July 20, 2007

Assorted mods & mtce


Today I spent about five hours working on the repeater tending to an eclectic assortment of items. Some related to the new controller installation, others that would fall under 'routine maintenance'.

I checked both the main and secondary receivers for both the 444.725 and 52.92 repeaters and found all four to be of normal sensitivity and without noticeable transmitter desense.

The Six meter Voter (an LDG unit) was checked and found to be favoring RX-3 (100' up the tower) over RX-1 (330' up the tower) by a fairly wide margin. This was corrected and the alignment of the RX-2 (the receiver located at the W8TVC location) was brought more into calibration. Hopefully the system will vote properly for the receiver with the best signal.

The DStar 145.36 repeater duplexer tuning was inspected and found to be 'on the money'. The RG213 coax jumpers from the duplexer to the DStar repeater were replaced with equivalent lengths of 1/2" superflex heliax and after measuring there were modest improvements with the better coax.

While on the topic of DStar, late in the evening on 7/19 it was noticed that the DStar repeater was off the air. On arrival at the site about 11:30am today I found the physical appearance of the system to be normal. The DStar repeater came back to life after power cycling. Power cycling of the DStar system is possible remotely via the 'FCC Legal' control system-- I should have thought to do that on Thursday night--- and save having the system offline for so long.

Hopefully, the apparent 'crash' of the DStar internal computer does not become a routine occurrence.

I opened up the new RC210 controller chassis and installed a hex buffer from two sections of a CD4049 in line with the COS signal from the six meter voter. This should improve operation of the switching circuits as the logic level was a bit marginal due to the high source impedance combined with a low load impedance. This brings into match the wiring of both the wI0OK/r and W8TVC/r controler chassis.

I attempted to upload the 'quack5.wav' file in order to reinstate the old DTMF test from the previous RC210 and the DVR (Digital Voice Recorder) in the controller but it did not play out. I tested the DVR with a local recording via RF and it does function. Perhaps the encoding of the WAV file required by the newer RC210 is different from the older versions.

Thankfully, the DTMF decoding of this new RC210 appears to still be functioning. I'm still crossing my fingers that the intermittant open trace on the board has been eradicated. I did notice that from my SyntorX certain DTMF tones from 6m were not decoded. All tones decoded properly from my Yaesu VX7, so it looks as if I may need to tweak the audio equalization at some point (most likely the culprit is the 'new' PL filter in the new RC210 chassis-- this filter is derived from an RCA700 radio, whereas the older chassis used a filter from a Motorola Motrac).

I also verified the reception sensitivity of the 440.05 link setup and found that the receiver had adequate sensitivity but that there was excessive loss (about 10db) in the duplexer (the 440.05 receiver is fed from a Cellwave Stationmaster at 100' and the RF is split with the control receiver and the 3rd 449.725 receiver via the duplexer.) I will replace this 'mobile' duplexer with a different unit later this year, hopefully providing a greater fade margin for the link from the W8TVC/r system.

Sadly, I neglected to verify that the transmitters and receivers were on frequency--- oops. while it's simple to verify the transmitters (I can do that from home), the receiver adjustments require a visit to the site. These adjustments will be tended to soon, on a 'seasonally cool' day. Today would have been a perfect day, but alas-- I didn't give frequency drift a second thought. Harumph. For the most part, the radios appear to be functioning properly, but there is a question as to whether the link from Empire to W8TVC/r is properly netted on frequency--- it sounds about 5khz off at times. It's been about two years since the frequencies were netted.

More news as it happens--- or within six days.. ahem....

John KF8KK

July 17, 2007

Repeater Controller Upgrade News


Yesterday and today the new Arcom RC210 controller for the GlenArbor / Empire repeaters was brought to the site and installed.

Sadly, it appears that the DTMF decoding problem which caused the RC210 board to be shipped back to the factory has developed the same problem again. After a bit of investigation I have concluded that this intermittant problem is due to either a cracked trace on the board or a bad IC socket. Pushing on the board in certain ways tends to change the failure mode.

At times the board would only detect DTMF '123' to call up the time, other times it would not decode any of the 'even' digits. Sometimes it would decode on 440, but not on six meters.

The last work on this problem was my replacing the IC sockets for the DTMF decoders. The RC210 has NOT failed since that work, but it is too early to tell as the problem was intermittant. I am encouraged, however. Any additional failure due to this problem will cause me to purchase yet another RC210 in hopes to finally get a good one (note that the RC210 put in service at W8TVC/r appears to be good--- this RC210 for Empire is meant to replace an RC210 built from a kit in 2003 that has some wierdities).

Otherwise, at this point MOST of the functionality that the repeaters had has been returned. The 'quack' DTMF tests are not operational yet (I need to find that quack file!).

Additionally, certain two meter remote base frequencies are not available, as is the VFO mode of the remote base. The 147.04 and 146.78 (the two frequencies most used) are functional. Apparently the frequency data to program the Motorola SyntorX is not satisfactory to the unit. Investigation of this problem will be taken up later in the year. The VFO mode (and the memory channels that rely on its operation) are almost totally unused and time is at a premium right now.

The 224.56 remote base, and the Weather Radio are also functional. Readback of A/C and battery power at the site also works.

This new controller is set to pass all DTMF tones. No longer is it necessary to hit '555' to unmute the dtmf. This is in preparation for the new Echolink DTMF decoder to be installed at the W8TVC/r site later this week. The plan is to eventually make it easy to command the Echolink node via 444.725 and 52.92.

While this new controller sports an interface to eventually connect to the 145.36 wI0OK-HD DStar digital voice repeater, that machine is not yet linked in. The linking of the digital repeater to the analog is slated for late fall. Sadly, there are interesting issues that must be addressed with circuitry in order for the link to function as one would expect. An example is that DStar users can send data through the repeater seperately from the audio. The interface must know to ignore the data-only transmissions of the repeater, otherwise it would seem as if there are many 'dead carrier' keyups from the DStar realm making for an annoyance in the analog world.

Overall operation of the new controller should be identical to it's predecessor. I will be publishing a newly revised 'Cheat Sheet' shortly.

John KF8KK

July 13, 2007

W8TVC/r controller upgrade update


Last evening more time was spent at the W8TVC/r repeater in order to finalize the controller installation. While the job is still unfinished, good progress was made.

The Echolink node interface (the computer side) was replaced, and additionally some circuit modifications were made to the controller side that communicates with the node. Apparently the 'phantom' COS signals were related to the PL Encode signals and the logic had to be rerouted somewhat for proper operation.

In addition, some capacitors were changed in the audio of the Echolink PC feed to the RF gear to provide additional high-frequency response (the Echolink audio sounded a tad 'muddy' previously).

An audio rolloff circuit was added to the outpu of the 144.67 receiver to bring the response of that into a more acceptable range.

Finally, an older multiport serial card was added to the Echolink pc in hopes of adding more serial ports (needed for the new Echolink DTMF decoder and the RC210 remote control software). Sadly, this older 'legacy' ISA board was not immediately recognized by the computer and the 'cheat sheet' which had the information on how the boards jumpers were set (and needed to manually tell the computer where the ports were) was elusive at the time. Being near 11pm at the time, Dave and I called it a day.

73, KF8KK

July 01, 2007

Empire Controller Upgrade Project Commences


Yesterday, June 30, the first outage associated with the upgrade/replacement of the RC210 controller took effect.

Part of this project is the replacement of the old 'kk standard' three pin molex/nylon power connectors with the standard 30amp Anderson PowerPole connectors. During both Saturday 6/30 and Sunday 7/1 the power connectors were replaced.

On 6/30 the 'FCC Legal Controller' chassis was temporarily removed in order for some minor modifications to be implemented. These mods deal with the new DStar system and the current packet configuration on that site. Additionally, certain wiring was removed as it was no longer needed due to the added voting system that was put in service after this module was created. The revised controller chassis was placed in service on 7/1 and appears to function well. It now awaits the installation of the new, revised RC210 chassis.

Sadly, on 6/28, the new Arcom RC210 controller was shipped back to the factory due to a very strange problem (infant mortality) wherin it would not accept commands from port-1.

With the new 'FCC Legal Controller' mods in place, the new Anderson PowerPoles in place, and the new chassis for the RC210 fully wired, the controller update project is nearly complete. The DStar interlink is a seperate project which did require the added functionality included with this controller update in order to be interfaced. I expect the DStar project to be implemented in the early fall (there's a host of 'interesting gotchas' that get involved in recovering an accurate COS signal from the received DStar signal)(ahem-- and the IC2200h linking radio is too nice next to my easy chair!).

In other, somewhat related news, the new Echolink interface board (the part that mounts in the pc) for the W8TVC/r system has been laid out and needs to be etched. This board will be a multi-purpose board that can also be used to provide switching and delays that are often needed in interfacing the links (see the fix for the echoey feedback in a prior post).

It was also discovered that it appears the 'phantom cos' that was being fed to the Echolink PC was most likely related to the 'courtesy tone with ctcss' mode that the W8TVC port-1 output is configured. The CTCSS driver also feeds the Echolink PC as COS, and hence, it was adding an undesireable signal to the system. I have worked out a minor circuit change to fix this (the fix will also help with the DStar integration in Empire too). The new echolink interface board will also help as it has built in adjustments and configurability to deal with such things as needed. I expect the new boards to be ready to install early in the week and installation to take place later this week (pending Daves schedule, etc).

The updating of the repeater and linking gear at the W8TVC site to Anderson PowerPoles will be done gradually over the next three months. There is much work to be done with the W8TVC repeater (Echolink pc updates, RC210 programming, antenna replacement, etc) and the power connectors will be changed as time permits.

Film at eleven...

John KF8KK