The Pi in the FiredogSDR can run 8 channel WD
|
Rob Robinett
Despite my vow last night to not work on the FD until its serious RF input problems are solved, this morning I decided that I had spent so much time setting up access to Tom NH6Y's FD that I would spend an hour to see if wsprdaemon (WD) could be installed and run on it. Surprisingly, it took only about 20 minutes any many automatic package installations to get WD installed and running. There are plenty of CPU cycles to support 8 channel mode, but the Pi CPU is getting pretty warm at 71C. So the FD would benefit from a better heat sink+fan. The FD is posting fewer than 1/4 the spots of a local KiwiSDR which is attached to a better antenna, so I don't know how much of that difference is due to the FD's RF input problems and how much is due to the Kiwi's better antenna. Hopefully we will get the chance to feed the same RF to the FD and Kiwi and quantify the difference in performance. |
|
|
|
Is that a typo for Flydog, or sarcasm, or is there actually another SDR called Firedog? I looked but came up empty?
|
|
|
|
Rob Robinett
It was a typo. I was working on the FlyDog On Sat, Jul 24, 2021 at 9:34 AM VE3VXO <ve3vxo@...> wrote: Is that a typo for Flydog, or sarcasm, or is there actually another SDR called Firedog? I looked but came up empty? --
|
|
|
|
I suspected that but wasn't sure. I don't expect to see my FD arrive till September probably but in the mean time if you have any chance to write down the steps you used it will help me a lot and would be much appreciated!
Best regards...joe |
|
|
|
Rob Robinett
Just log on to the Flydog's Pi as user 'flydog' password 'flydog-sdr' execute 'sudo apt-get install git' then follow the installation instructions on my github page. there are many, many packages installed when you first run WD, and you may need to re-run once or twice to get the WD installation to complete. in your conf file specify the IP:PORT of your kiwi as 'localhost:8073' On Sat, Jul 24, 2021 at 9:51 AM VE3VXO <ve3vxo@...> wrote: I suspected that but wasn't sure. I don't expect to see my FD arrive till September probably but in the mean time if you have any chance to write down the steps you used it will help me a lot and would be much appreciated! --
|
|
|
|
That looks pretty straight forward and yeah I already edited my local copy of your example .config file so I wouldn't forget to set localhost:8073. Well I guess now it's just hurry up and wait time.... I guess in the mean time I can wire up a regulator. Do you know if the FD will like a little generous 5V like the Kiwi does or should I go for 5V bang on? Thanks for your advice. Joe ---------- Original Message ----------
|
|
|
|
Rolf Ekstrand
Rob
Just curious about the "serious Rf problem" on the FD cape. I took a peek on the website and most certainly it looks to me that their enclosure is not optimal when it comes to cooling the raspi cpu. 71 deg is pretty close to the point where the raspi will start throttling down the CPU. I am running one of my rpi 4 at 2Ghz at 50 - 60 deg, but with lots of cooling and a big heat sink. On the other hand my rpi 400 is running at 2.2 Ghz as is and it seldom gets over 60 deg. One solution is to put an extension spacer between the FD and the raspi so you could fit in a larger heat sink and design the enclosure with the fan mounted on the side so that you put he majority of the airflow between the cape and raspi. Well, it's an easy fix, but first what are the RF problems? Rolf |
|
|
|
Rob Robinett
On the Kiwi I have been measuring the voltage on the USB-A port of the BBG and then adjusting the PSU to get 5V on that unloaded USB-A port. I have found that the the resulting PSU output is typically 5.5V under that condition which is when the Kiwi is running. I recall that there is DC loss in the PSU to Kiwi cable, and then something like .25V in the DC input filter on the Kiwi board. The cable losses were even worse when I used cheap 22ga DC pigtails. The BBG is speced to run from 4.4 to 5.6V and is protected against overvoltage to I think 20V. I have accidentally connected the Kiwi to a 12VDC PSU without damaging it. So adjust your PSU while connected to a running Kiwi and measure someplace on the Kiwi, not at the PSU. On Sat, Jul 24, 2021 at 1:06 PM VE3VXO <ve3vxo@...> wrote:
--
|
|
|
|
Rob Robinett
As I understand it there is at least one RF problem because they have two RF inputs, one with a 30 MHz LPF and a second with a 60 Mhz LPF and the filter outputs are both always connected together at the input of the RF amplifier. That results in severe interaction between the filters and a resulting dip in frequency response around 10 Mhz. The Kiwi's RF input amplifier design is already a weakness and this double filter amplifies the problems. Since the HW schematic is not open source and early owners report 'alpha-looking' hacks on the difficult to modify RF section, it seems the FD is really at an alpha HW development stage. I will be testing an unmoidfied FD against a Kiwi with both fed the same RF, so in a week or two I may be able to quantify the performance difference between them. Perhaps the HW problems I describe won't significantly affect WD sensitivity. On Sat, Jul 24, 2021 at 2:14 PM Rob Robinett <rob@...> wrote:
--
|
|
|