MAX9924 configuration and use cases

[Fred]

Wow, that really *is* magic! So If I reinstall my resistors and drop the grounds it will work? I don't understand how, no, wait, I think I do understand how, but not well enough to explain it. Marcos told me to ground those suckers, naughty Marcos...

I will try this later today post writing more code. I have a simple hack to make it work for now, so no big deal.

Thank you very much Jean!

Fred.

[Fred]

PS, I knew that you'd "just know" ! :-) You're, as always, a legend! :-)

[TonyS]

I would guess he used the 10k and 1nF noted in the MAX data sheet. However, Marcos mentioned upping the R values due to some heat issue. I'm not sure about the heat issue. 5V and 10K is small power. Can't a VR produce 100V if you get it really close? Even at 100v and 10K were still talking .01w.

Actually, It was I who pointed out this issue to Marcos. 100V and 10k gives you 1.0W of dissipation, not 10mW. I also pointed this out to Jean last year after seeing a picture of the small sized resistors on his board (hopefully changed).

Fred -
Can you "show" me how your Hall sensor interfaced to the MAX chip initially (I guess per Marcos's suggestion). I do have an interest because I used the MAX chip on my design and although I only have "paper" to work with, I did provide what I thought would be a functioning Hall sensor interface option (pull-up and grounded (-) input).

Thanks
-Huff

[jbelanger]

...
Fred -
Can you "show" me how your Hall sensor interfaced to the MAX chip initially (I guess per Marcos's suggestion). I do have an interest because I used the MAX chip on my design and although I only have "paper" to work with, I did provide what I thought would be a functioning Hall sensor interface option (pull-up and grounded (-) input).

Thanks
-Huff

That's not correct. The pull up is of course correct but grounding the VR- input will not work as mentioned before. You need to leave it floating which will use the internal 2.5V reference. Check the datasheet.

And 100V is 1W if this is a continuous 100V DC signal. A VR signal will dissipate much less than that but I did change to 20K 1/4W and there are 2 of them. So it should be safe for most cases and it is always possible to put a shunt resistor if the sensor signal is really too high.

Jean

[Fred]

+1 to Jean, Huff, your shlt won't work! And yes, our shit does now work with the grounds floating :-) Hooray. Thanks so much Jean! :-)

Easy mistake to make, seems like common sense, counter intuitive... This thread should be plenty googlable now, maybe I'll throw in some more keywords:

max9924 max9925 max9926 max9927 maxim vr variable reluctance interface conditioning sensor circuit, etc :-)

Fred.

[jharvey]

Actually, It was I who pointed out this issue to Marcos. 100V and 10k gives you 1.0W of dissipation, not 10mW.

Sorry when I did the Power equation in my head, I buggered it. You are correct 1 watt at 100v and 10K.

100V seems to be what I was expecting. When I look at the wave forms from that fellows posting, I see mV and 5V signals. 1 Watt of power seems a bit higher than I would expect from those little VR's. Also I'm surprised the datasheet notes 10K. Are we sure it's up to 100V at 10k? I wonder if I'm running under some assumptions. I don't have much hands on experience, so I'm a bit ignorant about it. Could the 100V have been a case with higher than 10k?

Hmmm, is the max chip 100 V tolerant, Wouldn't it arc the SMT leads? This is the smallest chip we have, I seem to recall the pitch spacing was .007 in or something like that.

[Fred]

That voltage is only seen at the outside end of the resistors, the inside end is V clamped by the device, hence the power dissipation concerns. Does that make sense?

[TonyS]

That's not correct. The pull up is of course correct but grounding the VR- input will not work as mentioned before. You need to leave it floating which will use the internal 2.5V reference. Check the datasheet.

Thanks Jean. I took a good look at the MAX datasheet and see why / how this works now. So to summarize sensor connections to the MAX chip -

1. Differential VR sensor: Connect one lead to the input resistor going to the (+) on the MAX chip and the other lead to the input resistor going to the (-).

2. Single-ended VR sensor: Connect the single lead to the input resistor going to the (+) on the MAX chip and GND the input to the resistor going to the (-).

3. Hall sensor: Connect the output of the sensor to the input resistor going to the (+) on the MAX chip and do not connect anything to the input of the resistor going to the (-). If the Hall sensor is of an "open collector" type, add a pull-up resistor from 5V to the output of the sensor.

Huff, swapping VR sensor leads can't be done, they have to be polarised correctly always. For an actual VR sensor setup this will always be correct, by default, period. For digital inputs it could change depending on buffers and other [shizzle] between cpu and sensor. However, if we standardise on the max setup, with just a pullup on the input, then probably it will work the same always and I can code to whatever the max provides and others can invert their setups if required. The only thing that is certain is that the code will not have configurable polarity for rpm inputs.

Please excuse my ignorance on this but I read the M...manual about this subject and since you have two leads, swapping leads can / will certainly be done and they do seem to provide a configurable polarity. Is this a philosophical difference?
Thanks
-Huff

[Fred]

View a VR wave form, and think about what the output of a VR conditioner is, and look at the two possible patterns and you will soon see why :-)

[jharvey]

A VR will generate a specific amount of energy. That energy is based on rate of change in the encoder wheel, strength of the magnet, coil windings, ect. I don't know how much energy it will generate, but per the above 100v on 10K it would indicate it generates 1 watt continues, with most wheels that at least a 2 to 4 watt spike. To me this would seem to be more than I would expect, but I'm ignorant about how much energy potential a VR has.

Using 1 watt as the the energy potential to dissipate, you can change your voltage based on resistance. So if you put in a 1meg instead of 10k, you'll generate a much higher voltage to dissipate the energy. If that voltage goes to high, it will arc across the leads on the MAX chip. So there is a balancing act between resistance and dissipated power.

Also wouldn't that be 100vac? So you would only dissipate about 70% of that 1 watt. It all gets down to the wave form.

I would like to see some more wave forms, such that I can predict the expected amount of energy in a typical VR, or at least a variety of VR's. Based on the package, I'm sure MAX doesn't expect to see 100v. The max chip is sensitive to mV, so if we are to see 100v signals, we would want to encourage small resistors of large dissipation potential.

I'm also curious where this 100V signal came from, such that we know the resistance. If it was just a VR on a scope, the VR was connected to a 1 meg ohm impedance, and the voltage would be much higher than it would be with a 10k impedance.