Hello all. I've picked up a couple MAX9925 chips to work with, but there's a couple things I'm not clear on.

The datasheet... http://datasheets.maximintegrated.com/en/ds/MAX9924-MAX9927.pdf

I will be using mode A.

The question I have with the application circuit on page 17... there are four resistors that have no values connected to the sensor. From what I can gather from the sheet is the inputs feed an op amp. Would I be right to assume that if I keep the values of the four resistors the same it would more or less be a unity gain type configuration? Is it possible I could just tie the OUT signal to IN - and eliminate all but the resistor between bias and IN+? I will not be using the OUT signal for anything other than for testing and would think I wouldn't need any sort of gain or other modifications to the signal.

I'm planning on connecting a standard Ford speedometer sensor to the inputs. All I know for sure is it is a VR type sensor.

 

There is more than one reason to use the configuration shown. Make all the resistors the same. 10K would work.

 

There is more than one reason to use the configuration shown. Make all the resistors the same. 10K would work.

Thanks #12. I was hoping to eliminate a few things... I already have too many holes to drill!!

 

Yebbut...reasons. Too much current available to the sensor, noise rejection.

Edit:
Another way to say, "too much current available" is, "too low of an impedance". If you do this single ended, the impedance at one end of the sensor (the inverting input) is very low compared to 10k. At that point, any AC noise that arrives on the wires is nulled by the low impedance feedback loop. The rest of the noise is applied to the non-inverting input, and gets amplified. The resulting amplifier configuration completely lacks the common mode rejection characteristics of a differential configuration. Instead of nulling the noise by tens of db, you amplify it.

 

... I already have too many holes to drill!!

Why not use SMT resistors then? I doubt they'd have to carry too much current anyway, since the chip is working with 5V

 

Why not use SMT resistors then?

5v^2/10k = 2.5 milliwatts, worst case. I don't think you can buy an SMT resistor that is too small to survive that.
Another advantage is that packing all the parts into the smallest possible space reduces the area of the virtual antenna which picks up noise signals.

So, yeah. Good suggestion all around.

 

Why not use SMT resistors then? I doubt they'd have to carry too much current anyway, since the chip is working with 5V

It would be nice... since this is a hobby more than anything I'm limited to one sided and one layer boards. I've found with a little creativity resistors and such make good jumpers without adding more holes for jumpers.

Yebbut...reasons. Too much current available to the sensor, noise rejection.

Edit:
Another way to say, "too much current available" is, "too low of an impedance". If you do this single ended, the impedance at one end of the sensor (the inverting input) is very low compared to 10k. At that point, any AC noise that arrives on the wires is nulled by the low impedance feedback loop. The rest of the noise is applied to the non-inverting input, and gets amplified. The resulting amplifier configuration completely lacks the common mode rejection characteristics of a differential configuration. Instead of nulling the noise by tens of db, you amplify it.

Thanks again!! I'll admit I have a lot to learn in these areas and it will take a little bit for all that to make sense. I pretty much just know enough to make myself dangerous at times. I drew up the circuit like the drawing shows. One more going over of the circuit tomorrow night and I'll be printing my transfer to make the board. I scrapped a few cars a while back and saved the sensors I knew I would be able to find a use for... kind of looking forward to playing with these things!

 

resistors and such make good jumpers

they do indeed... in fact, I use 0 ohm SMT resistors as jumpers in my circuits all the time!

 

There is more than one reason to use the configuration shown. Make all the resistors the same. 10K would work.

I finally got around to getting it to work. The 10K didn't seem to work, but I found another application where they used the 9924 and showed 100K for the unmarked ones (which are internal on the 9924). I swapped them out and it's working.

Thanks for the suggestion though... maybe has something to do with my sensor.

Just thought I'd update this and call it working!!