Analog fuel injection circuitry help (and Speeduino chat)

[MarcWolfe]

Can anybody give me a reason why this circuit wouldn't work as expected? All info needed should be in the picture.

[MarcWolfe]

Just now, James Evens said:

Can you show the calculations for the resistors?

What is the observed behavior?

I haven't observed anything yet, and won't have an oscilloscope to observe with once I do build something. I went off this 

from this video 

But obviously I added resistors. Should be a gain of 3x if it goes the way I want. 

[MarcWolfe]

Because the inputs to it are fixed and divided to be so. I'm asking because I'm not sure if the OP-amps feedback will change that. I don't think it will, if I properly understand how OP-amps work. 

[MarcWolfe]

3 minutes ago, James Evens said:

 

If you don't get it use the simplification (no current flow in and out, it does what ever it can to keep both inputs equals) and think about what the OpAmp actually does.

 

Did that, now I'm here.

[MarcWolfe]

4 minutes ago, James Evens said:

How can the voltage (red dot) be constant?

Because OP-amp with feed back resistor. Sounds like my concerns are unwarranted. Well, thanks I guess. Man, fuel injection with a 555 timer, couple pots and couple OP-amps as the entire brains of the operation. Wasn't expecting it to be so simple.

[MarcWolfe]

21 minutes ago, James Evens said:

This point is a 1:1 copy of the input signal (ideal opamp). Since this points changes there the "1.6V" from the voltage divider also changes and that's probably the point where your calculations, you haven't shared, went wrong.

 

If you keep this circuit there are two things which need tobe fixed:

1. resistor values

2. right opamp or a negative supply voltage.

I never said anything about calculations that went wrong. I'm here because I'm pretty sure I'm done with this circuit. Peer review basically.

[MarcWolfe]

2 minutes ago, James Evens said:

I tell you this won't work like you expect.

Are you under the impression that I'm working with anything A/C? Because I'm not. Just stop.

[MarcWolfe]

If you want to be sure to know how it goes https://www.youtube.com/channel/UC8oyY_JJKEDH79pSwBXjv8Q/videos sub and I post small bits here https://www.facebook.com/marc.wolfe.568

[akio123008]

On 5/20/2021 at 4:31 AM, MarcWolfe said:

I never said anything about calculations that went wrong

Have you simulated the circuit and looked at what happened? 

 

You have a voltage divider (the 5V - 0V thing at the bottom) and you've written 1.6V next to it. That's not correct. A current will flow from the opamp output into your voltage divider, messing up your standard voltage divider calculation.

The voltage in that node will not be 1.6V, in fact it will change with the input signal.

On 5/20/2021 at 3:15 AM, MarcWolfe said:

I'm asking because I'm not sure if the OP-amps feedback will change that.

So there's your answer to that; yes it will. I built the circuit for you in an online simulator so you can see for yourself:

http://falstad.com/circuit/circuitjs.html?ctz=CQAgjCAMB0l3BWcMBMcUHYMGZIA4UA2ATmIxAUgpABZsKBTAWjDACgBDEQm8YlEGjx8BxcEhZIw8eOGg5ICGmH4EwS1nIUJCMsNgx5sYQshlsAToPyC8wnreFVp8S914o77kNkJOKrlYOvsJgGKYhUAFwbsF+IEw0wpHO5gBK3mGmTA5ZUfQ0VDB4SCjQNEhF0AhsAO7eQg1ekHXWwp7tNioCLRmJyfFMxBHxRbRUlVEwNQDm4OE+8XmFYy31jd2CGAKba96RwXi8e4e8DitQredwIHhUFy0A9tQO95Ck1DAyZIQ6KJMCJDYNjPaTeN4fMFfeA-P5SQTIHwg8D0V7jD6VWAw8JwhE7ejAoA

 

So the next question is, how do you get what you want? From your piece of paper it seems you want to re-map 1.6 - 3.3V to 0 - 5V. This isn't trivial as there's an offset to deal with. 

 

What you can do is:

- create a voltage divider to produce 1.6V

- use a buffer to maintain 1.6V (separate the voltage divider from the rest of the circuit so that it's not affected like in your original circuit)

- use a differential amplifier to subtract this 1.6V from your input signal and apply a gain of 3.125x

 

Have a look at the working circuit here:

 

EDIT: I posted the same link twice (oops). Here's the actual working circuit:

http://falstad.com/circuit/circuitjs.html?ctz=CQAgjCAMB0l3BWcMBMcUHYMGZIA4UA2ATmIxAUgpABZsKBTAWjDACgBDcQ8MFcFHl79i4JCyRh48cNELzCleGAWEaosHNXbtk6ZE4hcNEChpUaNIWapCw4+8n2yMUjCWzEU2DMTB5sHk1XSAw8MLC8SBpWEylpNgAnECYokGIeVKpsfip4uCSUtIyivPcoCnhC41MzcHKUOryE5Jqwcss7cuy+Smro0wQeGuw8Ex6UPtaByaCGoYr8gwB3UvSgnhKV+p5GuPLWXLZV9qC+ATtz7bbymr2oY9orHafrcwfVztN3hEFvqm2Kgu3BegLmQT+pw+wKhfEuRwASmscik4UZchUfhUAdAEGwAOZrKFZUz4bEEowDUYmGrmAEPJFQ2a0awLbK0HF4JAoaA0JA4vEAe1JIDUmMgpGoMDg2F6FX4aCMbGFYHoYosEo0OOkZEUhEm4FM4Ho2DYQA

 

 

 

 

[akio123008]

On 5/20/2021 at 5:23 AM, James Evens said:

I am somewhat interested if it the circuit you are feeding this into blows up, just misbehaves or works.

Haha no it's not that spectacular. Just doesn't work.

[MarcWolfe]

On 5/22/2021 at 3:42 PM, akio123008 said:

Have you simulated the circuit and looked at what happened? 

 

You have a voltage divider (the 5V - 0V thing at the bottom) and you've written 1.6V next to it. That's not correct. A current will flow from the opamp output into your voltage divider, messing up your standard voltage divider calculation.

The voltage in that node will not be 1.6V, in fact it will change with the input signal.

So there's your answer to that; yes it will. I built the circuit for you in an online simulator so you can see for yourself:

http://falstad.com/circuit/circuitjs.html?ctz=CQAgjCAMB0l3BWcMBMcUHYMGZIA4UA2ATmIxAUgpABZsKBTAWjDACgBDEQm8YlEGjx8BxcEhZIw8eOGg5ICGmH4EwS1nIUJCMsNgx5sYQshlsAToPyC8wnreFVp8S914o77kNkJOKrlYOvsJgGKYhUAFwbsF+IEw0wpHO5gBK3mGmTA5ZUfQ0VDB4SCjQNEhF0AhsAO7eQg1ekHXWwp7tNioCLRmJyfFMxBHxRbRUlVEwNQDm4OE+8XmFYy31jd2CGAKba96RwXi8e4e8DitQredwIHhUFy0A9tQO95Ck1DAyZIQ6KJMCJDYNjPaTeN4fMFfeA-P5SQTIHwg8D0V7jD6VWAw8JwhE7ejAoA

 

So the next question is, how do you get what you want? From your piece of paper it seems you want to re-map 1.6 - 3.3V to 0 - 5V. This isn't trivial as there's an offset to deal with. 

 

What you can do is:

- create a voltage divider to produce 1.6V

- use a buffer to maintain 1.6V (separate the voltage divider from the rest of the circuit so that it's not affected like in your original circuit)

- use a differential amplifier to subtract this 1.6V from your input signal and apply a gain of 3.125x

 

Have a look at the working circuit here:

 

EDIT: I posted the same link twice (oops). Here's the actual working circuit:

http://falstad.com/circuit/circuitjs.html?ctz=CQAgjCAMB0l3BWcMBMcUHYMGZIA4UA2ATmIxAUgpABZsKBTAWjDACgBDcQ8MFcFHl79i4JCyRh48cNELzCleGAWEaosHNXbtk6ZE4hcNEChpUaNIWapCw4+8n2yMUjCWzEU2DMTB5sHk1XSAw8MLC8SBpWEylpNgAnECYokGIeVKpsfip4uCSUtIyivPcoCnhC41MzcHKUOryE5Jqwcss7cuy+Smro0wQeGuw8Ex6UPtaByaCGoYr8gwB3UvSgnhKV+p5GuPLWXLZV9qC+ATtz7bbymr2oY9orHafrcwfVztN3hEFvqm2Kgu3BegLmQT+pw+wKhfEuRwASmscik4UZchUfhUAdAEGwAOZrKFZUz4bEEowDUYmGrmAEPJFQ2a0awLbK0HF4JAoaA0JA4vEAe1JIDUmMgpGoMDg2F6FX4aCMbGFYHoYosEo0OOkZEUhEm4FM4Ho2DYQA

 

 

 

 

Well, at least was able to poke at it to get this. 

I originally had an option drawn up with a buffer, apparently I wrote it horribly wrong. I should manage to sort it out with that simulator. Thanks.

[MarcWolfe]

2 hours ago, MarcWolfe said:

Well, at least was able to poke at it to get this. 

I originally had an option drawn up with a buffer, apparently I wrote it horribly wrong. I should manage to sort it out with that simulator. Thanks.

After poking around at this thing a bit, and looking at it's scope readings, I have a hard time trusting it. 

[MarcWolfe]

58 GIGA amps on output and keeps climbing... 

[MarcWolfe]

Either this simulation is fucked, or my understand is. Can't figure out how to change the gain without also getting a change in offset  

[MarcWolfe]

I guess I should have started with this. The circuit drawn on there may be garbage, but I explain what I'm going for as far as wave (pulse) shaping.

 

[akio123008]

6 hours ago, MarcWolfe said:

After poking around at this thing a bit, and looking at it's scope readings, I have a hard time trusting it. 

You need to not poke around with it. The values of the circuit I sent were already correct. 

 

Just to make sure, this is what it's supposed to be: (I've put a square wave source on for this picture, triangle works fine too)

 

 

 

If you want to change the gain of the circuit, you need to change the 31.3k resistors to a different value (but they do need to be the same!)

For instance 20k would give you a lower gain, 40k would give you more gain (again, on both). 

 

31.3k gives you the 5V you asked for.

 

6 hours ago, MarcWolfe said:

58 GIGA amps on output and keeps climbing...

This is no surprise, given you connected an insane 10F capacitor to the output, which also has no internal resistance because it's an ideal simulator model. You're essentially shorting the output, and in a simulator where wire resistance doesn't exist, that gives you enormous current.

 

 Why did you do that? I don't think you actually own a 10F capacitor.

 

 

 

5 hours ago, MarcWolfe said:

Can't figure out how to change the gain without also getting a change in offset 

I mentioned that earlier in this post, you need to change both of those 31.3k resistors to a different (but the same) value.

[MarcWolfe]

5 hours ago, akio123008 said:

You need to not poke around with it. The values of the circuit I sent were already correct. 

 

Just to make sure, this is what it's supposed to be: (I've put a square wave source on for this picture, triangle works fine too)

 

 

 

If you want to change the gain of the circuit, you need to change the 31.3k resistors to a different value (but they do need to be the same!)

For instance 20k would give you a lower gain, 40k would give you more gain (again, on both). 

 

31.3k gives you the 5V you asked for.

 

This is no surprise, given you connected an insane 10F capacitor to the output, which also has no internal resistance because it's an ideal simulator model. You're essentially shorting the output, and in a simulator where wire resistance doesn't exist, that gives you enormous current.

 

 Why did you do that? I don't think you actually own a 10F capacitor.

 

 

 

I mentioned that earlier in this post, you need to change both of those 31.3k resistors to a different (but the same) value.

Guess I'll pretend I understand why that 31k to ground matters and try it.

[MarcWolfe]

14 minutes ago, MarcWolfe said:

Guess I'll pretend I understand why that 31k to ground matters and try it.

Oh, because feedback to the side that has the offset. I'm getting there I guess.

[MarcWolfe]

Guess I'll deal with the weird curve of an audio pot so I can use something I (probably) have laying around. Looks like I'll want a double pot with pretty high max resistance. 

[MarcWolfe]

Looks like I should also have my add subtract offset for tuning the engine be at the end, to avoid changes in gain from changing the offset further.

[MarcWolfe]

Ought to do 

 

[akio123008]

13 minutes ago, MarcWolfe said:

Ought to do 

Just a tip for that simulator, if you right click a resistor, and go to "sliders" you can make it variable. That way you don't need to draw in potentiometers with one pin disconnected.

[MarcWolfe]

So my latest problem to solve is realizing a 555 timer in monostable mode doesn't go between 1/3 and 2/3 Vcc, and it doesn't have a symetrical charge/discharge curve. Could potentially work in my favor because the MAP sensor might also not have a liniar curve. Haven't seen any example of integrator OP-amp circuits the aren't meant to be astable. If I had super fast fuel injectors, I'd make the circuit even more comparable to a class D amp and have a 555 timer enable that circuit for a specific amount of time per revolution.

 

Any suggestions for circuits to research would be great.

[MarcWolfe]

Side note.

 

[MarcWolfe]

On 5/28/2021 at 11:38 PM, MarcWolfe said:

Side note.

 

There we go.