input linearisation" and is the easy way of building nonlinear controllers (thanks mr Slotine)

you must have read some of Keyue Smedley's & Dragan Maksimovic's one-cycle control papers :)

Cheers Terry

Yes I remember I did (Maksimovic for sure, others don't sound familiar) but I didn't though about them. Rather I have a nice collection of old

70's semiconductors appnotes (you know the spirit: do everything with nothing) that I have pretty well assimilated when young (in French I'd have said 'integrated' but I don't know if English carries the same meaning) which bent my mind quite a bit make all this natural for me.

Now, how do you as easily invert some value? :-)

--
Thanks,
Fred.

Yeah. I figured the response of the resistor/NTC would be worse. I was really just thinking of anything that could read an RMS value. It's a design to fit inside a customized D cell maglite. Some of the real wacko designs run as high as 200+ watts (10-10.5A) The battery is no further than 1.5 inches away, and the bulb is only about 1-1.5 Inches away. I can use an RMS converter that requires only one external cap and a few external voltage divider resistors. I was hoping someone could come up with an idea like "Oh sure, you use an op-amp and a cap like in the attached schematic! F*wit newbie!" There are some great thoughts in the thread, but most have a parts count higher than what I'm using now. I guess I'll get into the shop and stat fiddling with the RMS converter feedback to the Ti TL5001.

That's just it. I'm using a TL5001 PWM controller, Not a uC. I just don't have the skills to implement a uC. I'd love to learn how. The AVR Tiny85 looks pretty nice. But I have no idea how to program it or how to write code to do regulation with PWM.

OK. It's a PWM regulator for a flashlight. It has to be small to fit inside. Voltage range at input will be between 6 and 30V depending on battery pack chosen. Output voltage will be between 7.2 and ~25V RMS. Duty cycle will vary widely. The PWM controller I'm using is the Ti TL5001. It has a built in programmable soft-start.

and if it's a square wave - with 50% duty cycle, you'll get 1/2 the power, or equivalent .707 times the peak voltage. For the RMS (root of the mean of the square), you'll get v^2, the mean is d*V^2 (d=duty cycle), and the root is sqrt(d*v^2) or sqrt(d)*v. For a 50% duty cycle, 70.7% of the square wave top. Mea culpa!

-Paul

CdS feedback from the incandescent lamp?

Best regards, Spehro Pefhany

--
"it\'s the network..."                          "The Journey is the reward"
speff@interlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com

** The whole thing must get very hot - or is all the IR light reflected away by the mirror? ** Why on earth have you selected 40kHz as the PWM frequency ????

With no LC filtering to consider - you should be using a frequency more like 400 Hz.

Then, even the most basic ( ie low supply current) ) RMS to DC converter IC s will work like a charm.

..... Phil

"Phil Allison"

** I see it is a limitation of the TL5001 that has you stuffed - it was never intended for what you are doing,

Better go find another PWM IC that will run at low frequencies.

..... Phil

"JMini"

** When the duty cycle is low - you need to watch out how high the pulse current becomes.

Example:

V batt = 30 volts, V out = 7.2 at 100 watts

Duty cycle = ( 7.2 / 30 ) squared = 0.0576 or 5.76%

I peak = V batt / R lamp = 30 / 0.52 = 57.8 amps.

That is way too high.

..... Phil