Alalogue OR Gates

[yngndrw]

I am wondering exactly what an analogue OR gate does and how it works.

I couldn't find much about it, except this:
https://www.freepatentsonline.com/4419595.html

This site seems to show that it is just two diodes with the anodes connected together to give the output.

Is this correct ? Are there any other types ? Is there a better op-amp version ?

Thanks for any help,
-Andrew.

 

[Bob Scott]

What is your definition of "analogue or" ?

The link you provided treats the definition as the highest peak value of two or more analog signals.

It seems to me that the OR function is logical, not analog. However, if you are using analog voltage comarators, the output will be logic. See the specifications for the LM339 or LM393. These are analog voltage comparator ICs with open collector outputs. They can have their outputs wired in parallel. When the outputs are wired in parallel, they are in "wired or" configuration. (negative logic wired or)

That way each voltage comparator can be used to monitor analog levels in a system, and if any level is exceeded, the one wired-or output can be used to shut down the system.

 

[ericgibbs]

I am wondering exactly what an analogue OR gate does and how it works.

I couldn't find much about it, except this:
https://www.freepatentsonline.com/4419595.html

This site seems to show that it is just two diodes with the anodes connected together to give the output.

Is this correct ? Are there any other types ? Is there a better op-amp version ?

Thanks for any help,
-Andrew.

hi,

The way I read it is, there are two diodes connect to a single input say 'A'
One diode as a +V rectifier and the other diode a -V rectifier.

A second input say 'B' is wired to its own diode pair as per the 'A' circuit.

The outputs of the +V rect of 'A' and 'B' are OR wired and the -V rect outputs are OR wired.

The outputs of the two OR'd circuits are connected to a summing amplifier,
the output of the summing amp is the instantanous value of the input.

At the moment I cannot think of a practical application, but I am sure there must be one.!

What analog signals are you trying to measure or OR.???

 

[yngndrw]

Well the problem that I have is that I don't know the definition of an "analogue OR gate".

It was part of a forum thread regarding a stepper motor driver, in particular the way the drive morphs from 10 microsteps to full steps after the motor is running at x speed.

The drive generates a reference voltage proportional to speed. This voltage is sensed by an offset/gain circuit that becomes linear (from 0V) at 4 revs a sec and goes non-linear again (clamps at 3.3V) at 6 revs a second. It's output analog ORs with the sine-cosine reference. The sine-cosine reference is always there, it is simply swamped by the ORed full-step morph signal. So big deal, decel to zero instantly. 2 milliseconds later the reference reverts from full-step to sin-cos. Enjoy the 141% torque bump while it lasts for the 2mS. It's all analog.

It seems that there are two signals, a sin wave as the PWM reference signal and a linear signal which is 0v when it should be in 10 microstepping mode and 3.3v when it should be in full stepping mode.

I was wondering what kind of methods of analogue OR gates that there are so that I could decide what one would make sense in this situation.

Atleast one of the signals (And maybe even the sine reference signal aswell) is only positive, so that may make things simplier.

It effectively has to convert a sine wave into a square wave when the second input is high, but also has to blend between the two modes.

That's about as much as I can work out, thanks for your input.

 

[Space Varmint]

I think that is actually a comparator. It would act more like an XOR.

 

[yngndrw]

Would that be able to blend / morph a sine wave into a square wave though ?

I have a feeling that what he (The person who I quoted) said is simplified - It is a combination of processes which he simplified to "analogue or".

 

[audioguru]

You want an analog blend circuit (variable resistors) not a digital logic gate.
With variable resistors you fade one kind of sound into another kind and fade out the original sound at the same time.

A long time ago there were circuits that had their gain changed with a variable DC voltage. They were called a VCA (voltage controlled amplifier). Fade down the original sound and fade up the new sound and you have morphing.

 

[yngndrw]

Ah I see, that could work. I'm wondering if there's a more application specific option.
A VCA would blend between two different signals, I am wanting to blend between a signal and a square wave version of it.

Because of this I'm sure that there is some kind of simplification which could be done - The second signal can easily be derived from the first.

I assume it's like a variable Schmitt trigger - It blends between -1x amp and Schmitt
https://upload.wikimedia.org/wikipe...plifier.svg/750px-Inverting_Amplifier.svg.png
https://www.electronics-radio.com/a...l-amplifier-op-amp/op-amp_schmitt_trigger.gif

 

[audioguru]

You might be able to gradually distort a sine-wave into a square-wave.

 

[yngndrw]

I'm at a loss as to how I would do that. I have no experiance with signal processing like this. I also have no experiance with audio filters, which would probably be a massive help.

 

[audioguru]

Feed a sine-wave into an opamp. Adjust the gain of the opamp so that the output is not clipped. Then slowly increase the gain of the opamp so that the output is clipped into a square-wave.

It is the same as turning up the volume on a little radio until the sound gets more and more distorted.

 

[yngndrw]

That sounds perfect and relatively simple as well.

Thank you very much for all of your help AudioGuru, Space Varmint, EricGibbs and Bob Scott.

Edit: I've had a look around but can't find out how to make a variable gain amplifier. I can adjust the offset but that's all. Is there an easy way to do this without using some kind of variable resistor IC ? (As simple as possible.)