Feedback on possible KerberosSDR V2 Hardware

Discuss KerberosSDR - 4x Coherent RTL-SDR
rtlsdrblog
Site Admin
Posts: 2425
Joined: Mon Nov 19, 2012 11:54 pm

Re: Feedback on possible KerberosSDR V2 Hardware

Post by rtlsdrblog » Fri Nov 08, 2019 4:16 am

Streets814 wrote:
Thu Nov 07, 2019 12:49 pm
A little background first. I have had my ham license for a little less than a year now. Other than radio communications, what attracted me to the hobby is Direction finding and data over radio (see AREDN project.) I have purchased another Direction finder (http://members.chello.nl/~w.hofman/pa8w/rdf.htm) and (https://radiodirectionfinding.wordpress.com/) if you are familiar with RDF mapper (http://www.musther.net/RDFMapper/) you most likely know about the direction finder linked above.

Having purchased the RDF42 (which cost between 2-3 times the price of the Kerberos SDR by the way) I came to the harsh realization it was a bit more complicated to assemble than I anticipated. Unfortunately, as a result, it sits in its original box with all the pieces unused. I have been following this project for a while now wanting to try it out, but hesitant to not have another unused device sit on my shelf. The idea of a plug and play ready to use version greatly appeals to someone like me. To be honest I want to pull the trigger on purchasing just to support the project.

With that said here are some things someone of my ability/knowledge (minimal) would find interesting or would like to see:

• Truly plug and play. Its easier (for somone like me) to tinker and learn than to have to build from scratch. while the Kerberos board is a lot easier than other systems I have seen, the easier to get started the better.

• If you add a small screen as mentioned something like the RDF42 screen (https://radiodirectionfinding.wordpress ... ot-screen/) makes a lot of sense. Right half has a direction bearing pointing in last X amount of directions averaged. The "ex" (bottom right) shows when an accepted signal has been sent to the computer. Center bottom is the symmetry meter above that (numbers) and to the left of that (bottom left corner) is the elevation meter which I am not sure how works to be honest. You can see the "quality" of the signal in the center of the screen just to the left of the direction pointer showing "Q8." I believe the signal is graded 0-8 and can be used by the user to set a rejection threshold. Meaning if the user sets the threshold at 5 any signal lower will be ignored and not sent to the computer for direction finding.

• Anything that will assist with using multiple devices together. As stated above I am also following the AREDN project a lot as well. The idea of a ham radio mesh network using wifi on the ham bands over long distances may have a lot of options with the kerberos. If I have multiple sites and install a kerberos on each site I could (in theory at least) simultaneously get a reading from the kerberos and see in real time the direction/source of the signal using mapping software like the rdfmapper.

•Ease of use. If there are easy to find simple to understand controls/features it makes life simpler for someone at my level. dials, knobs, sliders (real or virtual) that are self explanatory go a long way. I am not suggesting dumbing down the project but rather quick access to learners rather than digging deep into the settings like advanced users. Maybe a basic GUI for new users?

• Price. While you have no complaints from me on the affordability of this device and project as it is by far the lowest priced direction finder option that I came across (by a large amount.) The ease of affordability on the wallet will keep this project in as may hands possible.
Wow I didn't know the RDF42 was that expensive. It's made from low cost components, so I thought it would be much less. I assume you still have to provide the radio too right?

Next steps will definitely be focusing on ease of use. Since KerberosSDR is so new, and code so fresh the first stage was always going to be to get it working well, but requiring setup and tinkering. Now that it's been proven to work well we're looking to start on a more complete product with an out of the box experience for various applications.

We already have a compass display and this could go onto a built in screen. But as mentioned in my other reply, a simple compass display like this is really quite useless in many situations. A GPS+mapping program that determines an estimated TX location is really key as multipath can have you going around in circles if you just constantly followed the bearing only.

Streets814
Posts: 8
Joined: Tue Oct 15, 2019 12:44 am

Re: Feedback on possible KerberosSDR V2 Hardware

Post by Streets814 » Fri Nov 08, 2019 11:33 am

Yes you still have to provide your own radio. Looked up the paypal receipt, for a assembled pcb (no housing no antenna wirering to the antenna array that you still have to build) for the rdf42 is 392.91 EUR ($433.47 Americian.) after shipping. The RDF41 is available for $100 less. I targeted the rdf42 because it appeared to be a little more user friendly, and it actually may be but there is still a lot of building to make it work even with it "pre assembled." Unfortunately with not many people knowing what it is, I do not think it will be something I can easily throw on ebay?

Also just for your own reference I looked into http://www.dopsys.com/ for a rdf system. They quoted me for a 100 Mhz-1000 Mhz base price of $2,800.00. So you can see, based on experience look into this as a rookie, the Kerberos system is by and far ahead for affordability and as a result probably a great draw to radio enthusiast wanting to get into this aspect of the hobby.

An on screen display is not a deal breaker for me by any means. While i am sure it would be a nice feature it would not turn me away if it was not there. In a ideal situation I would have a few of these at a fixed site working together and maybe one in the vehicle for something similar to RDFmapper.

To help support the project and learn a little myself I have literally just pulled the trigger and purchased the Kerberos from the online store. =D

eb2
Posts: 1
Joined: Sun Nov 17, 2019 12:56 pm

Re: Feedback on possible KerberosSDR V2 Hardware

Post by eb2 » Sun Nov 17, 2019 1:34 pm

First thing is that a ready to use device would be very useful and popular for most people.

I would suggest a version that allows the an all in one, and a basic version, without the built in computer plus display. Maybe standardise support for Raspberry PIs. This would allow you to produce an all in one box that is easy to support. It would also be very customisable.

For the basic device it would be nice to support Gigabit, USB3 C and multiple computer boards(PI Clusters as an example) The switched ports for calibration would be great, or sell it as an option for new and old boards. I like the idea of daisy chaining two units together. Are you going to increase the support for more antennas in a passive array?

Basically I would love your effort to be spent on a better main board and includes an all in one off the shelf expandable option. The SDR and open system is what sets you apart.

Just my 2 cents worth.

User avatar
Airwind
Posts: 39
Joined: Wed May 23, 2018 8:27 pm

Re: Feedback on possible KerberosSDR V2 Hardware

Post by Airwind » Mon Nov 18, 2019 6:37 am

A few thoughts from me:
  • Personally I didn't quite like the enclosure of the V1 hardware, and I think it could have been done better. Perhaps use some extruded aluminium enclosure with side panels used for connectors?
  • I would put all the digital ports like USB on a single side of the enclosure while the RF connectors (SMA) should be located on the opposite side to minimize cross-cable interference.
  • Could you please add a dedicated DC power connector rather than having to power the board via USB?
  • You think the SDR chip DIP switch could have been done in software?

MountainLogic
Posts: 1
Joined: Wed Oct 30, 2019 9:12 pm

Re: Feedback on possible KerberosSDR V2 Hardware

Post by MountainLogic » Tue Nov 19, 2019 1:01 am

I would love to be able to explore higher into the LTE bands say at least 2200 MHz so a version with an e4000 front end would be very attractive to me. I would be interested in helping with such an effort.

-scott

PS: RF Switches Yes, please. In fact ability to switch multiple inputs would be helpful. Less interested in on board computing, etc.
Last edited by MountainLogic on Wed Nov 20, 2019 2:51 am, edited 1 time in total.

awdmesh
Posts: 19
Joined: Tue Oct 22, 2019 4:23 pm

Re: Feedback on possible KerberosSDR V2 Hardware

Post by awdmesh » Tue Nov 19, 2019 1:58 am

You guys familiar with the rashawk?

https://www.rtl-sdr.com/signal-directio ... k/rashawk/

In all honesty I recall getting in touch with those people and it never really worked, but essentially what I’m doing with the current Kerberos setup is using mapping software and plotting stationary pis/Kerberos setups with the gps location of where it’s setting and networking back the DOA information in order to plot lobs on a map, my thinking is if I have the array positioned correctly the lobs and bearings should match the map data in correct relation to headings and thus giving me intersections. Say all that to say,

- keep usb power as the use of battery packs make it easy to set them up in remote locations

- add more software to the pi or built into the Kerberos that helps average the lines of bearing when the Kerberos is stationary and dfing

- if you were to build in WiFi, maybe have the ability for the Kerberos to “mesh” forming a closed mesh network of dfing hardware without requiring a cellular network

rtlsdrblog
Site Admin
Posts: 2425
Joined: Mon Nov 19, 2012 11:54 pm

Re: Feedback on possible KerberosSDR V2 Hardware

Post by rtlsdrblog » Tue Nov 19, 2019 3:53 am

Airwind wrote:
Mon Nov 18, 2019 6:37 am
A few thoughts from me:
  • Personally I didn't quite like the enclosure of the V1 hardware, and I think it could have been done better. Perhaps use some extruded aluminium enclosure with side panels used for connectors?
  • I would put all the digital ports like USB on a single side of the enclosure while the RF connectors (SMA) should be located on the opposite side to minimize cross-cable interference.
  • Could you please add a dedicated DC power connector rather than having to power the board via USB?
  • You think the SDR chip DIP switch could have been done in software?
Good points.

For DC power, we could just provide a USB-C to barrel adapter. Would what work for you?

The idea of the DIP switches was to physically disconnect them if you didn't need some units, and to help with troubleshooting if you're having power supply issues. The future version probably won't have those switches, or maybe software switched if we have onboard computing with GPIO.

rtlsdrblog
Site Admin
Posts: 2425
Joined: Mon Nov 19, 2012 11:54 pm

Re: Feedback on possible KerberosSDR V2 Hardware

Post by rtlsdrblog » Tue Nov 19, 2019 3:57 am

awdmesh wrote:
Tue Nov 19, 2019 1:58 am
You guys familiar with the rashawk?

https://www.rtl-sdr.com/signal-directio ... k/rashawk/

In all honesty I recall getting in touch with those people and it never really worked, but essentially what I’m doing with the current Kerberos setup is using mapping software and plotting stationary pis/Kerberos setups with the gps location of where it’s setting and networking back the DOA information in order to plot lobs on a map, my thinking is if I have the array positioned correctly the lobs and bearings should match the map data in correct relation to headings and thus giving me intersections. Say all that to say,

- keep usb power as the use of battery packs make it easy to set them up in remote locations

- add more software to the pi or built into the Kerberos that helps average the lines of bearing when the Kerberos is stationary and dfing

- if you were to build in WiFi, maybe have the ability for the Kerberos to “mesh” forming a closed mesh network of dfing hardware without requiring a cellular network
Yep, but RasHAWK uses doppler, which is a bit different to how the Kerberos works.

If you're not using you're own mapping software, keep an eye out for a post very soon, either today or tomorrow as we'll release simple script that allows KerberosSDR bearings to be used on RDFMapper (http://www.musther.net/RDFMapper/). RDFMapper is third party, and he charges $25 for it, but it can be used for networked units.

User avatar
Airwind
Posts: 39
Joined: Wed May 23, 2018 8:27 pm

Re: Feedback on possible KerberosSDR V2 Hardware

Post by Airwind » Tue Nov 19, 2019 6:36 am

rtlsdrblog wrote:
Tue Nov 19, 2019 3:53 am
For DC power, we could just provide a USB-C to barrel adapter. Would what work for you?
I was trying to say that USB may not be the best connector for delivering power to the device. What I had in mind was to use a separate and proper standardized DC-socket instead. For example when I am driving around in a car I would often hardwire my devices directly to the car battery over a 12VDC to 5VDC buck converter via plain copper wires with the PC connected separately.

You could use either the barrel socket or a green DC terminal block. Both are pictured in the link:
https://www.reichelt.com/de/en/terminal ... 12975.html

The barrel connector is more widespread, but a lot of industrial gadgets use the terminal blocks. For example:
https://www.cnx-software.com/2019/11/12 ... or-module/

rtlsdrblog wrote:
Tue Nov 19, 2019 3:53 am
The idea of the DIP switches was to physically disconnect them if you didn't need some units, and to help with troubleshooting if you're having power supply issues. The future version probably won't have those switches, or maybe software switched if we have onboard computing with GPIO.
Yes, that makes sense. If you're planning to migrate the switches to software GPIO, then you could still have some unexposed physical switches present within the enclosure for just in case. Maybe something simple like pin jumpers?

Streets814
Posts: 8
Joined: Tue Oct 15, 2019 12:44 am

Re: Feedback on possible KerberosSDR V2 Hardware

Post by Streets814 » Tue Nov 19, 2019 11:06 am

rtlsdrblog wrote:
Tue Nov 19, 2019 3:57 am
awdmesh wrote:
Tue Nov 19, 2019 1:58 am
You guys familiar with the rashawk?

https://www.rtl-sdr.com/signal-directio ... k/rashawk/

In all honesty I recall getting in touch with those people and it never really worked, but essentially what I’m doing with the current Kerberos setup is using mapping software and plotting stationary pis/Kerberos setups with the gps location of where it’s setting and networking back the DOA information in order to plot lobs on a map, my thinking is if I have the array positioned correctly the lobs and bearings should match the map data in correct relation to headings and thus giving me intersections. Say all that to say,

- keep usb power as the use of battery packs make it easy to set them up in remote locations

- add more software to the pi or built into the Kerberos that helps average the lines of bearing when the Kerberos is stationary and dfing

- if you were to build in WiFi, maybe have the ability for the Kerberos to “mesh” forming a closed mesh network of dfing hardware without requiring a cellular network
Yep, but RasHAWK uses doppler, which is a bit different to how the Kerberos works.

If you're not using you're own mapping software, keep an eye out for a post very soon, either today or tomorrow as we'll release simple script that allows KerberosSDR bearings to be used on RDFMapper (http://www.musther.net/RDFMapper/). RDFMapper is third party, and he charges $25 for it, but it can be used for networked units.
OMG this is amazing news!

Post Reply