Post by thread:profiling

I use MPLAB to test run parts of my code for timing information. Currently I am coding RB0 to generate an approximate 38 kHz signal that is required for the IR detector. Testing this code segment under MPLAB reveals strange drifting within microseconds. If I "step over" the code as opposed to stepping through it, or setting a break point after the loop; the timer/counter within MPLAB reflects different profiling times. Sometimes its off by a few microseconds, other times by atleast 50+ us. Has anyone else noticed this? My guess would be that the PIC16F84-04 and its 1us instruction cycle isn't exactly "1 us", and MPLAB is aware of this, which profiles a different time for execution of larger pieces of code then single instruction step-through's. But without a serious hardware ICE, I cannot be certain. So does the PIC16F84-04 have an exact 1us instruction cycle, or does it fluctuate at certain times? Sebastian

The stopwatch is quite accurate in the simulator when you are running. It does exhibit some errors while single stepping in emulator or simulator mode due to the way breakpoints are set. If you want to time a routine, do not step through it, but zero the timer out at a breakpoint at the start, set a breakpoint at the end, and use Debug>Run>Run (almost like a Phil Spector song, if you're old enough to remember pop music)... -- ___________________________ | Darrel Johansen | | tempe, arizona | | spam_OUTdarreljTakeThisOuTprimenet.com | |_________________________|

NASA / NOAA Ninth International Winds Workshop (9IWW), Annapolis, Maryland, USA, 14-18 April 2008 Abstract Over 20 years of investigation by NASA and NOAA scientists and Doppler lidar technologists into a global wind profiling mission from earth orbit have led to the current favored concept of an instrument with both coherent- and direct-detection pulsed Doppler lidars (i.e., a hybrid Doppler lidar) and a stepstare beam scanning approach covering several azimuth angles with a fixed nadir angle. The nominal lidar wavelengths are 2 microns for coherent detection, and 0.355 microns for direct detection. The two agencies have also generated two sets of sophisticated wind measurement requirements for a space mission: science demonstration requirements and operational requirements. The requirements contain the necessary details to permit mission design and optimization by lidar technologists. Simulations have been developed that connect the science requirements to the wind measurement requirements, and that connect the wind measurement requirements to the Doppler lidar parameters. The simulations also permit trade studies within the multi-parameter space. These tools, combined with knowledge of the state of the Doppler lidar technology, have been used to conduct space instrument and mission design activities to validate the feasibility of the chosen mission and lidar parameters. Recently, the NRC Earth Science Decadal Survey recommended the wind mission to NASA as one of 15 recommended missions. A full description of the wind measurement product from these notional missions and the possible trades available are presented in this paper. http://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/20080017203_2008016805.pdf R __________________ I was looking for a map of global wind speed distributions.

Hey all, I want to compare the Eneloops that I got from Dealextreme with the ones I got from Dell to see if the DX ones are genuine Eneloops (there are very slight physical differences, but the performance never led me to doubt their authenticity). I figured a simple way to profile them was to measure under-load voltages while draining the cells with some resistors at around 500 mA. But this would mean current drain would decrease as the battery voltage decreases. 1. Would this still be a reasonable way to profile the batteries? Can I still mathematically compute the battery capacity using the graph? Another option would be to make a simple constant current source circuit that would ensure a constant current drain on the batteries. 2. Can anyone recommend circuits for making one of these

On 2/2/2011 5:15 PM, V G wrote: > Hey all, > > I want to compare the Eneloops that I got from Dealextreme with the ones I > got from Dell to see if the DX ones are genuine Eneloops (there are very > slight physical differences, but the performance never led me to doubt their > authenticity). > > I figured a simple way to profile them was to measure under-load voltages > while draining the cells with some resistors at around 500 mA. But this > would mean current drain would decrease as the battery voltage decreases. > > 1. Would this still be a reasonable way to profile the batteries? Can I > still mathematically compute the battery capacity using the graph? > > Another option would be to make a simple constant current source circuit > that would ensure a constant current drain on the batteries. > > 2. Can anyone recommend circuits for making one of these? > This months Circuit Cellar: http://www.circuitcellar.com/magazine/ ABOVE THE GROUND PLANE Parallel NiMH Cell Measurement by Ed Nisley (USA) Got batteries? You need to test them. Here you learn how to build an eightchannel parallel cell tester, with an Arduino Mega microcontroller board providing PWM DAC outputs to control the discharge currents, ADC inputs to measure the cell voltages under load, and serial output over a USB connection for data logging. p. 5

V G wrote: > I figured a simple way to profile them was to measure under-load > voltages while draining the cells with some resistors at around 500 > mA. But this would mean current drain would decrease as the battery > voltage decreases. That's not a big deal. Make really sure you have a circuit that cuts off when the battery gets low enough though, else you will damage it. You could simply measure the time from fully charged until the circuit cuts out to compare capacity of different cells. You may not know the exact capacity, but the comparison will still be valid. NiMH cells don't drop that much in voltage over most of their life, so you'll actually have a pretty good absolute idea of capacity too. If you really want to measure capacity, you can measure the voltage over time. Since you know the resistance, you know the current, then integrate that. Even once per minute would be good enough since you're talking about a discharge time of a few hours. ******************************************************************** Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products (978) 742-9014. Gold level PIC consultants since 2000

On 3 February 2011 11:45, Olin Lathrop <.....olin_piclistKILLspam@spam@embedinc.com> wrote: {Quote hidden}> V G wrote: >> I figured a simple way to profile them was to measure under-load >> voltages while draining the cells with some resistors at around 500 >> mA. But this would mean current drain would decrease as the battery >> voltage decreases. > > That's not a big deal. Make really sure you have a circuit that cuts off > when the battery gets low enough though, else you will damage it. You could > simply measure the time from fully charged until the circuit cuts out to > compare capacity of different cells. You may not know the exact capacity, > but the comparison will still be valid. > > NiMH cells don't drop that much in voltage over most of their life, so > you'll actually have a pretty good absolute idea of capacity too. > > If you really want to measure capacity, you can measure the voltage over > time. Since you know the resistance, you know the current, then integrate > that. Even once per minute would be good enough since you're talking about > a discharge time of a few hours. > > > ******************************************************************** > Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products > (978) 742-9014. Gold level PIC consultants since 2000. >

On 02/02/2011 22:15, V G wrote: > I want to compare the Eneloops that I got from Dealextreme with the ones I > got from Dell to see if the DX ones are genuine Eneloops (there are very > slight physical differences, but the performance never led me to doubt their > authenticity). > Surely the main thing of Eneloops is not Capacity. Many cells have 1800mAH to 2200mAH capacity. I'm a bit sceptical of 2,500 mAH and esp. of 2700mAH claims. The main thing is lack of self discharge. Many NiCds I've had are OK. EVERY NiMH I have had has self discharge unacceptable, 1,800mAH cells the best (maybe a month) and 2,500mAH about 1 to 2 weeks. If they are real Eneloops they should work after being in a drawer for 2 months. I can't see how else you can test. Capacity testing will only tell you the capacity!

On Thu, Feb 3, 2011 at 4:49 AM, Michael Watterson <mikeKILLspamradioway.org> wrote: {Quote hidden}> On 02/02/2011 22:15, V G wrote: > > I want to compare the Eneloops that I got from Dealextreme with the ones > I > > got from Dell to see if the DX ones are genuine Eneloops (there are very > > slight physical differences, but the performance never led me to doubt > their > > authenticity). > > > > Surely the main thing of Eneloops is not Capacity. Many cells have > 1800mAH to 2200mAH capacity. I'm a bit sceptical of 2,500 mAH and esp. > of 2700mAH claims. > > The main thing is lack of self discharge. Many NiCds I've had are OK. > EVERY NiMH I have had has self discharge unacceptable, 1,800mAH cells > the best (maybe a month) and 2,500mAH about 1 to 2 weeks. > > If they are real Eneloops they should work after being in a drawer for 2 > months. I can't see how else you can test. Capacity testing will only > tell you the capacity! > I'm just curious

Suspect Eneloop testing: What they all (or most of them said). And (i) Weighing cells often tells you much. (ii) Charge N cells to full charge. Make some of N "ordinary" NimH cells. Leave stand for as long as convenient. Probably 1 week enough. Charge cells again, measuring the mAh needed to regain full charge. Real and on-trial Eneloops should be low and similar. Normal NimH should lose more charge when standing. SOME std NimH do much better than others in my experience. LSD and non-LSD should be distinguishable. (LSD is in the sky with diamonds) (ii) Discharge tests Note that this test is NOT about capacity per se but about curve shape- Eneloops claim to be different. See below. 1 ohm will give close enough to 1A discharge across the range as to be considered constant for this purpose. Charge batteries fully. Measure and record O\C voltages. Starting simultaneously, apply 1 ohm load (2 Watt rated) and measure battery voltages immediately and at N minutes breaks. N is up to you. If I was doing this manually and had up to 8 cells and was going to sit in for the duration I may make N=1. ie 1 minute readings. This would tell you interesting things. BUT 5 minute steps would be good enough until you neared end point. Set an auto reset timer to 5 or whatever minutes and read when summoned. If you haven't got an auto reset roll over timer get or make one. You will find much use for it over the years. Take cells off discharge at predetermined endpoint voltage - maybe 1 Volt.Look at spec sheets and decide. Eneloops re liable to have lower mAh per mass than non Eneloop. LSDs may be real LSD but clones. Note that Eneloop claimed voltage per time discharge curves CLAIM to give higher voltages than std cells. IF this is true it will stand out for real eneloops and maybe for proficient clones. (iii) The chances of them being genuine are only modest. The chances of them being proficient clones are fair. The chances of them being bad clones are fair. The chances of them not being LSD at all are finite. Russell McMahon Applied_Technolog