Advertisement
Advertisement

dt_hayden

's profile
image

Biography has not been added


dt_hayden

's contributions
  • 01.24.2008
  • 8-bit microcontroller implements digital lowpass filter
  • Somewhere in the cloud exists an EDN design idea I submitted back in 1995 (I think) which used this idea and took it one step further. In order to get fine resolution of the low pass cutoff frequency, I kept the filter coefficients (1/16 and 15/16 if I remember correctly) constant, and varied the time sampling of the data using the onboard timers of the 8-bit controller. This allowed (0.4%* if I remember correctly) resolution in the filter cutoff frequency. *This could have been much smaller but the target system was battery powered and I elected to run the microcontroller from a 32KHz watch crystal to minimize current consumption, and 0.4% resolution was plenty. The microcontroller measured the voltage across a diode as a crude temperature indication, and compensated the temperature sensitive pulse response of an analog meter. The low pass filter cutoff operated around 1 Hz.
  • 02.13.2018
  • Teardown: Amazon Dash Wand with Alexa
  • I'm surprised the magnets are not glued in place, perhaps gluing metal to plastic is not robust enough? (I'm using 'glue' in a liberal sense, including a web of RTV sealant.) I'm always interested to the method of getting battery power to the PCB in portable electronics. The metal tabs used here to solder to the PCB look pretty robust, are the springs then welded on to them or just wedged on? Are the batteries in parallel? It looks like the batteries are installed in the same orientation and the far ends of both batteries returned to the PCB using the long metallic tabs.
  • 01.22.2018
  • Teardown: A19 LED bulb
  • Feit Electric is complete garbage in my opinion, I had extremely poor reliability results from their CFL bulbs (also from Costco). As far as the 25K hour lifetime, it is quite common for companies to take the rated life of a single LED itself and apply that to the entire assembly, regardless of standard reliability calculation methods. Very misleading claims, but hey, that's marketing. On the early Philips bulbs I bought, I figured they were lying about expected life but were planning on bulb prices dropping so much over time that any replacements (probably very few - who really takes the time and expense and maintains documentation to return a 3-4 year old bulb?) would be covered by the high prices of the early bulbs.
  • 12.30.2017
  • Reflecting on Apple's latest products: Are you impressed?
  • Although I have never been an Apple fanboy, and not a big gadget freak in general, I do have a new Apple TV 4K, obtained "free" with 4 months of prepaid DirectTVNow service. My primary reason for wanting to try the ATV4K was the application support of Zwift, a virtual cycling trainer/race platform/game. The touchpad remote for the ATV4K is horrible, and I don't use voice enabled (SIRI) services which seems to be the primary intended interface, but luckily the older Apple IR remotes are compatible so I bought one from Apple for $20. The Dolby Vision image quality, when it's working, is astounding with my LG 4K OLED TV. However, the ATV4K device so far has proven finicky and unreliable, needing constant resets and continuous logins to Apple's cloud service. Based on my overall experience, I'm still no fanboy.
  • 12.12.2017
  • The pros and cons of mesh networking
  • I'm enjoying reading your experience with Google's mesh Wi-Fi networking. My household situation could probably benefit from a product like this but I'm still uncomfortable giving Google access to everything and permission (theirs or a hackers) to update or make changes at their whim. I have a Zwift trainer bike setup in a 3rd bedroom that is very sensitive to Wi-Fi dropouts, and occasional Wi-Fi access needs in the small office upstairs in my shop that could benefit from a local Wi-Fi meshed beacon. Did you have any similar concerns and if so what has changed your mind?
  • 11.28.2017
  • Cancel PWM DAC ripple with analog subtraction
  • Nice idea, one of those ones that you see and think it would have been thought of long ago. I had a design that replaced mechanical potentiometers with PWM DAC for very sensitive calibration of operating frequency. I had to heavily RC filter the PWM output for stable operation, requiring fairly large C. This idea looks like it would have been much more elegant solution.
  • 11.08.2017
  • Deriving a rule of thumb
  • Good info! I have derived this myself, after using it many years and wanting reassurance that I was not misapplying it.
  • 11.10.2017
  • Are “beam array” headlights in automotive’s future?
  • The diagram above shows one driver servicing many LED via controllable switches, so probably not a 1:1 ratio between LEDs and drivers. Also, if the failure mode is dominated by LED degradation (supposition, not a claim) rather than driver reliability, increased drivers could have lesser impact on reliability. Then there are the wiggle words of 'similar to conventional', a target number would be better. In any case, I hope they do something, I'm tired of the dazzling lights.