Help with opamp

I'm new to Arduino and electronics in general. My goal is to make a quickshifter for my motorcycle. I will be using a strain gauge curcuit to kill the ignition when the gear lever is pressed.

Currently I'm trying to figure out how to amplify the signal from a wheatstone bridge. I made a test circuit using three 10k resistors and a potentiometer. I can get the bridge to almost balance using the pot and check the voltage with my multimeter.

My next problem is amplification. I have a LT1077 opamp but am not quite sure how to connect it up. Do i just connect each side of my strain gauge circuit to the -IN and +IN inputs on the amp and use the V+ and V- pins for 5v and ground from arduino, leaving the OUT pin for the amplifed signal? I assume the two Vos TRIM pins are for setting the gain.

Link to the datasheet is here: Redirect Notice

I'm curious, what is a quickshifter.

A device for shifting (gears) quickly?

On a bike, you can upshift gears without using the clutch. All you have to do is unload the transmission by shutting the throttle for a few milliseconds. You can then flick the gear lever up to engage the next gear and open the throttle again. On a race bike, you can eleminate the need for shutting the throttle by using a quickshifter to cut the power for you. It saves a few milliseconds but every little helps in a race.

HX711 is made just for this. Built-in instrumentation amplifier with programmable gain and differential inputs, and a 24 bit ADC.

http://www.icstation.com/icstation-hx711-weight-sensor-module-p-5251.html

I see. The reason I bought the LT1077 was because I saw it used in another similar project online. Since i started trying to put this project together, I have realised that there isn't much info available for a beginner about this chip. I will order the hx711 and see if I have better luck.

I notice on the spec sheet it states differential voltage as +-40mV. Does this mean the input from the weighing sensors must be balanced to within 40mV? Sorry if this is a stupid question...electronics are a little more complicated than i was expecting lol

I notice on the spec sheet it states differential voltage as +-40mV. Does this mean the input from the weighing sensors must be balanced to within 40mV?

The differential voltage is simply the voltage from your wheatstone bridge.

It's a voltage difference because nether voltage is referenced to ground. For example, a regular-old multimeter measures differential voltage because the black lead doesn't necessarily have to be connected to ground.

You probably won't know the voltage from your wheatstone bridge in your application until you experiment. The HX711 has adjustable/programmable gain so hopefully you can find a gain that works for you.

I haven't read the datasheet for the chip and I'm not sure what "40mV" means. It's an analog voltage so it's not fixed at 40mV and the thing has variable gain.

I have a LT1077 opamp but am not quite sure how to connect it up. Do i just connect each side of my strain gauge circuit to the -IN and +IN inputs on the amp and use the V+ and V- pins for 5v and ground from arduino, leaving the OUT pin for the amplifed signal?

A basic [u]op-amp differential amplifier[/u] requires 4 resistors and (like any op-am based amplifier) the gain is determined by the resistor ratios.

polymorph:
HX711 is made just for this. Built-in instrumentation amplifier with programmable gain and differential inputs, and a 24 bit ADC.

http://www.icstation.com/icstation-hx711-weight-sensor-module-p-5251.html

Too slow perhaps, I suspect something capable of several kHz bandwidth would be more
appropriate, the HX711 has 10 to 80Hz sample rate, so probably slower than human
reponse time for settling...

I have a diagram of the LT1077 used for another wheatstone bridge circuit. It's on my pc at home so will have to wait until tonight to have another look. I'll post it up here and hopefully someone can make sense of it for me.

Thanks for all the replies by the way. This seems like a very helpful forum for a beginner!

The HX711 has programmable gain of 32, 64, and 128. But it is wired as one set of inputs has a gain of 32, and the other set can be 64 or 128. Channel A has a gain of 64 or 128, Channel B is a gain of 32.

Yes, +-40mV means your two inputs must be matched within that. 0mV offset will be mid-scale, -40mV will be zero, +40mV will be 2^24-1 or about 16.7 million.

Using Channel B with a gain of 32, it'll be +-80mV full scale.

Expect most of the lower bits to be noise. But you should still be able to get at least 18 bits of resolution out of it.

It is easier to balance the bridge if you use a four wire load cell.

As for speed... really? 1/10th of a second too slow for human response time? How about 1/80th of a second?

http://www.humanbenchmark.com/tests/reactiontime

Or try it with your own ruler and see this chart:

Or just use a piezo pressure sensor and a comparator.

sean61:
I notice on the spec sheet it states differential voltage as +-40mV. Does this mean the input from the weighing sensors must be balanced to within 40mV?

It's 40μV, not 40mV (EDIT: for LT1077). That's the maximum voltage across the differential inputs when the output voltage is 1.4V (under specified conditions). As you would need to connect the operational amplifier as a differential amplifier (see DVDdoug's post and page 11 of the datasheet), you will not be connecting the output from the Wheatstone bridge directly to the operational amplifier inputs. Therefore you can ignore the 40μV.

Would it not be better to use a microswitch to detect when your gear lever is pressed?

It is 40mV, not 40uV.

From the datasheet:

The input multiplexer selects either Channel A
or B differential input to the low-noise
programmable gain amplifier (PGA). Channel A
can be programmed with a gain of 128 or 64,
corresponding to a full-scale differential input
voltage of ±20mV or ±40mV respectively, when
a 5V supply is connected to AVDD analog power
supply pin. Channel B has a fixed gain of 32. Onchip
power supply regulator eliminates the need
for an external supply regulator to provide analog
power for the ADC and the sensor.

I do think that a microswitch or piezo or some other form of pressure switch would be simpler than a load cell.

polymorph:
It is 40mV, not 40uV.

Thanks, I thought sean61 was referring to the LT1077.

I could use a micro switch but according to guys who have used them on their bikes, they wear out fast from all the vibrations and cause problems. All the commercial products now use strain gauges so I'm guessing there's a reason for that. Also, I have gone to the bother of buying strain gauges and an arduino to make this work so I'm gonna stick with it. I'll get there eventually and will learn loads in the process!

I think I'm starting to understand what's happening with the opamp now anyway. I just couldn't figure out what all the resistors were for and I'm not used to reading circuit diagrams. I'll post up the info I have at home and I think I should be able get my head around it now.

Here is the diagram of a circuit which I know works. It seems a bit different to the ones I can find online because most have the +IN pin connected to ground. And why is there both 5V and 13.8V at one end of the bridge?

Usually one of the excitation connections to a load cell is connected to ground.

I expect the 13.8V is explained in text that accompanies that circuit diagram. I guess it is saying that you can connect either +5V or a fully-charged car battery at +13.8V to the load cell. However it is quite likely that 13.8V is greater than the maximum rating of a load cell.

Hi,
The biggest problem with strain gauges and devices that output mV and need amplification is the ignition noise that could get amplified with it.
If you want high frequency response, then you will be asking for problems.

Tom.... :slight_smile: