Electrical:
- Working Flow Rate: 1 to 30 Liters/Minute
- Max current draw: 15mA @ 5V
- Output rise time: 0.04us
- Output fall time: 0.18us
- Flow rate pulse characteristics: Frequency (Hz) = 7.5 * Flow rate (L/min)
- Pulses per Liter: 450
So what do you think is happening? Too much resistence on my sensor extension? Measures .07 ohms on one cable and .03 ohms on a large wire cable.
1) It looks like you're using the N+2 pin a 5v tolerant input digital signal.
2) The rise and fall times are "slow". That should make it not too hard to fix.
3) I'm not sure I understand the max pulse rate per second. Is it 7.5*30 == 225? If so, what does the "pulses per Liter" mean.
The "DC" resistance of the cable isn't a problem at 3feet and with a 5V 15ma driver.
It's the "AC" impedance (the "resistance to a 'change in voltage')) that could be the issue. The rise and fall times are slow - which is good. Still, I'm wondering if there's too much over and undershoot on the signal input of the N+2 with the cable.
Any time a signal changes value, the AC impedances effect the waveform.
At 3 feet, you have approximately 6-10ns of delay in the signal from the start to the end of the cable. If the signal rise and fall times were faster than ~10-20ns, then you'd get very noticeable reflections, overshoot, undershoot, ringing - all of the bad stuff.
Your signal has a rise time of 40ns and a fall time of 180ns. So, definitely anything past an ~10 foot cable would have to be treated as a transmission line.
How about the following for a quick and dirty test?
Get some cat3/4/5/6 Ethernet cable, and use that to hook up the signal. Ethernet cable has an impedance of 100ohms.
The traces on the N+2 PC-board likely have an impedance between ~70 to ~90 ohms. So, not as good as a 75ohm coax, but better than regular wire, or a ribbon cable.
See if you can get a resistor between 1K to 330 ohms (620ohms if you have your choice). Connect that between the signal and the ground.
Very roughly, the digital output of the sensor has an impedance of ~5ohms to ~40ohms.
The trace impedance on the PC-board is ~75 ohms.
The input impedance on the pin on the N+2 is ~1 mega-ohm.
A single wire has an "infinite" impedance (not really, but we're talking quick and easy to remember).
When the output signal goes from the sensor to the input on the N+2, all of those different AC impedances effect the rising and falling edges.
With a 3 foot cable, 0-to-5V output, 40ns rise time, I'd guess that the signal would have noticeable overshoot, undershoot, and ringing, but it would still "work".
You should also guard against too much overshoot on the input signal, and therefore damaging the N+2.
After you do first do the test above, try adding an 10ohm to 33ohm resistor in-line (in series) with the signal. Put the resistor right at the N+2 connector. That limits the current from the overshoot.
It also limits the current going into the N+2 input - which then slows the signal's rise/fall times - which may reduce the overshoot and undershoot.
It also effects the overall trace impedance(increases it) between the connector and the N+2 (it'll be the pc-board trace impedance + the resistance). "Basically", the AC impedance of a resistor is the same as the DC resistance of the resistor.
If you can get/borrow a scope and probes, that would help. The 40ns rise time is equiv to 25MHz. You should use a scope and probes that have a bandwidth of ~100+MHz. You could still use passive 10X probes. Just make sure they are rated for the needed bandwidth.
BTW, I did a super quick search on ebay:
I found: (Analog non-storage scope)
Tektronix 2245A Oscilloscope 100MHz 4 Channel with 2 Tek P6133(150MHz) 10X Probes
Buy it now, under $250 with shipping.
I'm sure there are many other deals and options.
BTW, you're not going to get a digital storage scope with the needed bandwidth sampling rate and bandwidth for a low price.
So, to use the above scope, you'd want a steady flow so that you get a steady/consistent frequency. You can use your digital camera to take a picture. Turn off the flash, and use either a small tripod, or some books.