Need a circuit to preceisely measure AC Powerline frequency. Please suggest methods or a direct link to a schematic. The freq measurement shall be of a 110VAC/230VAC 50/60Hz line.

Don't want a readymade product.

regds

 

There are many ways to measure AC Powerline frequency. But you did not specify how precise the measurement must be. This will have a large impact on the method to be used. Another spec is how much sampling time you are allowing for the measurement. Obviously too much precision and too short a sampling time can make the job impossible. On the other hand, if you allow long sampling times, the job becomes very easy. I can't suggest any specific method until you set these parameters.

Robert Scott
Real-Time Specialties
Embedded Systems Consulting

 

Google frequency counter. You'll need a pretty good crystal, the line frequency is pretty accurate.

Regards

cww

 

Ummm... Honestly, I havent yet decided on the accuracy range. All I know is that I need to be able to measure two frequencies and raise an alarm if they dont match within 0.3Hz of each other (i.e., F1 - F2 > 0.33Hz should trigger a Relay or some other alarm mechanism). But I realise it was wrong on my part not to have thought over the accuracy band. I didnt really know if the accuracy requirements would change the design methodology all together.

I thought I could use any of the F to V (Freq. to current or voltage) converter chips from Linear Technologies or National Semiconductors. They have a pretty standard & "accurate" way to do that in a fairly economical way. All I was wondering is how to condition the input 110VAC/230VAC so that it can be fed as an input to one of these ICs. (Am sure that none of these ICs would accept a 230VAC directly. So I guess I need to condition the input to use these ICs). I am, in the process, open to any ideas or suggestions or direct links to any schematics that may help achieve the desired functionality.

Please advice, if an exact accuracy requirement would still be needed.

Regards
Grishma

 

Regarding F to V chips, have you looked at their accuracy specs? 0.3 Hz out of 60 Hz is 0.5%. This is very tight for passive components - resistors and capacitors. I think the chips you refer to are based on a RC time constant. If you want to compare frequencies for the purposes of setting a 0.3 Hz threshold, then you should have a measurement method that is accurate to within 0.03 Hz. (0.05% of full scale) I doubt if you are going to get that from a typical F to V chip.

You would be better off staying in the digital realm. Use a counter circuit. Count cycles for 30 seconds. That will give you frequency accurate to 0.05%

Robert Scott
Real-Time Specialties
Embedded Systems Consulting

 

Any commodity freq. counter should handle that requirement. Get a pair with RS232 outputs, read them with your PLC and compare the numbers. That would give you the flexibility to also alarm if they both go out of tolerance, etc. Google low cost rs232 frequency counter. Newport and other "automation" vendors like Red Lion would probably also have these. But the test equipment folks might be cheaper with better specs.

Regards

cww

 

Thanks, but that still leaves me without any information on conditioning the input. The input shall be a a continuous sine wave. 230VAC/110VAC. Should I use a step down transformer? Should I then digitize the sine wave and then use the counter ckt to find out the difference?

I am seeking a broad block level idea. Later, I can carry on from that on my own. But in this case, am just too confused to begin with.

Regards
Grishma

 

As for signal conditioning, I think you are right to use a step-down transformer. But then you should also think about noise. If you digitize with a simple zero-crossing detector, then you will be open to noise causing multiple pulses when the sine wave is near the comparator threshold. Use some hysteresis to ensure that you only get two transitions for each complete cycle. Hysteresis is commonly used in comparator circuits. Also, depending on how you process the digital signal, you could do the equivalent of hysteresis in the time domain by using a one-shot, either in hardware or in software.

Robert Scott
Real-Time Specialties
Embedded Systems Consulting