Metal Detector as a battery SoC sensor

[Mosaic]

Hi All:

I am developing a sensor coil to detect the condition of lead acid state of charge based on the metal content in the plates.

I am looking for suggestions as to frequency of the LC tank to achieve the penetration (>4") in order to detect the overall metal in the battery. As u may know a discharged battery has the least metallic lead.

I am also interested in a stable LC tank circuit to feed to the MCu . My design requires an external oscillator circuit as almost all MCU pins are used up already.

If the frequency is high I may need to do a local oscillator to create a beat freq. with the sensing LC tank thereby providing a beat freq. difference for the MCU to count.

 

[alec_t]

I'm no expert on battery charging. Can we assume that lead formation during charging is evenly distributed throughout the plate? If so, then isn't penetration depth unimportant? I can see that battery size and geometry would be factors to be taken into account.

 

[Mosaic]

Only new batteries have even distribution. Older batteries have weak cells. I did a test with a 4" , 400uH air coil. Did a battery charge / discharge cycle, however the resulting change in inductance of the coil @ 100Khz is insignificant.
The coil was mounted flush agains the battery sidewall on the positive side.

The coil inductance does respond (drop by 20% @ a 100hz test freq.) as I move it toward the battery, so I am thinking the lead connectors in the battery swamp any changes in the plates during charge.
Coil inductance responds poorly at low frequency testing . 1Khz elicits little change in the inductance due to the metal proximity.

 

[alec_t]

Only new batteries have even distribution.

Oh. So the age of the battery would be another factor to allow for.

the resulting change in inductance of the coil @ 100Khz is insignificant.

Pity. The method looked promising initially.

 

[dr pepper]

A very interesting circuit, and probably a reliable indication.
As for the LC tank circuit, if you want a stable freq then a xtal osc as you probably know gives a stable o/p, I dont know why you want a lc osc circuit, unless the l is the pickup inductor if thats the case than you could 'ping' the lc tank with the micro and count the number of rings or the duration of rings with a simple circuit maybe using the micro's comparators and counter/timers.
If you really want a continuous oscillation then maybe you could look at building a xtal disciplined (I think that what they are called) oscillator, I've seen these used for fm transmitters, where the xtal controls the centre freq to an extent but the actual freq is modulated by an external audio input, might be worth googling.