VoltAmp Meter: Input Circuitry 1

Since I want to start my exploration of lab test equipment with some relatively simple things, I am working on a basic two-channel (DC) voltage and current meter for bench use.  The idea is:

• Link the grounds of the meter and the Device Under Test.
• Report the voltage (relative to that ground point) of each channel.
• Optionally provide a return path for the test point, so I can splice it into the circuit and report the current passing through that point.

I'd like to aim for good accuracy and resolution (but not insane), and make the user interface simple and pleasant to use.  And, because I like small things, I'd like the whole thing to come out relatively compact.  Since the purpose of this project is not to learn about dealing with high voltages or currents, I want it to be able to handle something like +/-50 volts, and measure current up to 5 amps.  However, I want the device to not be damaged by accidentally hooking up some reasonable level of overvoltage, or by hooking it to a circuit that wants to draw more than 5 amps.  And finally I want it to do a decent job displaying voltages down at least to 1 millivolt, and currents down at least to tens of microamps.

Note that this is not really a great requirements spec... for example, "good ... (but not insane)" is not very specific!  Since I'm just playing at the moment, though, I'm not going to worry about that.

There will be five inputs to the device:  IN1(+), IN1(-), IN2(+), IN2(-), and GND.

Making the device simple and pleasant to use means (to me) that it should auto-range rather than have buttons or knobs for the voltage and current range.  And having decent resolution through four or more orders of magnitude does require multiple ranges.

To start with, let's consider the input protection circuitry, starting with voltage measurement.  We need to not let anything bigger than 50-ish volts through to the main circuitry, without messing up the measurement.  I considered putting something like a 10:1 voltage divider on the input, but I don't really like that solution because it adds an order of magnitude to the amplification may need to do later (for small voltages), which will cause unnecessary noise to enter the system.

A better approach, hopefully, will be to prevent the higher voltage from reaching the sensitive circuitry by blocking it or shunting it away.

The idea that struck me right away is to use a high(ish) voltage FET as a switch, and set up the gate to only turn the switch on if the voltage is acceptable.  This is complicated a bit by the fact that the voltage coming in might be negative... I'm not really used to thinking about negative voltages (one of the reasons this project is a good learning experience!).

I found some schematics on-line for a couple of multimeters, and watched a great YouTube video on this exact subject (multimeter input protection) by the mighty Dave Jones of EEVBLOG, and nobody seems to use a FET-based scheme.  For now I'll put the idea on the back-burner and fiddle with it now and then.

I finally narrowed the options down to two different schemes:

1. Use a varistor (e.g. a MOV) with an appropriate voltage as a shunt to GND.
2. Use a pair of zener diodes (one for positive and one for negative) as, basically, regulators ... again shunting the excess to GND.  The problem with this is that dropping high voltage through a zener causes massive heat dissipation.  Even 50 mA of current when burning off 100 volts will produce 5 watts of heat.  So, instead of passing the current through the zeners, what about using the zeners to trigger a low-drop/low-resistance shunt device?  A triac seems ideally suited for this (since the voltage may be positive or negative).

In either case, we don't really want to shunt the input directly to ground; the current through that path could be very high depending on the characteristics of the DUT.  So I decided to put a current-limiting resistor right at the input (which shouldn't do any harm).

The varistor approach is much simpler and from what I can tell it is the standard method.  Still, I think for my first test board I will put in circuitry for both and just populate the pieces I want to experiment with.  Here's the schematic as it stands now:

Next up: Input circuitry for the current measurement!