hi n advice.

[drm]

Hi All!

This  is my first post.

Am looking for advice about a reliable zero cross detection, for a SCR DC motor drive.

If someone can help with this issue, will tell the full story.

By the way. here is lil picture of the notch that is disturbing  my zero cross detector made with a simple NPN transitor.






Thanks in advance.

[Paul Price]

Can you draw a hand-drawn schematic and scan it into your computer, be sure to show power supply voltages and attach as a p icture.

Tell mel what you mean by zero-crossing...in relationship to the mains? Do you need a mains AC  60-Hz zero crossing point reference for triggering the triac?

[drm]

hi Paul, thanks for your replay

Right, need a way to detect zero crossing on a 50hz line, to fire the scr
in syncro with mains.

When firing point is near 0 crossing, notch distorts input signal to
0 cross detector.

Now its passed 0:00 AM here, tomorrow will draw the circuit and add more info


in my utube channel, theres more videos about the drive.

[drm]

Heres the circuit diagram.

Am making a video showing important TPs, that upload when ready.

[drm]

Video with TPs

[drm]

TP LOAD, is 50 v-div, sorry.

[David_AVD]

What happens when you remove the R/C network ("delay") ?

[drm]

Hi David.

It was added to override diode barrier of the first transistor, and make zero cross pulse, closer
to real cero voltage. Shifting it, a bit to the¨right¨

if I take it out, circuit looks a lil more prone to sporius firing near 180 degrees, but not big deal.

Was reading that a 0 cross detector with OP AMP and histeresys, and PIC micro programmed to reject
out of time 0 crossing, can make it more reliable .May be

Keep watching ...

[David_AVD]

Actually, that base resistor looks really low in value.  I'm sure I've used 1K in a DMX AC dimmer project.

[drm]

Thanks to mention, watching now can see that I made a mistake about that part of the circuit diagram.

68 ohms R, goes to the union of 300k and 1uf.

Like correction.jpg shows now.


It should take 3,2 amp !! connected to 220 vac, as I drawn it before.

[David_AVD]

Oh, it's a feed direct from that mains!  Mine was via a transformer secondary.

[David_AVD]

Also, please resize your pictures down to around 800 x 600 max or thereabouts.  Having to scroll around all over the place really turns people off from looking and replying.

[drm]

OK.

[drm]

motor ready to fly...

[Paul Price]

Double Crossing Circuits that produce a clean zero-cross output.

[miceuz]

I've used an optocoupler-based zero crossing detection circuit for a recent project. It's just an "AC input" optocoupler with two LEDs in antiparallel on the input and two power resistors to drop the voltage. Works a treat and not much mocking with mains voltages. The drawback is that those two resistors are getting really hot - around 95C.

[David_AVD]

You should be able to increase the value of those two resistors to dramatically reduce the heat.  Picking an opto with good transfer characteristic should help a lot with input current requirements.

EDIT: That 200R pull-up could be way higher.  That's probably the reason you have to drive the opto input so hard.

[drm]

Thanks for those circuits 

Will simulate them , and discover if Msim is able to show the notch produced by a 220uf
condenser, connected to 310v peak via a 2,5 ohms R ...

[miceuz]

EDIT: That 200R pull-up could be way higher.  That's probably the reason you have to drive the opto input so hard.

oh, yes, the pullup is 10k, it's just a mishap in schematics. I'm using two 8.2k resistors. Those give me 340V/16.4k=20mA peak. Maybe I could go somewhere to 10mA peak CTR will be only 5% lower, but this would give me a more narrow ZC pulse and much lower power dissipation. Will try in a moment, thanks!

[miceuz]

With double the resistors (32.8k) and 20k pullup I'm getting 400uS wide pulse (the bottom of it, the digital pulse is more narrow) and they are rising 16 degrees above ambient. It's a win! 

But now I have 4 resistors in series - they all are sharing the load. With two 18k resistors I would get somewhat 240V^2/36k = 1.6W -- 0.8W dissipation per resistor, I would guess that would effect in quite substantial rise anyway?

[David_AVD]

For 0.8W dissipation I'd probably use a 2W resistor.  Try it out and measure the temperature.

[drm]

So Msim can show a notch over ac mains.

u can press space bar, and watch violet trace on virtual oscilloscope,
simulating attached file: making a notch.

[drm]

was no luck to get a reliable zero cross detection  (didnt try with PIC even) , because of the big notch caused by load conmutation.

So i decided by now, to replace the scr by a 6 amp and 200v mosfet that got at hand.

And feeding the output stage with 110 vdc.

Load voltage is modulated by pulse with 555.

Look the inrush starting currrent of my DC motor, and how its decreasing, as rpm and rotor back emf  goes high and get stable.

[drm]

good info about notches.

[drm]

very useful app notes about scr, triacs and power mosfet from philips

http://www.4shared.com/file/x_IWLEAY/scr_triac_pmosfet_app_notes.html