Tapping into a digital signal wire without interfering...

[jimdeane]

I have a piece of equipment that produces a digital signal pulse that feeds into a data aquisition unit. The DAQ has some low-level hardware processing that only sends cumulative packets to the high-level hardware, so I can't write a program to access the individual data pulses.

I would like to create an interface that would sit between the sensor and the DAQ, and would let me use that digital pulse for other things (in this case, to create a brief beep when a pulse arrives).

My first thought is that something like a FET would let me tap into the digital signal without adding or taking away much from the signal. I've used FETs before to detect and quantify static charge, so my impression is that it would drain so little current from the signal wire that it would not affect the signal at the DAQ.

In this particular application, I would then take the output from the FET as an input to a 556 timer set up as a one-shot and as an oscillator (so each pulse in would result in one discrete "beep").

However, I've not designed a circuit like this before, so I'm not sure what "traps for young players" are waiting for me. Am I on the right path, or is there a better way to tap into this circuit?

[suicidaleggroll]

You haven't told us anything about this signal yet.  Voltage, frequency, drive strength, line length, impedance, etc.  If this is just a simple low frequency, slow rise time TTL pulse, you can do pretty much anything you want without causing a problem.  If it's a controlled impedance high speed pulse train, then you probably can't.

As blueskull mentioned, a buffer IC would probably be a much better choice than a FET due to gate capacitance.  Eg: The SN74LVC1T45 can run at 1.65-5.5V and up to 420 Mbps, depending on the voltage levels you're translating between.  Run the signal through one of those and then do whatever you want on the output.

[jimdeane]

The digital signal is probably 5v/0v, but I need to take it to my home lab to put it on a scope and watch the signal. The maximum signal frequency is about 3500 Hz.

I'll be testing it (maybe a couple of weeks, but eventually) to get a better idea of the signal parameters.

As I wrote this, the company who makes the sensor offered to swap it for the updated version that produces the desired beep, so I may not need to do this -- but I will probably still work on it.  (For the curious, the sensor is a Vernier VRM-BTD radiation monitor, and the interface is a LabQuest.)

[wraper]

The digital signal is probably 5v/0v, but I need to take it to my home lab to put it on a scope and watch the signal. The maximum signal frequency is about 3500 Hz.

I'll be testing it (maybe a couple of weeks, but eventually) to get a better idea of the signal parameters.

As I wrote this, the company who makes the sensor offered to swap it for the updated version that produces the desired beep, so I may not need to do this -- but I will probably still work on it.  (For the curious, the sensor is a Vernier VRM-BTD radiation monitor, and the interface is a LabQuest.)

Just use usual 10:1 passive scope probe or any CMOS IC input you want. At such frequencies you do not bother when it is digital signal.

[DaJMasta]

As mentioned, 3.5kHz and normal TTL is quite robust and there is a lot of leniency for tolerances when dealing with digital, so almost any normal method is permissible.  Even a 1:1 scope probe should be no problem.  Feeding it into a buffer or driving a few directly would probably both be fine - the capacitance on a fet is not going to be so large as to interfere with something that slow.  Provided your source has enough sink current, even a BJT would likely be fine in there.

[tggzzz]

As mentioned, 3.5kHz and normal TTL is quite robust and there is a lot of leniency for tolerances when dealing with digital, so almost any normal method is permissible.  Even a 1:1 scope probe should be no problem.  Feeding it into a buffer or driving a few directly would probably both be fine - the capacitance on a fet is not going to be so large as to interfere with something that slow.  Provided your source has enough sink current, even a BJT would likely be fine in there.

You don't have enough information to assert that.

The two key parameters are the rise time and when other signals change relative to the one being probed.

The rise time defines the highest frequency; the period is irrelevant.

If you add too much capacitance (e.g a *1 probe on a clock) then hold or setup times could be violated.

Often the best probe for a digital signal is a resistive-divider low-impedance Z0 probe. Since it is merely a 450ohm resistor feeding 50ohm coax, it is also easy to build at home 😀 And the 450ohm resistor can often be left installed in the circuit.