Did you know that there are cases where you can add noise to your signal to get more resolution? It is referred to as dithering and works well, provided a few conditions are true in your system. (This explanation assumes you understand averaging/filtering to smooth data and what an analog-to-digital converter does.) • If there is time to average multiple samples • If the signal is quiet relative to ADC resolution • If there is a way to generate and combine a noise source with signal before the ADC sampling. The ADC would read these (again truncating) as: 22 23 22 22 22 23 23 23 avg = 22.5 So eight measurements will have some 22 and some 23, with the ratio close to the fractional value. In this example I only used values of 0 or 1, but a wider range of values between and shaping the noise distribution can further improve the estimate. The smoothed value (purple) of the noisy samples (green) is closer to the actual value (blue).
In practice, any random or pseudorandom source will work. 4x oversampling is enough to get a benefit, and 8 or 10 is quite good. If the signal is noisy more than 1-2 LSB (as it usually is), then there is no benefit to adding noise. Just average as you would normally to get better effective resolution. This article clarifies each parameter that goes into filter design, and provides an analytical solution to calculate k value.
Site of Jeff Steidl. This is the same equation as that for an which is a highly useful and low-computation smoothing technique similar to, but generally better than, an N-point average.
A starting point for k is 0.5, which is roughly equal to a 4-point average.] The formula for a first order lag filter is fairly simple: new_filtered_value = k * raw_sensor_value + (1 - k) * old_filtered_value The 'magic' is determining an approprate value for k. Calculate k To calculate k, enter Tr, Ts and Frac from above into the following equation: k = 1 - e ln(1 - Frac) * Ts / Tr For details, see. In linux, process an incoming serial data stream (from program a.out) which appears on screen (stdout) and route to a file at the same time../a.out tee >logfile.txt It is not smooth, as the terminal is buffering data. To change that, use unbuffer./a.out tee >logfile.txt To use unbuffer, you must install the 'expect' package* References *to get it on Mac unix, the package installer recommended is homebrew. 1) install homebrew if you don't have it. Ruby -e '$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/master/install)'. Fourier Series how to derive series waveforms more detail on pg 324 of this pdf harmonic content of PWM double-edge vs.
Nov 22, 2017 Product: Driver Rider Backrest Pad Cushion For Harley Davidson Touring 1988-2017SKU: GB1151287Available options: >>>Please let us.
Leading/trailing edge PWM harmonic content from the performance of the three bipolar PWM methods can be compared on the basis of their spectral models (2.11), (2.13) and (2.15). To that end, note first that the carrier harmonics and their sideband harmonics produced by trailing-edge bipolar PWM are equal in amplitude to the corresponding harmonics in leading-edge modulation. Hence these two methods have the same performance as far as harmonics are concerned. To compare them with double-edge bipolar modulation, the amplitudes of the first three carrier harmonics (m = 1, 2, 3) and their sideband components are tabulated in Table 2.1 for n = ±1,±2,±3. The comparison shows that double-edge modulation produces far fewer harmonics. In other words, double-edge modulation eliminates all even-order sideband harmonics of even-order carrier harmonics, as well as odd-order sideband harmonics of odd-order carrier harmonics.
The reduced harmonic content is a significant advantage of double-edge modulation and makes the method more attractive for AC–DC and DC–AC converter application. First off, my list is of dual op amps, because in virtually every design, there is a requirement for two or more, and if there's a spare, there is some other function that can be enhanced by using it, typically as a buffer or a filter. Besides, two in the space of one is almost free. (There are also opportunities for quads, but there are fewer choices, and often a quad is not conducive to a good pcb layout. The Mentalist Season 5 Episode 14 Torrent. ) A good op amp for me is one that has enough speed for my typical application and still maintains precision. Very general guidelines for applications from DC to 1 MHz.
The favorites meet or beat all of these requirements, and have been used successfully in production designs. GBW >10MHz and 5V/us Ibias = 10V Standard pinouts Rail-to-rail operation... AD823 AD8676 LT1498 LT1804 LT6244HV To replace an LM741 or similar, almost any general purpose op amp is better with regard to performance, e.g.
Bandwidth, slew rate, bias current and offset voltage. Here are some common and inexpensive good choices. Note that these are not rail-to-rail though, and are less suitable for lower voltage applications due to requirement of headroom between min/max output and rails. TL081 (TL082 and TL084 dual and quad) LTC6082. It took a lot of discovery, but, here are my reminders to set up a new xbee board. • Get the right xbee.
('Series 1' for plain point-to-point connection) • Before trying to find the device (using XCTU.exe), be sure the baud rate matches; it must in order to connect. If you change the setting, you must delete the device from the list, and re-find and connect, with the new setting for baud rate. • Set the PAN ID to 4 • Disconnect the Rx and Tx from your mcu or pcb even if it's off; otherwise you won't be able to link it with the XCTU configuration program.
Its output voltage follows the input voltage, meaning the output is ideally the same--with the caveat of a voltage offset.* It doesn't amplify or attenuate the voltage, but its high input impedance and low output impedance act to isolate or buffer the incoming signal from downstream circuit loads. *When using the follower with an NPN or PNP transistor, the output will be offset by a silicon junction drop of about 0.7V. The offset amount will change with temperature as the junction.
In many cases this has little or no impact on the application. To drastically reduce the offset and the temperature variation, add one transistor. The Dual Complementary Emitter Follower The solution is to introduce another circuit with the same temperature dependence but moving in the opposite direction. They will tend to cancel each other out. Also, if the circuit could add the opposite offset voltage, the buffered voltage would be back where it started; a Voltage Follower. Here's a way to do it. The input resistor is not needed in theory, but is useful in practice to provide some protection from spike energy and to current limit under fault conditions, with no negative effects.
Start with the standard emitter follower. Then add a 2nd follower, but using a PNP transistor. This addresses both issues. It doesn't matter which one comes first for operation.
But to maximize headroom, if your voltage input is closer to the supply, drop it with NPN, then boost with PNP; and vice versa. If the PNP and NPN transistors are similar in specification (size, power, bandwidth), they will track well. If they are located near to each other and have similar thermal environment, they track better. Even better is to use a (BC846B or similar) that contains one of each. Pictured here is the response of the & ) over a range of temperatures (-40 to +80C) that closely matches the simulation data pictured here.
I've gotten similar performance on the bench with this circuit. Zoomed in, we see that the spread over four temperature steps is large for the 1st stage (green), but very small (aqua) with the 2nd stage buffer added, as compared to the original signal (Tan). Also the output voltage offset is reduced from 0.7V to ~0.1V The discrete solution shown is a very rugged circuit and usable over a very wide range of input voltages, limited by the selected transistors. As shown, it easily works with a supply of 5 to 36V. For higher voltages, I like to use the similar parts: 2N5550 or NPN and PNP which can go up to 150V.
This is a favorite circuit for its simplicity and performance; it's a good illustration of real-world transistor application. I just received some new Teensyduinos for an ongoing project. They happen to be the new version 3.2. I tried two of them on windows, and then on Linux. However I cannot see the devices. Should be software-equivalent.
On Windows, neither shows up as a com port on Windows. On Linux, it shows as 'Teensyduino raw hid'. Tried swapping to T3.1s that I've been using; they work on both machines. I discovered indirectly from troubleshooting guide that you must - Open the Teensy Loader program - Then press the Teensy's reset btn Then all works fine. It's a 'first time' startup thing. I've never noticed this in the past. Apparently this applies to v 3.1 and v 3.2.
• Rate of pulses affects speed of motor response. My maximum was with an interval of 8-10 ms. • The exact time for min and max and actual angle of rotation vary for different motors, and usually need trimming to offset the exact pulse durations and a measurment to find angles. • Motor rotation rates vary a lot and are specified in time per 60deg turn. • 0.15 - 0.25 is typical; less than 0.1 is actually 'fast,' although lots of parts are referred to as fast are 0.18 - 0.2/60deg. 0.06 is the very fastest I've seen on more expensive, fast servos, at their highest input voltage.
Price jumps hugely from $15 to $200 for ones I've found so far for a seemingly small speed increase. (ex: ) Programming Directly coding using time will work. Here is the code I used to try out control; it sweeps the value up and down to spin motor from ~min to ~max position, with a little margin near edges. To control more than one motor, use one of the RC Servo libraries (such as the one supplied with Arduino examples). Higher resolution control with library.
To include the compile date or time in your program (as a way to keep track of versions), I discovered that there are compiler constant strings for this in Arduino's compiler. To include in your arduino code, use these lines near the top of your program. // get date and time of compile. Char compileDATE[] PROGMEM = __DATE__; char compileTIME[] PROGMEM = __TIME__; Then, to use it in your program, include the array in a print statement. Serial.println(compileTIME); Prints: 20:38:56 Or, in a formatted print string using sprintf(), char txt[40]. CompileDATE[3] = '.'
; compileDATE[6] = '.' ; // replace spaces sprintf( txt, 'Date%s Time%s', compileDATE, compileTIME); Serial.println(txt); prints: Date Oct.20.2015 Time 20:38:56 I use the date in a splash screen msg on my OLED display also. Handy to view when not hooked to a computer. ( shown elsewhere.) outputOLED(' ', 1, 3, 1, 1); outputOLED(compileDATE, 4, 3, 1, 1); outputOLED(compileDATE, 4, 3, 1, 1). Discovered that the following print optional argument will output integer numbers in various bases. (As of Ver 1.6.5) Serial.print(nnnn,base) where nnnn is an integer, and base is an integer, In addition to base 8 ( OCT) and 16 ( HEX), it can even do other bases such as 3 or 2 ( HEX) or even 7 (should you think of a reason!). Base = 7; Serial.print(millis(),base); base = 2; Serial.print(millis(),base); Outputs in base 7 and base 2 respectively 10111000 Didn't know you could do that.
--- 3/15/2016 Addendum: if the first number is a floating point number, then the second argument is the more conventional number of digits after the decimal. Serial.print(xxx.x,nn) where xxx.x is a float, and n is an integer. Then the device should (and did) show up in Windows 7 device manager and as a comm device in XCTU software when selected first tab/button 'Add a radio module.' After a few seconds it discovers the device and shows up as a box in list with Port definitions and MAC address. I was planning to start directly with two arduinos and two Xbees. However, there is a built-in terminal to allow direct communication. I proceeded using it.
The manual shows communication check with adjacent windows from two separate computers. I duplicated the setup on a 2nd PC. The next step is to configure the device PAN ID (personal area network). Manual says a pair of devices must have same ID to communicate.
I set it to 4. I set the same value on the 2nd system.
Select the Terminal tab; define a packet by clicking '+' and then typing some text. I switched from ASCII to HEX tabs and added 0A to include a linefeed.
Further reading says to set addresses for MY and DL. If all at DL = 0, should work, but doesn't. Says MY must be set as well, but it is Read-Only. Only reference to this is a blog post saying MY is read-only for XBee 2, but not 1. To verify PC-to-Xbee connection, the AT command +++ can be typed; it should respond with OK.
++++ OK This did work. (Seems you have to touch the Baud rate function first in some cases.) Couldn't make additional progress beyond this step. Decided to contact Digi. Have to register in order to send a tech question. -- tech support helpful. Turns out there are several flavors of the modules. The units I started with are not the right choice for a direct connect, nor are directly compatible; in the lingo, there are Series 2 and original (now called Series 1).
The Series 2 are more suited more mesh networks and have much lower effective data rates. I was able to configure by setting one xcvr as 'coordinator' and is sufficient to send some data. I ordered a pair of the right ones XBP24-ASI-001 (decoded means XBee Pro @ 2.4GHz Series1/802.15.4, version with SMA antenna connector). Necessary steps with software are still to set to matching PAN ID. It is optional but not necessary to set the MY and DL values to complementary values on the pair. If both are 0, every device gets everything.
Else, suggested values are 10, 11 and 11, 10 respectively. No mention of range of other value pairs.
I left values at zero. After seeing transmitted and received messages show on the respective terminal screens (yea!), I proceeded to set up a repeated message, possible in the terminal. I carried the laptop and xbee through the building to confirm that it maintained communication at 50-100 feet through a couple of walls. The data stopped appearing in the terminal window after a few minutes. Tried clearing the window with the provided button (but it didn't), and eventually found that disconnecting xbee USB and restarting XCTU were necessary (but not always sufficient) to re-establish connection.
Long story short is that there are limitations with terminal window: after 4000 characters are sent (~2-3 minutes at 500ms repeats), it stops updating in text display window, though communication and data transfer are actually still functioning. Also, a bug in the program was revealed: the Clear Buffer button only works if the received data is less than its maximum. Once it's full, the button no longer works.
I found no way to recover from this. But once I knew, all was well. One other problem appeared soon after; some glitch occurred when I read back data from XBee, and nearly all values are scrambled and appear red. Descargar Driver De Joystick Eurocase there. Couldn't clear by re-writing defaults. Resolution was the usual de-power the board, and restart everything.
Problem did not re-appear. Finally, I got this screen frequently: 'Action Required.' The suggested troubleshooting choices did not apply in my case. It is caused by having the Rx,Tx connected to microcontroller when the Add or Discover Radio module is pressed. I had to physically disconnect serial lines (Rx,Tx) in order for XCTU to find it.
Needless to say, this is easy to forget about if you close and relaunch. I do not know why this would be necessary.
Suitable arduino software based on echoBoth.ino receives and re-transmits text from either end. 1st arduino code.