Friday, April 5, 2013

Code Mostly Ported and Other Things

image

I’ve been working in the background on things that most people will find dull – more coding. No major breakthroughs, but no breakthroughs are needed at this point. I’ve ported about 65% – 70% of the code and what is left to do is below. More after the break.

  • Write a few functions to average torque + ancillaries
  • Going to stick to basic power profile to begin. I’ll add in Crank pages and the TE/PS page afterwards. However, I’m not seeing an added benefit to Crank pages except greater resolution (basic is 1 watt, crank depends on the rotational speed so it can be more accurate). I’m also unsure how I want to calculate TE/PS – Should it be done on the last revolution or averaged since the last TE/PS send?
  • Evaluate Gyro’s – Decide on a magnetic switch or gyro – I forgot to order supporting electronics so I can’t test the gyro yet. It uses a PLL that requires a filter circuit that I overlooked. I’m also looking at the AH1806 Hall effect sensor. Unlike a reed switch, that mechanically moves to make electrical contact, the hall effect detects magnetic fields. I dislike the little glass reed switches, and this would be surface mountable to the PCB.
  • Sort out using the ADS1247 internal reference
    • I was getting a lot of instability with this, however I almost wonder if I fried it wiring something wrong or setting a register incorrectly with something wired – the upside to using this is that the ADS1247 then powers the strain gauges directly and turns itself off when it is not doing a conversion.
  • Measure current of nRF51422 EK board and everything connected (ADS1247 + strain gauges). Then decide on a coin cell or lithium ion rechargeable. – I’m in the coin cell camp, but my last prototype forced my hand.
  • Design PCB – based on gyro or not – pending above choice
    • I am trying to have provisions for the ADS1248 which has two more inputs. These could be used to do what Pioneer is doing – vector data – but that would be down the road
  • Order PCB
  • Assemble PCB
  • Design enclosure – Have it rapid prototyped
  • Tune the Trace Antenna (UGH!, anyone who’s an RF engineer and interested in helping with this drop me a line! kwakeham@gmail.com)
  • Order new crank(s)
  • See if there are any 1000 ohm strain gauges that are suitable to replace the 350 ohm ones – this would drop current consumption which may be needed.
  • Instrument new crank(s)
  • Put it all together, test it myself
  • Build 5 – 10 of these for the beta test
    • Low cost circuit, but strain gauges and the cranks are high cost investments that I need to absorb.
  • Upgrade to ANT+ Alliance member – go to symposium in Sept in Alberta. Mainly for fun / Vacation. Maybe Ray will make an appearance again for another keynote?
  • Long term, evaluate carbon cranks. From what I’ve pieced together is that most manufacturers are now using uniaxial CF weaves which aren’t as susceptible to non-linearization of the stress-strain curve with higher usage factors. I’ve been keeping an eye out for a heavily used carbon crank to instrument to see the stress strain curve it produces. However people treat these as almost as valuable as a new one. I have a hard time paying 160 dollars for a crank I can get 200 dollars new. Looking at ordering a SRAM Force GXP and riding as much as I can all summer to measure before an after.

So around the web the "RFduino” was launched on kickstarter. What is this? Well it’s is certainly not the traditional Arduino. It doesn’t use an AVR Chip. Instead it used the (drumroll please) nRF51822. That’s the BTLE version of the one I plan to use. The main difference is that the 822 is BTLE but I think (correct me if I’m wrong) that it can also be used for other wireless protocols. The nRF51422 is a drop in replacement! You know what that means? Reproduction of the open source board should allow for an ANT+ based version (ANT+ Network key access required via ANT+ Adopter / Alliance membership).

Sadly I’m posting this way to late. I only found out about RFduino a couple of days ago and that was at the end of the kickstarter. I did pre-order one. The upside is that I am hoping that this will kickstart (pun very much intended) the BTLE version of my powermeter. Not so much on the PCB side, but on the coding side. If some other savvy individual were to build a BTLE powermeter library then I could piggy back on their work and combine it with my ANT+ and voila! – BTLE version! Not certified of course, but I’m still building prototypes. I’ll sort that issue later.

Estimated delivery of RFduino is July – I’ll have my V4 prototype meter up and working long before then.

No comments:

Post a Comment