Problem: Schmitt Trigger not giving ON/OFF output

[InterlinkKnight]

yeah. Tell us who your suppliers are and we'll try to identify an appropriate op-amp.

I live in venezuela and here things are complicated. I can only buy what is available on electronic stores near by. So, I can only buy basic things with limited options, and many times even the basic things (like resistors) are hard to find.

 

[(*steve*)]

OK that makes things very tricky for you.

Perhaps if you can describe exactly what you are trying to do we can figure out some alternative methods.

One thing which would help is to add a negative supply rail. You might be able to get away with something as low as 1.5V. This still isn't perfect, but it will allow the 741 to pull the output significantly lower.

to try this, just disconnect pin 4 and connect a 1.5v battery with the -ve terminal to pin 4 and the +ve to your 0v rail (the negative terminal of your 5v supply).

 

[BobK]

If a Zener diode is too slow, a '741 is certainly also too slow.

Perhaps I am dense, I still don't see how a Zener diode can force a voltage to zero.

Bob

 

[InterlinkKnight]

What I am trying to do is a tachometer for arduino, but the tricky part is that it also use a real car cluster gauge. They work separately fine but when join together, the cluster adds some voltage and arduino don't recognize the LOW anymore.

I am surprise it works at all, separably. The cluster do the measurement and move the needle gauge very accurately, and arduino by itself also measure RPM very good. But when I join them, the LOW is not longer 1.6V, but 2.26V. And arduino interprets 2.26V as HIGH, so never sees LOW.

The RPM sensor is IR LED + IR receiver, BTW.

I experiment a lot and since they work great by themself, I put another op-amp to make 2 different Schmitt Trigger, one after the other. I suppose could also work 2 Schmitt Triggers in parallel, but what I did is put them in series. When the first Schmitt Trigger goes LOW with the cluster connected (2.26V), the second Schmitt Trigger goes LOW because is set to have a 3V threshold.

So, problem was solve, but not in the way I thought would be. And maybe too expensive since I use 2 op-amps, but is what is available to me.

But I am thankful for all of your replies because I learn that what I needed was a rail-to-rail op-amps, and with low voltage range. I will take notes on that.

Thanks.

 

[CDRIVE]

If a Zener diode is too slow, a '741 is certainly also too slow.

Perhaps I am dense, I still don't see how a Zener diode can force a voltage to zero.

Bob

Bob, you read my mind. I was just about to post that the 741 is not considered to be the AA Fuel Dragster of the OpAmp world.

Chris

 

[mgrass]

You can make a Schmitt using an inverter cmos gate. As was posted, the 741 cannot swing withing 2 volts of the rail. So using ground and a 5 volt supply, you only have a 1v swing. Depending on the hysteresis of your circuit, you may not be able to overcome the feedback signal. https://en.wikipedia.org/wiki/Schmitt_trigger has a good explanation of the schmitt trigger. Look about halfway down under op-amp implementation. They show both an inverting and non-inverting circuit. They site also shows you how to design it using discrete transistors.

 

[CDRIVE]

If Kris was still with us (RIP) he would have given you a complete schematic, parts list, list of suppliers and a whole lot of stuff I've probably missed. Above all and particularly because of your component procurement issues, it would have been completely DISCRETE! If you couldn't get the transistors he recommended he would have sent you out mining minerals to make them.

Damn, I miss him!

Chris

 

[CDRIVE]

After I posted, it hit me that a very common 555 can be used to make a very simple schmitt trigger. Do you have any? If not, can you get them? It will be fast and will give you more than acceptable TTL logic level output levels. IE, .. Lo ~= 0V Hi~= 5V. This will be especially true if it's the CMOS version. That said the common Bipolar version will more that fill your requirements and be recognized by any chips in the TTL family..

Chris

 

[InterlinkKnight]

All good ideas, but I already solve it using 2 op-amps, one for arduino and the other one for the car cluster. Was not an ideal solution, but works.

But I will keep it in mind for the future.

 

[bpark1000]

An op-amp is not the proper device for comparison. Op-amps are optimized to operate in the linear region, where all transistors are operating in the linear region. 741 really needs +/- 15V to function. When an op-amp "runs into the rail", its transistors saturate, from which recovery is slow. Some (older, especially) op-amps can even "reverse polarity" leading to "latch up". much better is to use a comparator. If you are in the "old camp", an LM339 will work. It is a quad comparator. The outputs must be pulled up to Vcc with a pullup resistor (4.7K works). You can set up the input resistors in the usual way, as you did for op-amp. The output gets to within Vce(sat) of ground which should be 50 to 100mV. One advantage of the "open collector" output is that you can "wire or" the outputs to make, for example, a window comparator.

Beware the LM339 and other quad comparator pinouts! They are not the same as for op-amps!

Regarding your confusion about offset voltage: Think of a tiny battery being in series with one of the inputs. The voltage the "ideal" input "sees" is offset from what you are inputting. This voltage ranges from 0.7V for a discrete kluge "op-amp" to handfuls of microvolts for fancy factory-trimmed modern parts.

There is also an offset current. This can be thought of as sneaking in and connecting resistors between an input and either Vcc or ground. This causes currents to flow in the resistor network you put on the inputs, producing voltage offsets. High values of resistors in those networks are most affected.