I have an optical probe which is NPN (or PNP) - whichever it is, it’s the wrong one. The probe produces a current between +VDC and signal output - what I need to take out to my data aquisition hardware is a measurable voltage above ground. Since I have a differential ADC, I can use the probe as is, with the downside that the BNC fittings between components are all at a positive potential, and I’d have to be careful when hooking things up. I would rather have those fittings connected to ground, which would not only be safer, but would enable the probe to be used with a single-ended ADC.
Short of buying the opposite (PNP vs. NPN) version of the optical probe, is there an easy way to convert the output signal to get what I want?
I’ve never heard of a probe of any sort being called PNP or NPN. Those are transistor configurations. In any case, it shouldn’t matter. Connect the most negative lead of the probe oputput and connect that to the ground terminal of the ADC and connect the most positive probe lead to the +V terminal of the ADC. As long as the voltage differential of the probe’s output falls within the range of the ADC, you’re good to go.
I don’t understand the PNP/NPN terminology myself. As has been explained to me, the probe is a current producing device. To generate a voltage, I need to put a resistor between signal out and +V, so when the probe is closed I measure that difference. If this worked between signal out and ground, that would be the end of my problem. As it is, this is the signal that I would use if the offset differential is my only solution, as you proposed in your post. Unfortunately, that doesn’t solve my problem, though. The probe is self contained with a 9V battery as supply power - meaning that the BNC fittings at each end of the cable that connect the probe to the ADC will be at a higher potential than the surrounding hardware, and hence I would have to be very careful about letting them come into contact with anything. If I could use the ground as one end of the differential, that would keep all connection hardware and chasses at the same potential.
The other problem is accomplishing the objective of making it useful with single-ended gear, which necessitates comparing a single output signal against a common ground, which would be the single-ended ADC chassis or its power supply negative. (I’m not the only person who will use this probe, and others use different data aquisition systems).
The only solution I have found so far is to use a NO relay, powered between +V and signal output, which when closed sends +V back to the ADC. When I do that though, the signal isn’t useable because the output isn’t tied to ground when the relay is open, and it is noisy.
I’m a mechanical guy - this electronics stuff is a little infuriating. Any other suggestions?
Might help if I could see a schematic for the probe. Maybe I can find one if you’ll provide the make and model number.
If nothing else, at least give a better description of all the outputs of the probe. I’m having a bit of difficulty understanding why it wouldn’t work if you connect signal out to the ADC ground terminal, and +V from the probe to +V on the ADC (or vice versa if signal out is more positive than +V). The resistor shouldn’t be necessary, unless the device relies on a current flow for accuracy; perhaps it needs a load to maintain linearity, liek some amplifiers do.
I’m not familiar with optical probes. Do they produce a binary (on/off) signal or a continuous/analog signal? In other words, does it tell you if the light is on or off (the former) or does it produce a signal proportional to the intensity of the light (the latter)? Do you know the make and model of the optical probe you’re using?
Without seeing specs I can only offer the following suggestions:
I assume the V+ you’re referring to is the ADC’s positive supply rail. If you connected the resistor to the positive terminal of the probe’s 9V battery (instead of the ADC’s positive supply rail), then you would have a floating voltage source. Or you can just use an external battery, which would probably be easier. But there are two problems with this approach: 1) A battery is not regulated, and 2) a battery discharges after a while. A zener or linear regulator (78XX series) might work O.K. to resolve problem #1. If #2 is a concern, consider using a separate, regulated DC power supply powered from a wall outlet. (The output of a power supply is almost always isolated. For “best” isolation, choose a linear supply over a switching supply. To increase isolation even further, consider plugging the supply into an isolation transformer.) For a single-ended ADC, simply connect it up and you’re done. For a differential ADC it’s also a matter of simply “hooking it up,” though you’ll probably want to install a 100K resistor between the negative input and power supply ground. This is because the input circuitry of a differential ADC needs a bias current path.
Based on your description I would assume the probe has an “open collector” NPN output, and that it is on/off (not analog/continuous signal). If this is true, why not have it drive a PNP transistor? Connect the emitter of the PNP to the ADC’s positive supply rail (V+). Connect the base of the PNP to a 1K resistor; connect the other end of this resistor to the probe’s signal output line. Connect the collector of the PNP to another resistor; connect the other end of this resistor to the probe’s common and ADC common. Connect the collector to the ADC input.
Are you looking at a pulse train or DC signal? If the former, you might try sticking a capacitor between the probe’s output signal and the ADC input. You would still need to use a resistor connected to V+.
If all you’re concerned about is the risk of having V+ on one of the ADC’s input lines, consider doing this: instead of using one resistor between V+ and the probe’s out line, use two resistors in series. Then tap into signal voltage across the resistor that’s connected the probe. That way, if one of the signal lines accidentally comes in contact with ground, you won’t blow anything up.
It sounds to me like the probe has an open collector output stage, as Crafter_Man suggested. If that’s the case, then the probe’s output has to be connected to the ADC’s +ve rail with a big (say 470k) resistor.
If the ADC is “single-ended”, then you need to tell the ADC or whatever is processing the ADC’s ouput to invert.
If the ADC accepts a differential input, then you can connect the ADC’s inputs to the probe’s output and ground, and do the inversion somewhere else, as above. Or, swap the differential inputs. Or, you can connect the ADC’s inputs to the probe’s output and the ADC’s +ve rail. Or do this and swap the differential inputs, and then invert elsewhere.
Note these none of these options involve connecting the BNC shells to the ADC’s +ve rail.
Desmostylus: I thinkFuji is tapping the signal directly across the pull-up resistor, and my responses assumed this was case. Is that the way you interpreted it? Now I’m wondering if I interpreted it correctly… at any rate, if he is tapping the signal directly across the pull-up resistor, I don’t know why; if it’s a digital signal, I would measure between the probe’s output and common/gnd.
Digital signal. It’s a Keyence PZ2-61, micro optical sensor, NPN output (100mA, 40V max). I found the instructions online - driving a voltage load requires a 4.7 kOhm resistor between +V (and the +V I am talking about is the positive terminal of the 9V battery in the probe enclosure) and voltage output. Upon further reflection, it occurred to me that a double pole relay would work, as I could send battery voltage out on the NO side, and tie it to ground on the NC side to eliminate the noise. The problem is that the whole mess has to fit in a tiny little box not much bigger than the 9V battery, which precludes using the relay since I’d have to put it on a board for a robust installation, and space is at a premium. The probe location is quite remote in comparison to the other gear - I can’t run power to it, hence the battery. From the probe, a BNC cable runs to a telemetry system on its own 24V supply, and the ADC is on the other side of the radio link.
I can use it as is, since I can use a differential signal with my gear, and invert it or do whatever else I want with it. The problem is making the probe compatible with less sophisticated equipment (meaning it must produce zero V WRT ground when off, and a positive voltage when on) - I should have purchased the PNP version of this probe, but for now, I need to make do with what I’ve got.