I'm wanting to experiment with circuits for signal conditioning, like precision rectifiers and auto ranging to best utilize the span of A/D converters. Is there a name for the type of circuit that multimeters use to select ranges or auto-ranging?
Also I'm looking for an oscilloscope that would let me look at the signal in and signal out to make sure I'm getting the signal my circuit is supposed to provide. I don't need real high frequency, mostly working with microcontrollers and 60hz - ~100Khz or so, but more is better for future projects I may decide to do. Space is primary importance of the Oscilloscope I need now, either a pocket scope, a usb scope, scope meter, or the largest would be something like the Rigol scopes. Are there any good open source or other small oscilloscopes? I wouldn't mind getting a usable pocket scope for now and getting larger scope later, I don't expect a pocket scope to do everything I want an oscilloscope to do.
The "ranging" I was referring to wasn't for the oscilloscope, the oscilloscope is to aid in adjusting the filters for best smoothing versus responsiveness of the signals I'm wanting to condition.
One idea I have is to use an analog multiplexer with a voltage divider input and the output goes to an instrumentation amp. If the analog input is 0-5V, range 1 would be 0-5V, range 2, 5-10V... using the common mode rejection to output 0-5V by increasing the reference voltage to different ranges. The idea is I'm wanting to get the readings in the A/D range to use the bit counts. For example an 8 Bit A/D could output representing 0-255 Volts or
0-255 millivolts. I'm wanting to have the signal conditioned for different ranges depending on the range selected or what I'm measuring.
So do the analog multiplexers work bidirectional (mux- demux)? Can they be used to swap different resistance values in an op-amp circuit? For example,
2 analog multiplexers, R1 connects 1 lead in Analog Mux 1 IN 1, other lead to Analog Mux 2 IN 1, R2 goes to Mux 1 & 2 IN 2, R3 - IN3, etc. Then select the gain by selecting R1, R2, R3 ... . Or would that work to select the gain resistor in an instrumentation amp?
...And adjustable offset, or adjustable attenuation? ...Or all three!?
An analog mux is simply an array of CMOS switches. Think of it as an nPmT (e.g., CD4051 = SP8T) switch with rather high (and somewhat nonlinear) contact resistance.
Out of 8 bits, you won't have as big a worry about switch nonlinearity, but you will have to be aware of it for best results.
If you're doing ranges by some multiplier sequence (probably not decades, but 1-3-10-... and 1-2-5-10-... are popular), what you do is leave the input circuit at maximum gain (0-255mV?) and switch the input source along a tapped resistor divider.
A decade divider for instance might start with 1k at the bottom, then 9k,
90k, 900k and 9M for a 255V, 10Mohm divider as a DMM might have. You also have the option of connecting the inputs directly (the divider is out of circuit), providing a "hi-Z" range. Obviously, with ample protection circuitry, in case one were to connect 255V to it while in the 255mV range.
For best results, you'd use a "hi-Z" buffer, such as a precision JFET op-amp follower, with the mux in front, so that no current is drawn through the mux and there's no nonlinear switch resistance error.
Disconnecting the divider probably requires a switch at the very bottom, connected to ground through a servo amp to eliminate switch resistance.
BTW, why 255mV or 255V? That's a rather ungainly range for any purpose -- AC line peaks go to 170V or 340V; most ranges go to 500V or 1000V; most displays are 199, 999 or 1999 range (and so on). It might feel comfortable to know that your digital number is always about something (give or take a few milivolts), but you go through an awful lot of trouble to attempt to ensure that: oddball reference voltages, suboptimal ADC ranges (255mV into the ADC's VREF is probably less than it's spec'd at, and will magnify the effects of offset voltage of the internal comparators, making the measurement that much worse), analog calibration, etc.
It's better to use a standard reference (which might be 2.50V, or 5.00, or even 2.43 or something funny like that; point being, as long as it's stable, it doesn't matter), design the analog front end so it's guaranteed to just exceed the range (you can design it with 10% resistors if you want, just so long as the important ratios are matched -- high ratio-tolerance resistor pairs are actually available), and do exactly one calibration step, in software, multiplying the digital value by a coefficient before display.
The main downside is if you want to have analog protection of some sort, for example an 'over range' indicator using a window comparator or the like. Then either your input gain needs to be calibrated anyway, or the comparator needs to be programmable by the same coefficient (which is just as easy, but tedious, and the DAC takes up more I/O resources).
You can, but I wouldn't recommend it, unless you can ensure that R >> Rsw so the error due to Delta(Rsw) is also small.
Don't know about you but building a 10 MHz scope would be a lot more fun than buying one. If you charged minimum wage for your time it would be a lot more expensive than buying one. Pushing it to 100 MHz would be even more fun and challenging, but may be past the edge of my reach. Jan P. and the others here should be able to coach you through it though. It depends a lot on how you want to approach it.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.