Neues Projekt Hilfe?
  • Projekte
  • Zeitraum
  • Badge
Neueste Kommentare
mfc
mfc vor 4 Stunden
No comment from me as two days after my last post I travelled to Spain for some caving (and sucessful bat detecting) and having returned still working through the backlog.... I fully intend to test what happens when adding the decoupling capacitor to the 'new' MEMs BoBs .. but it will be weeks before I get everything out to test that.
Bat detector with amplitude recovery
DON W
DON W vor 5 Stunden
That certainly does make sense. And this was a regulated supply, even though the clock itself has its own internal regulated power system. I've never seen one quite so unstable, to where my DVM was in constant change..I didn't bother to scope the output, but I can imagine how it would look. So far, with a new supply, no further issues.  
6-digit VFD Clock with ESP32
TomR
TomR vor 1 Tag
I haven't seen any comments on this project in a month so I can't tell if any progress has been made on the issues that were brought up. Maybe they were moved to a different forum that I haven't seen. In the meantime, I have been looking into possible explanations for the variability in sensitivity reported among users. I have focused on two causes:The sensitivity of the MEMS microphone to the ultrasonic signals emitted by bats. The frequency response of the Knowles microphone included in the kit is not specified above 10KHz. Thus Knowles is not required to report it or even control it during manufacturing so both the part-to-part mean and standard deviation are unknown to us. Since the echolocation if bats is on the range of 10KHz to 200KHz, we must assume this is a kit-to-kit variation. Generally, devices manufactured in a given lot will have little part-to-part variation, but larger lot-to-lot variations are common. Knowles makes a version of the same part number that is documented for typical operation to 80KHz, but using it requires a different PCB layout.The large input offset voltage of the TL062 opamp as reported by TonGiesberts.  The C version of the device is specified to have a maximum offset of 15mV at room temperature, with additional variation over temperature and aging. This is high enough to cause a decrease in sensitivity of the Schmitt trigger IC3B.  The AC version is specified at 6mV maximum - an improvement. Switching to a device that has been laser trimmed for low offset voltage given a further improvement. The AD648JNZ is drop-in compatible to the TL062 and has input offset voltage specified to be 1mV maximum.  The cost of this device is much higher however.  There are other devices with very low offset voltage available, but they tend to be single-supply and in surface-mount packages, These could be used but will require a socket adapter and re-design of the biasing.
Bat detector with amplitude recovery
Axiris
Axiris vor 1 Tag
We've seen power supplies that can't cope well with a (substantial) change in load. E.g. when the device draws more current, the output voltage of the PS drops, and as a result the device draws even more power, etc. until the device powers down. Since the PS is still active it then could produce its normal voltage output, and the device may power up again. In some cases, a periodic power-cycling occurs. From the device's point of view, it's like plugging out and plugging in the PS at a constant pace. When the clock goes into sleep mode, it draws considerably less power as the VFD tubes are turned off. When the clock awakens, the clock fires up the VFD tubes and the load on the PS goes up. At this point the issue may occur. Good luck!
6-digit VFD Clock with ESP32
Mathias Clauszen
Mathias Clauszen vor 1 Tag
Hallo Russin Sir Bullar, about the non worling oled, i think they kindly asked you to contact costummer service so they may can arrange a new OLED for you. About the firmware, the arduino IDE is for precompiled hex files simply not build, and far from perfect ( won't say it's one of the worst things to programm an avr but it's defnitly in the top 10 ). Besides this some more about the way to get new firmware to the station: The atmega32u4 has a leonardo compatabile bootloader onboard. If you connect to the USB device with 2400 baud, the bootloder will get started and the board, afterwards the avrdude can send new firmware to the board. Thats what the given avrdude line is basically doing. On a linux system that won't be COM1 but /dev/ttyACM0. For teh faulty OLED, if the avr reacts to serial commands this simply means the OLED was dead on arrival and you need a new OLED. For the soldering station you also need to be aware that the display is not running in i²C mode. The station uses the SPI mode with following pin defines:OLED_CS 12OLED_DC 14OLED_CLK 15OLED_DATA 16OLED_RESETso you need at least four channels on your scope for data analyses. But seeing signals on SCL and SDA is expect the avr is sending data to the screen. Also note there is no way that the avr gets any response if the display actually got any data send, meaning, if the display is dead it will simply not able to detect if it is showing anything or not. For the serial communication you can also use CuteCom ( Ubuntu has it as packet ) and connect to the station. On Windows it will be plug and play, on ubuntu you have to deal with the user right for the serial ports and also in some versions arround 18.x there are some problems that the user will get no permisson to serial ports, even if he is in the dialout group ( something in the udev rules was broken ). If the connection is possible you can at least rule out tht the avr may is in some bad state. Also you have on the PCB a 6 pin connector ( 2x3 ) near the AVR. This is the  ISP header, so you can attatch a avr programmer of choise and flash the hex file directly to the mcu if you like to. I hope this will at least help a little. I wish you a nice weekend and i hope you'll have a working station soon. Best Regards Mathias  
180348 DIY soldering station