Converting Sinusoidal to Digital

So i have to confess that I am not an electrical engineer, just a plc & controls operator, though we did study op-amps in trade school.. Could you provide me with a part number and a connection diagram for the op-amp that you think will work?

First, you should read the hyperlink (blue lettered op-amp) to understand the circuit topology. The schematic in the link outputs a 0 to 5V square wave.

You did not specify several critical things: the range of the input voltage (voltage swing), the frequency range of the input voltage, the desired trip voltage level, the PLC model number, or the I/O module model number. To prevent excessive jitter, I often designed for a ±1% to ±10% hysteresis of the desired threshold voltage.

Without that critical information, I therefore cannot provide anything more than another comparator tutorial. Still and all, learning how to design reliable PLC interfaces will do you well in your programming career.

You need a pulse conditioner or a pulse input conditioner. It's a module that mounts on a DIN rail that takes in weak pulses and cleans them up to provide a consistent pulse amplitude. Make a google search or look in manufactuers like DataForth, Pepperl and Fuch, Efector, Phoenix Contact, etc.

Before that, however, what supply voltage are you using for the flow sensor? If you're using a 24 Vdc supply, then try the signal conditioner. If it's lower than 24Vdc, try using a 24Vdc power supply. It could also be that your power supply is current limiting, which can affect your sensor's performance. In that case, get a higher power power supply.

regards,

Vulcan

The sensor isn't powered, it creates its own ac pulse from 1-20 vac(acts like an ac generator). At a low pulse rate, generates only 2volts, at higher pulse rates, up to 20volts. So what i really need is some kind of a transistor, that is 100% on at 2vdc, to send 24vdc pulses to my plc input..

OKay, I looked up the 515. I used to use a different model that had a powered output and I thought they operated in the same way.

So, your output is a sinusoidal signal not pulses. You can use a transistor circuit, but I'd rather use an op-amp as a comparator so convert it into a clean, constant amplitude, square wave. I can't tell from the manual if the signal swings positive and negative (I think it does), so you probably need to half-wave rectify it before feeding it into the comparator. I'd make the reference voltage variable (with a trim pot) so you can adjust the lowest voltage that it'll trigger cleanly.
However, my earlier suggestion of a pulse converter module (which can convert a sinusoidal signal to clean square-wave pulses) would be a faster, albeit more expensive) solution.

regards,

Vulcan

Thanks you for the help. So I usually get my electronic components from digikey.ca, they literally have thousands of different op amps. Could you reccomend a part number that will work for this application?

Yes, there are thousands of different op-amps available at Digikey. But if one uses their convenient selection filters, the number of devices quickly drops to a manageable number to choose from. While virtually any op-amp can be configured into a comparator circuit, it is worth noting that in their filter selection, one of the first choices is "comparator." These components have been designed to prevent a "latch up" condition. Next, I selected a few additional filters (availability, through hole, supply voltage), which (I hope) are relevant to your fabrication skills, and dropped the number to a few dozen parts to choose from.

The packaging of the comparator circuit will then depend on several things yet to be discussed, such as the environment, cable lengths, and cable terminations. Then there's the question of the anticipated installation lifetime.

It would be useful to know the output impedance of this sensor. I suspect it to be relatively low, like a few hundred ohms. Knowing this will help determine what resistors will be preferred for the rest of the comparator circuit.

GF Signet 515 pdf

Electrical

Frequency 19.7 Hz per m/s nominal 6 Hz per ft/s sinusoidal

Amplitude 3.3 V p/p per m/s nominal 1 V p/p per ft/s

Source Impedance 8 KΩ

Cable type 2-conductor twisted pair with shield, 22 AWG

Cable Length 7.6 m (25 ft) can be extended up to 60 m (200 ft) maximum

That's a higher output impedance than I expected. I would then design a comparator circuit input impedance in the range of 24~100 KΩ to not significantly attenuate the input signal while reducing triggering on noise. An LM311 comparator should work well but it has an open-collector output. You'll need a suitable pull-up resistance to the 24V supply to see a 24V output.

I should've said "comparator" rather than "op-amp". (",)

I don't order from Digikey (I'm from the Philippines), but their website allows you to search for them. You need a single comparator, but the LM111, LM211, or LM311 are a bit complicated to use. The LM393 is my personal favorite. It's a dual comparator, but if you don't mine having an unused comparator, that's the one I'd use. It needs a pull up resistor but that shouldn't be a problem.

regards,

Vulcan

I suspect that the frequency alone is proportional to the flow rate.

I wonder if you might try a lower voltage input, and insert a fast switching diode like a 1N934 in the circuit.

The BRX specs call the input module 12-24Vdc, perhaps the 5 volt input with the rectifier might saturate the input.

At worst case, a zener diode could clamp the voltage to the max input voltage of the module.

So the output wire connects to a diode, followed by a zener diode rated for 5-6 volts and a current limiting resistor to the return wire, maybe a 500mA rated diode, starting with the max voltage you measure from the meter.

Save you designing a circuit and power supply for an analog chip, inverter or follower, op amp or Schmitt trigger circuit, which was my first thought. A long time ago, I used current source amplifier array, Norton amplifier current to voltage CDA which works well on single ended supplies like your generator output, LM3900, using that circuit to (attempt to) count words uttered into a microphone. It kinda worked. You would use a 24 volt supply, so your input would always have plenty of voltage. Still need power supply to run the chip, and you’d need to know the current that your generator can develop.