Hello ... new guy here, but I'm having some problems with my LM2907N-8 chip. I've been stuck on this too long, and wondering if anyone can help my with some fresh ideas. This whole thing seemed very straightforward from the datasheet

My input frequency is 25 - 200 Hz, 5 V square wave. Within that frequency range, the duty cycle of the 0V portion of the square wave can change.

My VCC is 12V, and my calculated R1, C1 and C2 are: 60k, 47nF, 3uF. I'm using a 68K resistor for R1 b/c I didn't have a 60k.

I attached my circuit.

Thanks to anyone who gives this a shot!!

 

The first thing that strikes me is that you are comparing the input signal with GND, but your signal never crosses GND. You need to level-shift the input in a negative direction, or AC couple it. AC coupling is easiest, but duty cycle might be a problem. What is the width of the positive portion of the waveform (you implied it was constant)?
Actually, the simplest solution would be to use the 14 pin version, because you can set the threshold with a resistor divider.

 

Thanks Ron. I saw some 'examples' from the datasheet about reading engine RPM from an ignition (0 - 12V square wave) so I figured the chip is capable of referencing a ground input. I think I missed the part about the ground referenced input interfacing with VR sensors ... grr

The input to the chip is pulled to ground by the sensor (requires a 10K pull up to 5V). The low pulse width will be between 1 and 5 ms, and the high pulse width will vary depending on the current output frequency. The sensor frequency output does not rapidly change.

I have the 14 pin version of the chip too. I will try setting the crossing voltage to 2.5V ... I assume applying 2.5V to pin 11 (inverting terminal of op/amp on charge pump circuit) will be a good crossing point for the chip to detect frequency.

 

Thanks Ron. I saw some 'examples' from the datasheet about reading engine RPM from an ignition (0 - 12V square wave) so I figured the chip is capable of referencing a ground input. I think I missed the part about the ground referenced input interfacing with VR sensors ... grr

The input to the chip is pulled to ground by the sensor (requires a 10K pull up to 5V). The low pulse width will be between 1 and 5 ms, and the high pulse width will vary depending on the current output frequency. The sensor frequency output does not rapidly change.

I have the 14 pin version of the chip too. I will try setting the crossing voltage to 2.5V ... I assume applying 2.5V to pin 11 (inverting terminal of op/amp on charge pump circuit) will be a good crossing point for the chip to detect frequency.

That sounds like a good next step. What resistor values do you plan to use?
BTW, let me warn you that I have never used this chip, but I am an old analog engineer, and I'm generally pretty good at interpreting datasheets.

 

I'm just going to use two 10k resistors off my 5V supply (supply for the 10k pull-up ... I have two regulators on the board). Not sure if that will restrict too much current ... we'll see. At least the voltage divider calc is easy!!

I'll keep you updated

 

Got it working. Ron H, you were right about the first issue I had with the inverting terminal being tied to ground on the tachometer side of the chip. My output is about 1.5 V @ 50Hz and 4.5V @ 160Hz. Good enough for me!

I ended up having a bad chip on my test circuit, and I replaced it with another bad one (someone but a broken chip back in my supply of good ones!!!!!!!). Anyway, this circuit works (finally). I think I could have used the resistor network instead of the diode to set my threshold voltage.

Thanks for the help!

 

Got it working. Ron H, you were right about the first issue I had with the inverting terminal being tied to ground on the tachometer side of the chip. My output is about 1.5 V @ 50Hz and 4.5V @ 160Hz. Good enough for me!

I ended up having a bad chip on my test circuit, and I replaced it with another bad one (someone but a broken chip back in my supply of good ones!!!!!!!). Anyway, this circuit works (finally). I think I could have used the resistor network instead of the diode to set my threshold voltage.

Thanks for the help!

Good news! Glad I could help.

 

Hello ...

My input frequency is 25 - 200 Hz, 5 V square wave. Within that frequency range, the duty cycle of the 0V portion of the square wave can change.

My VCC is 12V, and my calculated R1, C1 and C2 are: 60k, 47nF, 3uF. I'm using a 68K resistor for R1 b/c I didn't have a 60k.

I attached my circuit.

Thanks to anyone who gives this a shot!!

twaroc, how you calculated R1, C1 and C2 ?. I already read the datasheet, but i don't understanding it.


Thanks

 

Try looking at page 4 of the LM2907/LM2917 application notes. They have formulas for R1, C1 and C2.

http://www.national.com/an/AN/AN-162.pdf

 

in page 4, i ve problem determine the value of C2,,,
confused about V ripple,please help