Signal Conditioning Problem 1

[lespaul]

I developed a circuit for a module that transforms several types of signals into a digitally compatible signal. I tested this circuit on a breadboard in January and it worked perfectly over the required range of amplitudes and frequencies. I had some PCBs fabricated about a month ago with the signal conditioning circuit and a bunch of digital components. The signal conditioning circuit does not work for a variety of signals now. Low amplitude, low frequency sine waves do not generate a clean signal like I was seeing with the breadboard. So I used the circuit on the breadboard again to make sure everything was the same and that circuit no longer worked! I originally tested the circuit in two locations and it worked perfectly in both locations. I tested it in two locations again and it failed in both locations. I'm completely dumbfounded.

 

[Skogsgurra]

lespaul,

Redflag my postings if you do not value them.

But before you do. Read my last post again. Try to understand what I am saying - and wy I am saying it. "Attacking" is not my usual behaviour even if I often is very open and direct in my comments. I think that other members can verify that.


Gunnar Englund

http://www.gke.org

 

[jimkirk]

lespaul,

How are you setting the internal gain of the AMP2?
You need a resistor connected between the gain set pins.
There is no resistor shown on your schematic.
Without a resistor, the instrumentation amp's gain is UNITY.

Also, I generally try to avoid the practice of having positive feedback element driving into what feeding it. The high frequency output impedance of the previous stage (the AMP02) is seeing some very high frequencies driven into its input through 201 Kohms. Doesn't sound like much, but I've seen issues with that topology.

You might try grounding the left hand side of the 1 K resistor and appling the input to the inverting input of the 311. Try adding a 1 K resistor to balance impedances and input bias currents as well. Good bypassing is a must.

If phase polarity is important, just swap the V+ and V- inputs to your circuit.

Glad to hear that there's a +/-15 volt power supply, but it does point out that the schematic you supplied doesn't have all the required information. (A common problem. I hate schematics that assume power supplies. Sometimes the software itself makes it overly difficult to include power on a schematic.)

Another question about your schematic...The LM311 needs a connection to ground for the output transistor, as logbook said. This isn't shown either. Just assumed.

 

[nbucska]

LESPAUL:
If I underrstand it, you want to convert a sine of variable
frequency and impedace to digital.

I think there is a better way. Can you please tell us
1.) min/max frequency your circuit must work with?
2.) Min/max. amplitude ?



Plesae read FAQ240-1032
My WEB: <

http://geocities.com/nbucska/>

 

[lespaul]

Max frequenecy is negotiable, but it must be able to do at least 30K. Min frequency is 24 Hz. Max amplitude is 50V and min amplitude is 50 mV.

This circuit must also be able to convert square waves with amplitudes .7-1.54 V, 0-4.2V, and 0-12 V. Same frequenecy requirements.

 

[nbucska]

zero volts signal can't be digitized.

I would suggest:
clamp the input with back to back Schottky diodes and amplify 10X. Digitize it with a comparator with about
100 to 200 mV hysteresis. Trigger with output a non-retriggerable single-shot with about 8 uS and use this to reclock a FF. The D input of the FF is the comparator's
output with an RC delay of about 10 uS.

Refinements:
1.) Change the gain, SS delay and/or hysteresis with the frequency
2.) Add a PLL.

i.e. Digitize first and then debounce it. Do not differentiate !!!



Plesae read FAQ240-1032
My WEB: <

http://geocities.com/nbucska/>

 

[lespaul]

I forgot about one of the signals. There will be a 700 mV p-p square wave signal that will be riding below vehicle voltage which is about 12-14 V. If I amplify that signal, it will run an amp right to the positive rail and never leave. Thats the reason for the differentiator.

 

[nbucska]

Use AC coupling with long enough time constant so it won't
differentiate.

Plesae read FAQ240-1032
My WEB: <

http://geocities.com/nbucska/>

 

[logbook]

Of course it is not a "differentiator" anyway, since the HF gain is simply R3/R1. C1 could stand being a bit bigger to make the LF signals stronger.

Put a scope on the output of the first stage and see how much noise is present. If the noise is bigger than the hysteresis in the comparator you will of course get noise spikes on your digitised waveform. So in this case you either have to reduce the noise or increase the hysteresis.

PS: I think scogs was having a go at me rather than you ... and I could see piece-wise true parts in his criticism (of me)

 

[felixc]

With R6 at 1K, it means that the comparator is switching currents of 15mA. This isn't necessary, and such transitions introduce noise in the rest of the circuit, especially since you're trying to detect low voltages.

 

[LionelHutz]

If the output of the LM311 is a NPN transistor connected to ground then the LM311 can't pull it's output negative so your positive feedback circuit that's referenced to ground (circuit built to work with input signal centered around ground) just won't work. The output MUST be capable of going negative for the positive feedback you're using to work.

Maybe I missed something in the quick look at the datasheet.