A smaller connector for the Chibi RF module. 20 pin .1" center connector is limiting the size of my sensor boards! I'd like a .05" single row, or .1" two-row, 8-10 pin connector. Configured correctly, it could be pin compatible with the 10-pin version of the AVR programming header, which would also save me the 6-pin ISP header.

Down side is (I think) Akiba has tried to create a standard 20-pin sensor connector for all his boards. If I am right, I will respectfully submit that he needs to designate 2 pins to supporting I2C.


I would also like a Chibi-like board for the Atmel ATA5428 433MHz radio. It uses about 1/2 the current of the 915MHz RF212 when transmitting or receiving, and the better path loss at less than half the frequency means better range.

Down side of this idea is that it distracts folks from standardizing on Chibi, and the ATA542X chips are getting harder to find.

Heh heh heh...pins 5 and 6 are I2C. Pullups are on the MCU side. For 433 MHz, I'd prefer the CC1101 which is still going strong. There's also a follow-on chip which is the CC430 and includes MSP430 + CC1101 for 433 MHz. The ATA5428 looks like it's either already discontinued or will be soon.

Here's a picture of the v1.0 of the universal serial connector that I'm standardizing on.


I'm going to release v1.1 of the connector will be coming soon and just trade some GPIOs for dedicated PWM and frequency counter pins.

Prob should add that the power consumption between chips needs to take into account the bitrate. The ATA5428 has a max bitrate of 20kbps while the RF212 has a max of 250kbps for 802.15.4 mode and 1 Mbps for non-compliant mode.

At 20 kbps, the ATA5428 will need to keep the transceiver on for 15-20x the time of the RF212 at 250 kbps (there's also some overhead that both will share so it's not exactly a direct multiplier). Hence the power consumption will go up proportionally to the longer time needed to transmit data. So its true the RF212 uses more power if both are kept on all the time. But in the actual use case where you sleep the nodes, the RF212 will actually probably consume less power.

I do agree that at 433 MHz, you will probably get better range though. I just don't agree with power consumption between two radios unless their bit rate is also the same.

Note that I said current, not power. My thought when I wrote that was that a joule-thief power system may be advantaged by a radio with half the current draw. Of course, using super caps or even a moderately large tantalum might fix that problem, but at a $ cost and size. I'm sensitive to size, as you can tell.

At extreme ranges of 915MHz, one might need to place the radio in a low kbps modulation mode, equalizing the transmission time.

I need to study the TI chips, but there are only so many hours in a day, and only so many hours I can steal from the night. I trust you will choose the right radio for us

I like your connector. It is much better than the "A" platform. Is it reasonable to have some GPIO's be optionally PWM and F-counters rather than dedicated?

QUOTE:
Note that I said current, not power.

Ha ha ha...duly noted.

For the connector, all pins other than power pins are optional GPIO. I'm standardizing the pinout just so that I can design add on boards that will work across different MCUs. The main reason I developed this platform is because I couldn't find a WSN platform that easily allows you to swap MCUs and radios. It didn't help that most WSN platforms are made by semiconductor vendors.

I want to stay away from having optional pins on the MCU side (other than GPIO) because someone (including me) designing a peripheral which requires a fixed PWM pin in a position where its optional will have compatibility issues. I'm still thinking more on it because it's pretty hard to get a standardized connector right. You always find some cases where you can't design something due to constraints you put on the connector. Hence the v1.1 coming soon.

As for the Arduino, I'm sure they have a much harder time. My connector pinout is pretty much just for wireless sensor types of circuits which makes the decisions much easier.