I started a new thing this last week. I've been impressed by Adrian Colyer's the morning paper for a long time. It's a project where Colyer read and commented on a computer science paper every week day during term time, and wrote a summary of the paper. Some of the summaries are quite long. It was really useful, for those of us who were working in a similar field, because it meant that we could get a quick overview of papers without going to the trouble of reading them.

I have no intention of displaying the level of diligence that Colyer did, but it did make me think about a less ambitious "read a thing every day" project. I've been thinking of it as "datasheet a day", but I've been reading a mixture of datasheets and application notes.

Here's what I read this week:

Datasheet: Maxim MAX7030 434 MHz transceiver

This is a fun little chip I'm thinking of using for a project. It's like a handy little toolbox of RF components, all in one package.

One thing I picked up from reading this is that there is a lot to learn about RF! I learnt a bit about ASK modulation, data slicers, image rejection, and some other things, but mostly learnt that I have more to learn...

Improving Resolution of SAR ADC (TI SLOA249)

This is a relatively pedestrian app note about oversampling for SAR ADCs. Still useful to read, but not exciting.

Datasheet: TI ADS1115 ADC

This part is used on some Sparkfun ADC breakout boards. I have one and have never got round to using it. I thought they were less good than they are. I have some experiments I want to do where I might be able to use these.

Some things I learnt from this:

I've always wondered why input voltage limits for chips are often quoted as something like GND-0.3V to Vdd+0.3V. I now think I know why, based on a schematic showing the ESD protection diodes on the inputs of this chip: that 0.3V value is just the diode drop from the ESD diodes! So, one mystery (for me) sort of solved.

This statement about I2C pull-up resistors is probably obvious to more experienced people, but it was useful to have it stated explicitly: "Pullup resistors are required on both the SDA and SCL lines because I2C bus drivers are open drain. The size of these resistors depends on the bus operating speed and capacitance of the bus lines. Higher-value resistors consume less power, but increase the transition times on the bus, thus limiting the bus speed. Lower-value resistors allow higher speed, but at the expense of higher power consumption. Long bus lines have higher capacitance and require smaller pullup resistors to compensate. Do not use resistors that are too small because the bus drivers may not be able to pull the bus lines low."

An Introduction to Acoustic Thermometry (Analog AN131)

This is a good one. (Jim Williams is involved.) It shows you how to measure temperatures with an ultrasonic transducer and a jam jar. (Well, they use an olive jar, but jam jar sounds less exotic!) The whole thing is worth reading, and the application is a lot of fun.

The first interesting thing here was that they were doing pulse length measurement with a relatively low-end microcontroller and getting 100ns resolution. That sounds pretty good to me.

The second and maybe the most useful thing is the idea of gating to help with noise in sensitive measurements. They gate the trigger for detecting reception of ultrasonic pulses to avoid spurious pulses, which is sort of obvious, but they also do something less obvious. The ultrasonic transducer they use requires a 150V bias voltage, which they generate using a switching boost converter. To eliminate switching noise during measurements, they switch the boost converter off during pulse reception. They use a big output capacitor on the boost converter, so there's little loss of transducer bias during measurement. (I assume the current draw on the bias input is very low.)

I found that all very interesting. It might be something I might even try myself.

There were also quite a few interesting looking parts I want to learn more about: LT1122, LT1220 (op amps); LT1011 (comparator); LTC6993 (monostable); LTC6991 (clock generator); LT1072 (boost converter: 150V!). All Linear Tech parts, because it's an ADI app note...

Brushed DC Motor Fundamentals (Microchip AN905)

Brushed DC motors for beginners.

Mildly interesting things:

  • If the sketch diagram showing the design of a DC brushed motor is to be believed, the split-ring commutator in these motors is offset from axis of the stator poles, presumably so the motor can't get "stuck". Had never thought about that before.
  • Top tip for using a H-bridge: don't switch both MOSFETs on one side on at once...

Datasheet: Microchip MCP4725 DAC

This was mostly useful for the little glossary of DAC terminology it has. Pretty handy.

RF Basics Guide (Maxim AN5300)

This is a longish (55 pages) grab-bag of miscellaneous RF information for people who don't know anything about RF electronics. I found it useful, and it has lots of references to other app notes, some of which I'll probably read next week.