Capacitance Meter Firmware

posted Nov 6, 2011, 9:28 AM by Chris Stratton   [ updated Nov 6, 2011, 5:56 PM ]

atmega_capmeter with open source firmware measuring 4 pF capacitor
I've always wanted to be able to measure capacitors.  I received an analog multimeter as a birthday present sometimes in grade school, but while that could measure resistors it couldn't do much with capacitors.  At the time I was familiar with the resistor color code, but it wasn't really after engineering school that I understood capacitor markings.

A while back when buying a new soldering station from Amazon, I chanced upon a a listing for an $11 capacitance meter kit. Rather doubtful about it's accuracy, I figured the price was right (or at least a good deal on a collection of re-purposable parts) so I ordered one.

(Measuring a 4 pF capacitor at 4.2 pF requires nulling the meter's own capacitance before inserting the test cap)

Well put me down as pleasantly surprised.  The thing works.  Just by measuring the time constant of charging through a 3.3K or 3.3M precision resistor, it can read from single-digit picofarads up to at least a hundred microfarads.  I can't test the absolute accuracy (caps aren't known for their precision) but it's quite repeatable.

Really, there's only one thing wrong with the kit: while they provide the schematic, they do not publish the firmware source code.  So a few weeks ago, I decided it would be an interesting project to write a new open source firmware from scratch, and after a long Sunday of coding, here is a preliminary take:

Whadizzit ???

posted Nov 6, 2011, 7:50 AM by Chris Stratton   [ updated Nov 6, 2011, 7:51 AM ]

When I ironed the toner for a recent circuit design from the torn-out Ikea catalog page on which I printed it onto the copper, I noticed this interesting "face" formed from the ground plane and two vias under the main chip.

But the real question is - what is this circuit for ?   First correct guess wins an already converted Anti-Clock (limited to US addresses only)


posted Nov 6, 2011, 7:27 AM by Chris Stratton   [ updated Nov 6, 2011, 7:41 AM ]

I was at my favorite source for project inspiration the other day (the dollar store) and picked up a simple battery powered alarm clock, thinking I'd try and see if I could make it run backwards.

I took it apart, and found what I expected - the single AA battery powers a circuit which uses a 32 KHz crystal as a timing reference.  It divides this down, and at 1-second intervals activates the primitive "stepper motor" formed from the coil, steel core, and a little (ferrite?) cylinder under the small gear.  This then drives all the hands through the gearing.  Viewed from the back of the mechanism, the small gear of the motor rotates anti-clockwise.  I need to make it spin clockwise, to make the clock itself into an anti-clock.

So I thought "No problem! I just reverse the wires to the coil!"

Except... NOTHING CHANGED.  It still spun clockwise.  I thought "maybe I mixed up the two black wires and accidentally soldered them back in their original places" so I swapped them again - STILL no change.

Then I thought "maybe if I slide the coil off the steel core and put it on backwards" (the wires weren't long enough to fully re-assemble it this way, but it was just a test)

Except... NOTHING CHANGED.  It still spun clockwise.

Then I thought - "Maybe if I pop the ferrite cylinder off the back of that little gear and flip it over?"

Except... NOTHING CHANGED.  It still spun clockwise.

Finally, I realized there was only one piece of the system I hadn't reversed - the steel core bar.  This is the only element remaining that can be determining the direction of rotation.  I already knew the coil could come off of it from the previous experiment, fortunately the coil would fit on the opposite arm if I flipped the core over.

Except... SUCCESS!  The clock is now an ANTI-CLOCK

I did have to clip off the squarish rear corner of the core piece in order to get the clock case back together, but the problem is solved (hint - make sure the alarm switch is set to "on" when re-assembling the rear cover, then turn it "off" once assembled).

I'm not exactly sure why this works, and really should take it apart again and investigate further.

  • TIP - when hacking complex electromechanical assemblies, snap a photo of all the parts in their proper place before disturbing them, to aid in re-assembly.

1-3 of 3