question about spd command

[plcdp 11]

Hello, I am trying to set up a tach to moniter rpm of a turning screw and I used the spd command. The problem is it comes up with an inaccurate number when I check with a infared tach gun. I suspect this is because I am not putting the right number in for a constant as I don't fully understand how this function works. My code reads SPD X5 K2000 D1 and I have gone up and down with the constant value. If anyone has any insight on this I would greatly appreciate it. Thanks, plcdp

[Crossbow 295]

The SPD instruction counts pulses over a period of time. In you case that is 2 seconds. So the number of pulses counted in 2 seconds is stored to D1, D2, and D3. D1 contains the number of pulses counted last time slice are in D1, D2 contains the number of pulses in the current time slice, and D3 contains how much time is left in this time slice. The number you put in for the constant depends on how long you wish to count. So it's really up to you. What speed are the input pulses coming in? And which PLC are you using? That will determine the maximum speed the PLC can record. JY992D88101 is the number for the FX Programming Manual II. That covers FX1S/FX1N/FX2N/FX2NC. JY992D48301 is for the older FX products, and JY997D16601 is for the FX3 series. All are in the Downloads section here.

[Goody 5]

You say you are using a 'tach' - how are you using this? What is the output from this tach?5

[plcdp 11]

Hello, I am using an FX-3U and the input pulses are coming in fairly quick. Basically the screw diameter is approx. 2.5 in. and there is a ring that goes around it with divots every 1/4in. or so. The proxy is on when not over a divot, and off when over the divots. Using a tach gun I measure 83rpm. So far using the plc I measure 17. Thanks for the help, plcdp

[Crossbow 295]

The high speed counter inputs are looking to count rising edges of the input. It should not matter that your system is normally on, it should still turn off and on the same number of times per rev. How long is the sensor staying off on each divot? If you're talking about 83 rev/min, at 4 pulses per rev, that's only 332 pulses per minute. That's well below the spec for the input points. if you are reading the number of pulses in 2000ms, then taking the number in D1 and multiplying by 30 should get you the pulses per minute. Then divide that by 4 since you have 4 pulses per revolution. That should leave you with revs per minute.

[panic mode 246]

i guess he is using hand held tachometer to verify rotational speed. some models allow contactless (infrared) pickup.

[plcdp 11]

Yes that is correct. I actually got this working yesterday, my problem was that I didnt have enough divots to pick up (only 6 per 1 rev.). I ended up putting a k4000 in for my time in the spd command and taking the number I was getting and dividing by 6 and then multiplying by 15 to get rpm. Only samples every 4 seconds, but for my application it works. Thanks for the help, plcdp

[Crossbow 295]

I fail to see how lengthening the time constant would have any bearing on the operation of the application, but the math is sound.

[plcdp 11]

The lengthening of time is because if I use a shorter time I can get a reading but because of the necessary math to scale it, the display is in increments of 10. I wanted it to be accurate within a couple of rpms. plcdp