I was asked to make up some quick-detach coupling tubes to allow two 24GHz round flanges to be clamped together. I made a tube with a 22mm x 0.75mm thread outside and broached a keyway slot, then made a brass key to fit and fixed it with Loctite. So far I’ve only made up one of the clamp rings, I’m waiting for a real example to turn up so I can get the tolerances exactly right. All its fine so far though.
Following the design in http://www.g4dbn.uk/?p=1333 I made up a trial version in brass. Anti-cocking flange, adjustable barrel and the standard sized coupler for a VK3CV design Silicon Radar 122GHz board.
This version has round 2.00mm waveguide.
Willi HB9PZK modelled a Rexolite lens for a “POTY” dual-band patch antenna to use on the 10GHz downlink from the QO-100 amateur satellite. https://rfantennas.wordpress.com/author/hb9pzk/
The design is a joy to make when compared with the dual-taper versions. I have some one-inch Rexolite 1422 round bar, so I’ve been making a few of these lenses to replace the lossy Nylon versions that some folks are using. Rexolite is a cross-linked polystyrene, which machines rather like plexiglas. It has a tan-delta around 0.0004 and almost zero water absorption.
Willi’s design has a stepped transition to the read. This latest batch is to fit UK 22mm plumbing pipe with a nominal wall thickness of 0.9mm and an inside bore of 20.2mm instead of the standard 20.0 bore.
Nice simple job squaring, facing, drilling, tapping and slotting some free-machining C111 copper for an oversized heatspreader for a W6PQL MRF13750 600W PA board. 4 x 5 x 0.5 inch thick. I have quite a number of these to machine up now. C111 is such a contrast from horribly gummy C101
Part-completed job making some giant heatspreaders for a pair of F5JWF 4 x MRFE6S9160 500W 23cm PA boards. Still a lot of detrail work to complete on the resistor pockets and precision-N socket fittings
Here is a basic setup for Meinberg NTP which is suitable for Windows 7 onwards. The download link for the latest version is at https://www.meinbergglobal.com/english/sw/ntp.htm#ntp_stable
When you install it, you will be asked about creating a specific user to run the service. Up to a point this is a good idea, as it reduces the risk of security compromise, but you must ensure that the password you create for it will never expire. If you are not concerned about the small risk of compromise, you can use the SYSTEM account. Don’t user your own account though.
It will ask about which servers to use. A really simple solution is just to pick the UK pool from the drop-down list. You need to select four or more servers to ensure that the system will ignore any insane outliers.
When you complete the config, it will ask if you want to view the configuration file. Do that and it should contain things similar to this:
restrict default noquery nopeer nomodify notrap
restrict -6 default noquery nopeer nomodify notrap
restrict -6 ::1
driftfile "C:\Program Files (x86)\NTP\etc\ntp.drift"
server 0.uk.pool.ntp.org iburst minpoll 6 maxpoll 7
server 1.uk.pool.ntp.org iburst minpoll 6 maxpoll 7
server 2.uk.pool.ntp.org iburst minpoll 6 maxpoll 7
server 0.nl.pool.ntp.org iburst minpoll 6 maxpoll 7
server 1.nl.pool.ntp.org iburst minpoll 6 maxpoll 7
# these are default settings, it does no harm to put them in the config file anyway
The NTP recommendations say this:
Forcing a poll interval that is more frequent than what NTP would normally select on its own, hurts accuracy and stability of time on the local system.
NTP polling does not directly synchronize the local system clock to the server clock; rather, a complex algorithm calculates an adjustment value for each tick of the local system clock
Shorter polling intervals cause NTP to make large but less accurate calculations that never stabilize, causing the local system clock to wander. They are also useful if you want to make sure that your NTP daemon will detect an outage of the NTP peers in less time.
Longer polling intervals allow NTP to calculate smaller tick adjustments that stabilizes to a more accurate value, reducing wander in the local system clock.
If you do make any changes to the config file, you need to restart the ntpd service, either using the services.msc GUI, or go via control panel/admin tools to services, or if you are command-line oriented, just type:
sc stop ntpd
sc start ntpd
You can monitor status using
You can also run w32tm /stripchart to check against another NTP server:
w32tm /stripchart /computer:fr.pool.ntp.org
Press control-c to stop the stripchart.
There is a neat GUI program on the Meinberg site at https://www.meinbergglobal.com/english/sw/ntp-server-monitor.htm
Note that it is fairly old and says it is for ancient Windows versions, but it works fine on W10
CAD drawing for a version of the 122GHz coupler terminated in a UG-387/U anti-cocking flange. Brass with 0.061″ steel location pins. The finished procuct, made in brass is at http://www.g4dbn.uk/?p=1367
I presented this talk at the 2019 RSGB convention on 13th October 2019 at Milton Keynes. It is more of a polemic and call-to-arms to work on ways to push the limits of VHF/UHF/SHF weak-signal communications in the way the LF enthusiasts have, working with Eb/N0 very VERY close the the theoretical limit, but we are up against stochastic processes which spread troposcatter and other DX propagation modes on the sub-metre wavelengths and make the LF approaches less than useful.
I am calling for the implementation of an adaptive feedback channel using perhaps 146 or 71MHz datacomms, perhaps New Packet Radio or something dedicated, to give us a way to close the loop and modify transmit parameters on the fly. This is not going to be easy, but it could be a nice challenge. It also might fit the Ofcom criterion of “Not just more of the same” on the new allocations.
This is a Microsoft Powerpoint pptx file. Please don’t copy it to other websites, just create a link pointing the this page.
Time to turn the CAD design into something physical. Brass rod and my ancient Colchester 1800 and Bridgeport have come up with this as a first try. The barrel adjuster permits positioning the back of the waveguide flush with the inner end of the 8mm cavity, and up to about 6mm inside the reamed 4mm tube. Normal setting will probably be around 1.6mm
Based on the excellent work being done in Australia with the boards using the SiliconRadar 122GHz chips, I decided to have a go at making a version of their diplexer with a thread and locknut instead of a sliding piston and grubscrew. Also I am using brass. No particular reason other than the fun of making things. I did some initial sketches in Fusion360, and made a few of the locknuts last night, now making some horns to see how they look in real life.
See here http://www.g4dbn.uk/?p=1312 for the first machining attempt.