Phase Locked Loop

[BobEdge]

scr996c.doc
Hi,

I have been looking into using a phase locked loop to detect 3 phase zero crossing points. I have found a very interesting paper (attached) on the subject, which does almost exactly what I want. The trouble is it is a little beyond me.

I was wondering if anyone hase done a similar thing. Any help or advice is greatly appreciated.

Regards
Bob...

[Kamikaze47]

If it's just zero crossing you need you can do that with comparitors. Detecting the zero crossings of 3 phases would just require 3 comparitors.

[Charles Linquis]

If you need to check for the correct phasing of 50 or 60 or 400 Hz AC 3 phase, that is easy to do - No phase locked loop needed.

Please give us more details about what you are trying to do.

[BobEdge]

Hi guys,
Thank you for your interest.

At the moment I am detecting the zero crossing point by feeding the 3 phase into the input through high resistance chains. This works great until one of my welding machines goes 3000 miles away, and is run on a generator. Sometimes we get a meltdown, where the zero crossing is not detected correctly, and the thyristors are fired at the wrong times. This tends to give huge currents, instead of welding at 1500A you get more like 15000A. This is why the paper I attached is of so much interest to me.

Regards
Bob...

[Charles Linquis]

I didn't see the attachment at first.

A few years ago, I "wrote" a PLL implementation that I used for a heart rate monitor. It does a good job of pulling out the hear beat rate in the presece of noise.

If I get any time at all in the next few days, I'll see if it looks suitable for your application.

[amgen]

if sensing zero crossing with an interrupt, how about keep checking that the xing pin or voltage sense pin stays at changed condition for .5 milli sec or 1 milli sec before starting timing firing counter. That should filter noise spikes and you only loose a few percent of phase control at top end due to delay from checking ?

don
amgen

[Charles Linquis]

At the moment I am detecting the zero crossing point by feeding the 3 phase into the input through high resistance chains.
Bob...

So, it sounds like you need a good zero crossing detector, not a PLL - correct?

I have lots of applications where I need to detect zero crossings and I use an opto-coupler like a MOCD207M (Fairchild) in a circuit with a 15K 2W resistor and a bridge rectifier. The transistor in the opto reliably turns off within 5 degrees on either side of the zero crossing. The LED is getting about 1mA at that point. If you increase the resistor's rating to 5W, you can use the same circuit for 120 and 240V. Use a 10K resistor in the collector of the opto.

[BobEdge]

Well I just read Darrels thread, & website on using instant interrupts. This should make what I want to do much easier. I can keep a running average of the frequency of the supply (which can dip using generators) then use the two opto method as in the paper to more reliably detect the zero cross point. Using a few timers to then sort out the firing of the thyristors.

Thanks to Darrel this may not be beyond me at all.

Just need to brush up on using hardware peripherals, rather than using PBP for ADC, timing etc.

Thank's for your ideas guys

Regards
Bob...