Parts: ST Micro’s VNH2SP30 and VNH3SP30 motor drivers

So a new project that has come to mind is a remote controlled treadded vehicle such as a tank or crawler, and I started thinking about what motor drivers to use. Having had some good success with ST Micro before, I decided to take a look on their site first.

Having taken a look at their motor driver applications page showed old Multiwatt-11 packaged drivers or newer power-SSOP packaged drivers with less power than I’d like, however automotive grade electronics components typically have a few key features: good packaging/ease of mounting and cooling, power, reliability. The automotive page has a motor driver page as well which has shown the VNH2/3SP30 drivers (datasheet) to have voltage, current, and thermal protection on a thermally efficient die pattern (3 separate dies with direct connection to the bottom, PCB with thermal vias acts as the heatsink).

What is even more interesting is that you can run both sides of the H bridge in parallel, and use 2 separate IC’s to get even more power. The chip itself will withstand 30A, but only as long as it doesn’t overheat, which is limited by the board itself. The datasheet itself shows some example patterns of PCBs to connect the inputs and outputs while having more copper to cool the chip itself, however the chip would also benefit from a heatsink mounted above and maybe below (with a layer of conformal material). Pololu has actually done testing itself to show that continual current load can be in the 14-15A range, and Phidgets has simply derated their IC’s down to 15A maximum duty cycle.

I’ve noticed that a few devices such as the Phidgets high current motor driver and Pololu’s motor drivers use these chips, and in particular the Pololu website has a good comparison between the VNH2 and 3 series. While the chip itself is pricey ($15+ on Digikey, Mouser, etc) the form factor is useful since it lends itself to miniaturized robust robotics, despite the original automotive usage.

If any ST engineers read this, some suggestions for a product targeted towards robotics:

• Non SMD format, thruhole preferrably
• A SMD format that simply has the pins on the reverse side so that heatsinking can be done directly to a metal surface via clamping. (Note a insulating layer needs to be applied, such as the one used in your ISOTOP/SOT-227 packaged power-fets).

Update: I have discovered that at my Eagle install (5.4.0) has the MultiPowerSO-30 package and the VNH3 part already in the st-micro library, unfortunately the VNH2 with current sense and diagnostic are not present (had to make my own part).

Desulfator build and Lead Acid Revisit

So my previous post regarding reconditioning dried out sealed lead acid batteries mentioned the use of a desulfator. I based my design off of a common desulfator circuit which uses 2 inductors to push pulses of 50V into the battery at 1kHz, with the only major modification being replacing the circuit between the 555 timer and the mosfet (originally a diode for fast turnon, and a resistor for bleeding charge from gate) with a 2N3904/2N3906 pair to drive the gate (and the other modifications being part substitutions.

Revive ‘dead’ sealed lead acid UPS batteries

Just to begin, this guide is not a fix-all solution to dead uninterruptable power supply (UPS) batteries, I will describe what the cause of one particular, common failure mode for sealed lead acid batteries is, and the solution.

Essential Hacker and Geek Skills

So two recent blog posts by UptownMaker and Wired covered different essential hacker and geek skills, they are good lists but both have their niches. I figured that readers would find these useful so without further ado.

The UptownMaker list is more oriented towards makers while the Wired list is more about computer skills and geek cred.

Probably the only thing I would add to that list is how to drill holes in tough materials such as steel, the proper use of lubricants, and tool care.

Data acquisition & sensors, Copper Corrosion, Failures, and ‘Name that device’

So my new recent project has been to develop a bunch of sensors and a data acquisition unit (DAQ), and I have already finished a few tiny boards with some more work to be done. Just as some eye candy, a voltage sensor (divider for a ADC on a microcontroller, selectable to 1:2, 1:3, or 1:5), and a current sensor (resistive 0.001ohm shunt with a current sense amplifier, MAX4372).

Another thing I thought I would bring up is the effect of silicone caulk on electronics. While on the ASV project, despite my protest that caulk would corrode circuit boards due to the release of acetic acid fumes, we ended up moving ahead with it anyways. Regardless, here are some photos of a circuit board exposed to some fumes from caulk, even at a far away distance.

Everyone has had some failure when trying to develop a new device, and as part of my line of sensors, I was going to use a strain gauge taken out of an electronic bathroom scale. Unfortunately I’ve noticed that the resultant sensitivity is far too low with a differential opamp, and that even 1% resistors are not good enough to get the desired value of 0V difference between the two ends of the wheatstone bridge. I’ll have to revisit it at some point, mainly because a strain gauge would be a useful piece of kit. I’m sure that if I get it trimmed out properly and maybe use a higher Vcc for more resolution that the strain gauge will be useful, but until then its a bit of a failure.

And now for a new thing, ‘name that device’. I don’t have anything to give away or award to whoever gets it right, but regardless here are some photos. If you have any idea of what this device is, leave a comment to this post. I’ll reveal its identity after a week or so.