HELP understanding SIGNAL PROCESSING schematic

[StudentSA]

I am building a project to read a vehicle engine RPM. I have been having sleepless nights trying to design the input signal filtering circuit.

The input signal is taken off the vehicle`s coil negative terminal and ideally should provide a pulse of 12V, 0V, 12V, 0V etc... as the engine cylinders fire.

This is not the case as the signal can have some serious resonance and voltage spikes.

Anyhow, I have found a circuit in a magazine that also uses the signal to do REV Limiting and would like to know if someone could help me understand the input filtering part of it?

I understand a fair fit of electronics so just maybe point out to me the building blocks.

Here is the circuit :



The blocks is just my assumed understanding, Please can you correct my understanding and elaborate on the rest of the circuit.

The Output Vout is used to pulse a input pin of a counter.
The Input Vin is from the coil -ve and has resonance and voltage spikes

Your help is appreciated

 

[MikeMl]

I though I answered this for you here?

 

[MikeMl]

...

Anyhow, I have found a circuit in a magazine that also uses the signal to do REV Limiting and would like to know if someone could help me understand the input filtering part of it?

I understand a fair fit of electronics so just maybe point out to me the building blocks.

Here is the circuit :

View attachment 40229

The blocks is just my assumed understanding, Please can you correct my understanding and elaborate on the rest of the circuit.

The Output Vout is used to pulse a input pin of a counter.
The Input Vin is from the coil -ve and has resonance and voltage spikes

Your help is appreciated

This circuit has some serious problems. Take a look at the output. It would cause multiple counts at each point opening...

 

[StudentSA]

Hi Mike

I appreciate your help. What I am trying to achive is some understanding of what this designer considered when approching such a problem. I hope to gain some understanding into the pro`s and cons such that I can design something even more acurate.

I mean no disrespect, but you seem to have constructed the circuit a bit wrong, please have a look.

Im hoping to understand the building blocks behind this circuit such as the Filters and why they have been designed for a specific frequency. Questions like why the 1W 22k Ohm Resistor?

Appologise if I am sounding confusing.

All help is highly appreciate, I only seek knowlage in the hope to share in future.

 

[mneary]

Another multi-thread question..... do you think you'll get different answers?

 

[StudentSA]

This thread is about understanding a specific schematic, Yes they do link up and we can argue endlessly to what extent they compare. I am new to this forum and in due course will learn the ropes.

I appologise for any inconviniance I have caused you mneary, However I belive forums would work alot better if people brought solutions and ideas to the table and left petty issues aside.

Kind Regards.

 

[mneary]

I am currently trying to help someone in another thread, and whenever I put in the work to find a solution for him in one thread he says his other thread solved it yesterday. It's not rewarding. It's not petty.

 

[StudentSA]

Good Point.
I agree, and will take your thoughts into consideration in the future.

Thank you

 

[StudentSA]

Good day Mike

You mentioned that my circuit attached in the first post has "some serious problems. Take a look at the output."
However if you consider your circuit that you simulated it is not the same as mine that was posted... Capacitor C3 is in the wrong place and a resistor is missing before R6.

Maybe this is why it has some serious problems?

Regards,
Umar F. Dockrat

 

[Gary B]

From Vin to AC coupling forms a band pass filter. The zener gives a rough clip to +10 volts maximum and attempts to ground out the negative half of the cycle. The 2.2K input resister also provides current limiting for the zener. Given the +5 clamping reference, the output from this circuit probably goes to a T2L Schmitt trigger for final cleaning. Its time constant would be just a little shorter than the maximum pulse rate at maximum RPM. Therefore contact bounce and noise spikes are blocked at that point from your counter rather than at the input coupling circuit.

Does that help?

 

[StudentSA]

Thanks for your reponse, The schematic starts making more and more sense daily.

How do you figure that from Vin till the ac coupling is a Band Pass Filter (BPF)? I understood that a BPF had to be comprised of a Low Pass Filter (LPF) and a High Pass Filter (HPF) like in the following picure.



Do you have any reference that the configuration used i.e. 22k,47nf and 10k form a BPF?

The output of this circuit is to be used to drive a TTL IC (Counter) the inclusion of a schmitt trigger might provide additional signal integrity, I will keep this in mind.

Regards,
StudentSA

 

[MikeMl]

Good day Mike

You mentioned that my circuit attached in the first post has "some serious problems. Take a look at the output."
However if you consider your circuit that you simulated it is not the same as mine that was posted... Capacitor C3 is in the wrong place and a resistor is missing before R6.

Maybe this is why it has some serious problems?

Regards,
Umar F. Dockrat

Ok, I went and fixed the error I created when inputting your circuit into my LTSpice sim. Here it is again, only now its performance is even worse. I am plotting the four nodes: A, B, C and Out. Notice that the output from your circuit has multiple edges at each points opening, and the output is not sitting at a clean "logic" level during the time the points are closed.

V(A) and V(B) are similar, except for the effect of the DC blocking capacitor, C3. V(C) and V(OUT) are identical, because the low-pass filter R1/C4 is totally ineffective. Like I said earlier, this circuit is not useful. Compare V(OUT) to the V(OUT) in the circuit I posted previously.

 

[StudentSA]

Dear Mike

Thank you for your efforts, things do look quite bad in the above circuit... Can I please request you to post the simulation files? I would like to play with it a little before I give up all hope.

These results you attained will certainly keep me in deep thought for a while.

Regards,
StudentSA

 

[MikeMl]

Dear Mike

Thank you for your efforts, things do look quite bad in the above circuit... Can I please request you to post the simulation files? I would like to play with it a little before I give up all hope.

These results you attained will certainly keep me in deep thought for a while.

Regards,
StudentSA

Here is yours and mine

 

[Gary B]

22K 1W and 47nF form a single poll low pass filter with a cutoff of about 90 KHz. The 10K and 2.2 mfd form a high pass single poll filter with a cutoff of about 10 Hz. Together they define a single poll band pass filter.

If the first gate of the selected counter is a Schmitt trigger, there is no increase of cost or complexity. Since I have been out of the country for a while, I can’t say if you can still easily find the things in the hobby shops but, one used to be able to find quad Schmitt NAND ICs for a buck or two.

 

[ericgibbs]

22K 1W and 47nF form a single poll low pass filter with a cutoff of about 90 KHz. The 10K and 2.2 mfd form a high pass single poll filter with a cutoff of about 10 Hz. Together they define a single poll band pass filter.

Is this correct or is it a typo.?

 

[MikeMl]

22K 1W and 47nF form a single poll low pass filter with a cutoff of about 90 KHz. The 10K and 2.2 mfd form a high pass single poll filter with a cutoff of about 10 Hz. Together they define a single poll band pass filter...

There is the biggest problem with SA's circuit. It should not be a BANDPASS filter, or have a DC blocking capacitor. The signal being filtered is basically a pulse with swings from ground (points closed) to +12V (points open) with a lot of ringing on the edges.

It stands to reason that the circuit should do only two things: clip off the positive/negative peaks (the Zener or diodes), and then LOW-PASS filter the ringing. The circuit MUST have response to DC, so as to pass the dwell.

 

[Gary B]

I believe it is. The 47nf cap passes high freq. to ground so it defines a low pass filter. The 2.2µf cap is in series with the signal path so it defines a high pass filter. I have a reference book (Reference data for radio engineers) with a nomograph for filters. I am assuming the frequencies I got from it are correct. If in doubt, one can do the math manually to verify.