Hi

See Below.
Still needs to be breadboarded to test.

eT

 

Hi

See Below.
Still needs to be breadboarded to test.

eT

View attachment 171900

That's cool! It is as well much more performant than lm331, less than half time to get stable result.
What is U2A & U2B (LM358?) and what are you using for simulation?

Can you drop here the circuit in higher resolution, may also the BOM list?

 

Hi

"That's cool! It is as well much more performant than lm331, less than half time to get stable result."

Might be a little premature for parts though. I have the circuit on a breadboard....still working out the filters. The 37hz requirement makes it tough. Probably change the buffer to a 2nd stage filter. Also, the output varies with the supply variations, so it probably will need a 9v regulator. More to come...

eT

 

Hi

"That's cool! It is as well much more performant than lm331, less than half time to get stable result."

Might be a little premature for parts though. I have the circuit on a breadboard....still working out the filters. The 37hz requirement makes it tough. Probably change the buffer to a 2nd stage filter. Also, the output varies with the supply variations, so it probably will need a 9v regulator. More to come...

eT

Is getting more tricky of course! the whole think i would need twice due to returning fuel line. Modern engines do have the fuel return line to keep the pressure constant. But the return line will have a lower frequency range i guess and i do not know yet how much will return during normal operation. The Teensy 3.2 will subtract the value of the return line from the supply fuel line before it will be dropped on the CAN bus (NMAE200 network).
Thought to use the same base as well to grep the RPM signal from the W connector of the generator of the engine, but this engine should also have a rmp sens with better signal (more square like) at the end of the steering chain. Have to count the teeth for frequency calc first to think more about that.

At the end it should be a complete digitalized engine logger/interface dropping all informations on the nmea2k network. This makes it possible to use the whole potential of the multifunctional display (Raymarine GPS cardplotter).
The teensy will use a library from ttlapalainen,
https://github.com/ttlappalainen/NMEA2000 , very good stuff.

I very appreciate your efforts! Thx

 

Can you drop here the circuit in higher resolution, may also the BOM list?

Is this clearer now? U2 is LM358 and the simulator is the LTSpice. It's available freely from the net.

 

Hello again

So...here's where I think would be a good place to start. I'll leave you with this design to experiment with only because the parts might get expensive. The LM358 might be OK but depends on your requirements. The low frequency performance requirement make this tough. The low pass filter opamps need to have high CMRR (~90>), with low input offset (<1mv). The ripple voltage needs to be kept down below a millivolt. I've show an LM2917 with the zener, but since there needs to be a regulator, an LM2907 can be used instead. If you use the LM2907, you'll need to remove R5 and recalc R1. If you breadboard this, keep all leads and connections short as practical to prevent noise. BTW-The output seems to settle much quicker on the bench then in the simulation.

Good luck.
eT



​

 

Hello again

So...here's where I think would be a good place to start. I'll leave you with this design to experiment with only because the parts might get expensive. The LM358 might be OK but depends on your requirements. The low frequency performance requirement make this tough. The low pass filter opamps need to have high CMRR (~90>), with low input offset (<1mv). The ripple voltage needs to be kept down below a millivolt. I've show an LM2917 with the zener, but since there needs to be a regulator, an LM2907 can be used instead. If you use the LM2907, you'll need to remove R5 and recalc R1. If you breadboard this, keep all leads and connections short as practical to prevent noise. BTW-The output seems to settle much quicker on the bench then in the simulation.

Good luck.
eT

View attachment 172180

​

Thank You!

The prices of the components do not matter really. I just make this for myself and good functionality is more important than to spare some money.
You showed me up a way I haven't thought about and since I have not been working with an electronic design the last 20 years I didn't come on the idea to use OP amps at all.
I didn't get this handled also due to the low input frequency and since the return fuel line will have an even lower frequency I guess it will not be possible to do this as I planned now. But I will try to count the impulses with the Teensy and calculate the return volume.

However, this is a perfect start and I never expected so much help here at all. Just to complete the post, I will post the result here and the end device. May somebody will have use of it as well.

 

@eetech

Just a question regarding the 5k pot. What would be the calculated total resistance of R1 39K and the 5K pot? The point I am asking is that I would not have any reference to adjust the pot. In fact, the highest frequency is 611 Hz like given by 100l/hr and 22k pulses per liter and highest FVOut for this circuit, what resistance is there to be used or how can I adjust this without doing the physical test which is almost impossible?

 

It’s a 10K pot And the circuit is set for 1000 Hz.
If you want full scale to be 3.3 V at 611 hz then you’ll have to recalculate

 

It’s a 10K pot

I deleted the post I wrote here before since I got the answer...

 

It’s a 10K pot And the circuit is set for 1000 Hz.
If you want full scale to be 3.3 V at 611 hz then you’ll have to recalculate

@eetech00

Regarding the recalculation to get the full scale for other frequencies, If I take the formula I am missing one constant to start with. But is this correct that I can easily calculate it directly proportional? let's say for 100 Hz I could divide C2 and the total of P1 and R1 by 10?

What are the criteria's to choose C2 accordingly?