| Products > Test Equipment |
| Unknown GPSDO but BG7TBL logo on the circuit board and OCXO ! |
| << < (19/21) > >> |
| Noy:
Yes i know. But i don't want to power up the sva to use the DG. But maybe you are right and the ref in/out is working without powering up the sva ( sva has not a real power switch...so maybe it works will try.) But for now first part will be to configure the neo9 to work with the remaining atmel / programm the new one. One advantage to use a new one will be i can use nmea 4.0 default output from neo9 and i can control the "stable" voltage border by myself. Does somebody know the original voltage at which the atmel says "OK"? Forgot to measure it before surgery.. |
| Noy:
Lift the Atmel today and mounted a new one. Now i have to program it. One thing i noticed, the pin 2 is connected over R33 to the 1khz output (R30). Now i am quite unsure whats they will do with this pin? Measuring the frequency? Thought they are measuring the control voltage every x seconds and if there is no change anymore its stable? Or maybe both? The neo9 is configrued in his flash, so i only need to rebuild the "supervisor" stuff : 1. led green off, red on 1. power GPS on 2. GPS fix reported on uart 3. Led green on 4. is the frequency stable (voltage + frequency?) or how can i measure stable + correct frequency? 3. Led red off I connected the output today with old atmel + working neo9 to my counter and it works. But i get sometimes a short frequency "hop" 10.000.000,00 to 9.999.995,2 and intant back to 10.... What can this be? Hopefully its something "old atmel" or no case related... Would it be good to add a cap to the controlling voltage pin at the ocxo to smooth small / fast voltage changes? Or at the opamp input? The atmel is not controlling anything else or? Like you can See i ripped accidantly the UART traces off.. (Pulled on the connected jumper wires :-( but its fixed instant glue ;-)) The other atmel pin which outputs 1.001khz Is not connected anywhere.. Don't know whats going on with this pin.. |
| TurboTom:
In the former version (without the dedicated decimal counters), the Atmel's clock was supplied by the 10MHz OVCXO and the programming arranged an OSC In -> PWM Out 10'000:1 divider. I guess this hasn't been removed from the code when the HC390 decade counter have been added, so it's still available at the corresponding pin, though unused. The Atmel isn't doing anything than configuring the GPS module and monitoring the "stability" of the PLL phase comparator output to identify a lock (ad signal it via the LED). Pretty lazy life i'ld say... ;) |
| Noy:
Not sure about that. Check the pictures from post #60 In the 2019 PCB version pin 2 connection + resistor r33 isn't there.. In my 2020 Version its added again. And they advertised it as e-GPSDO which should be more accurate than the PLL gpsdo (never proved this marketing. But seller said slightly other bom and Software to be Mord accurate.) How is the stability monitored? Its measuring the voltage and looks for what? Voltage drifts/hops? Maybe they are measuring frequency with pin2 and then they are switching analog input into gpio / pwm and manipulating the compator output voltage to regulate it or something else? |
| TurboTom:
You're probably right with your assumption regarding pin2. It seems, the output of the frequency divider is fed there. But I don't think it's meant for frequency measurement (which would require the MCU's oscillator to be more stable than the OVCXO). If the GPS 1kHz signal is also supplied to the the Atmel somewhere, the phase walk between the two signals could be measured, and this will probably be a more accurate means to identify the "Locked" condition than looking at the A/D-converted phase comparator filter output. If the GPS-supplied 1kHz signal isn't found at the Atmel, I'm running out of ideas... ??? ;) |
| Navigation |
| Message Index |
| Next page |
| Previous page |