Printable View
Hi, I have a very simple circuit setup to detect AC zero crossing. There's a mains AC transformer (230 to 12V) which connects into a full wave rectifier. There are no other components.
When I measure the raw AC out of the transformer with a scope I see a nice sine wave as expected. When I measure after the rectifier I see what almost looks like DC ripple but with every 2nd trough cut off. (screenshot attached)
Has anyone see this before and does anyone know what's causing it?
Attachment 8023
Do you have some kind of load?
Nope. There's no load. Could that be the problem?
Yes I think. Try with something like 1k ohm. Probably the capacitance of the junction or some other esoteric phenomenon.
I have seen that before , I was due to the ground reference for the scope not being the same as for the rectified waveform
I've tried connecting a resistor as the load but it didn't have any effect.
I've actually got 2 circuits. The one described in the initial post is a test circuit to try to isolate a problem I was seeing in the full circuit. The scope ground is connected to the negative side of the rectifier output in both circuits.
The full circuit has 2 resistors used as a voltage divider between the rectifier output and a PIC chip. It also uses 6 diodes for the rectifier to ensure the rest of the circuit doesn't affect the zero crossing detection.
The PIC chip is also only seeing half of the zero crossings (which is what prompted testing with a scope).
I've attached a rough schematic with the relevant parts of the full circuit.
Attachment 8030
I'm not sure the regulator will like to have a capacitor at the output and a full ripple at the input. I've learned (30 years ago) you have to put a diode between the in/out pins of linear reg to protect them for any situation where there is a lower voltage at the input vs the output. Also that may explain the waveform you have.
Anyway, take a look a this very interesting circuit:
http://www.dextrel.net/diyzerocrosser.htm
I didn't know that but it makes sense as I've blown regulators up by soldering them in the wrong way round before.
My original thought was that the capacitor was causing the strange waveform so I setup the test circuit which has only 4 diodes (full wave rectifier) and now a resistor for load. There's no capacitor or regulator attached.
Thanks for the link. It's definitely more complicated than what I've got now but it will be really useful as I'm gradually moving towards switching power supplies
The regulator needs a capacitor on it's input. Read the datasheet - they usually have an example circuit. But this may not explain the waveform you are seeing. It could also be your ground reference for the scope. The scope ground should be on your regulator REF pin. Also, you don't identify the IC1 device, so no idea the characteristics of that pin you are sensing with. Sometimes this can be caused by differences in the diode forward drop characteristics, but 4 volts is too much to be explained by that. You will likely be exceeding IC1 input voltage, possibly triggering internal protection circuitry, and so on. Your 10K and 1K probably need to be reversed,
I specifically used 2 extra diodes to avoid any of the capacitance problems I'd have had using the same positive that supplies the regulator. In theory the main part of the circuit shouldn't interfere with the zero crossing part (not that much anyway).
The schematic I posted above is just a guide for what my circuit looks like. The scope readings were all taken from the test circuit which I've attached below. There's nothing that should interfere with it at all. This test circuit doesn't have any capacitors or regulators or anything, just diodes and now a resistor.
Attachment 8036
Just a couple of thoughts:
Do you get the same waveform when AC coupled as DC coupled?
How about changing channels on you scope. Same waveform?
How about trying a different probe?
It's doubling frequency so it appears the all your rectifiers are working but maybe change the location of two and see if waveform changes. Could be a bad one or two rectifier.
cro1 as per test circuit
cro2 as per test circuit with diode in series with cro ground
looks the sams as per post 1
Not really. Bottom plot in post 12 looks almost normal. It's much more symmetrical than original plot but only slightly better than to plot in post 12.