Reinventing the square wheel / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

Reinventing the square wheel

Move Thread

LAN_403

Sedi

Thu Oct 22 2009, 08:32PM

Registered Member #2432 Joined: Tue Oct 13 2009, 11:12PM
Location:
Posts: 2

Hello all, I've been reading for a while here in the forum and on various coilers' websites, and enjoying it all, especially some of the hardcore *SSTC explanations... really great stuff. I think I understand some of it but in EE, I'm a noob at best.

Anyway, here's my first question: Is there an easy way to create a square wave from a sine wave, in which the square wave's pulse duration is equal to one quarter of the sine, so that the square is also in phase with the sine? I want to create a pulse that begins at a zero crossing and ends at a peak(or trough, if it's not rectified). Put another way: is there already an IC or a published circuit with a digital output that can toggle when a signal peaks, or passes 90 degrees at any frequency?

I've been trying to pound this into LTSpice... the latest idea is based on watching for the amplitude of the wave to match the amplitude of a slightly delayed sample of the same wave, meaning the original signal has just peaked and started falling. The original signal is sent into a comparator and a unity gain amplifier, and the propagation-delayed output of the amplifier is also sent to the comparator. Which is which input on the comparator doesn't matter until I know whether I'd need a rising or falling edge on the digital side of things. So far it seems I can't very easily match the amplitude and shape of the amplifier's output exactly to the original signal, which I understand will be utterly essential (and somewhat easier if it was all one IC). Probably the only reason this hasn't worked in LTSpice is because I don't *really* know how to wire up an opamp and LTSpice knows this

Also I need to get some better models into it, to say nothing of my head.

I could be very wrong but I don't think I want a PLL or a PWM controller... nothing continuous. I've been searching for phase, peak, angle, detector, etc. on the web and forums and wikis for something that ought to be simple but I keep getting buckets of PWM circuits that do the opposite of what I want. I don't want to detect the maximum voltage on a repeating pulsed DC signal or show whether two signals are in phase or how far apart, I just want to know right at the peak, how long ago it was at zero. This would probably be easy to do with a microcontroller, especially since it could then immediately use that information for the next stage, but I want to get my mind around analog things for now.

I expect you'll ask why I even want to know, and the purpose is ZCS with an utterly chaotic Fres. I'd say almost unpredictable, but I hope that the first quarter wave will allow me to cleanly end the second quarter. The other 180 degrees are just the same thing again, and the higher the instantaneous Fres, the lower the chances of it changing significantly in that cycle. Again, I hope.

Thanks for any hints, and all the info so far.

p.s. The closest thing I've seen yet which I meant to include is US Pat#5852794. I'll be reading.

Mads Barnkob

Thu Oct 22 2009, 09:13PM

Registered Member #1403 Joined: Tue Mar 18 2008, 06:05PM
Location: Denmark, Odense C
Posts: 1968

A hex Schmitt inverter as 74HC14

would be useful, series up two inverters to get a positive square signal.

It sounds to me like you describe something simple like steve wards sstc5 driver:

klugesmith

Thu Oct 22 2009, 10:02PM

Registered Member #2099 Joined: Wed Apr 29 2009, 12:22AM
Location: Los Altos, California
Posts: 1716

Sedi wrote ...
I want to create a pulse that begins at a zero crossing and ends at a peak(or trough, if it's not rectified). Put another way: is there already an IC or a published circuit with a digital output that can toggle when a signal peaks, or passes 90 degrees at any frequency?

Set up one comparator to detect zero crossing of the input sine wave.
Use another comparator on a differentiated or integrated sine wave to get the peak event.
If both comparators have open collector outputs, they can be wire-ANDed to get the pulse you want, with no additional logic.
The differentiator or integrator could be as simple as a RC network. Start the design by setting bounds on the frequency and voltage range to be handled.

klugesmith

Fri Oct 23 2009, 02:52AM

Registered Member #2099 Joined: Wed Apr 29 2009, 12:22AM
Location: Los Altos, California
Posts: 1716

This would be an edit to previous post, except it's a pain to add pictures.
You mentioned trouble with LTSPICE, so I hope this example helps. From their components menu I selected the first comparator, and it offered a demo circuit. That was the starting point for:

Sedi

Mon Oct 26 2009, 01:08AM

Registered Member #2432 Joined: Tue Oct 13 2009, 11:12PM
Location:
Posts: 2

Thanks for that, but it's all even more pointless than I realized. I was hoping to automatically minimize deadtime in a halfbridge, but after reading more about Class DE I can see that's the opposite of what I want-- and now I can think with a smaller secondary to boot! My timer never would have worked, I can see that now just from Richie's note about the shape of a spark-loaded primary current waveform.

Hey Steve, thanks for putting that paper (classde.pdf) on your site. edit: and Richie too, I knew I had seen a copy of that before! I forgot what it was called and went looking again

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob