RF Patch Panel

Having worked in Broadcast engineering for 40 or so years, I have become addicted to the use of patch panels for signal routing in complex instillations.  Patch panels allow for an almost total rewire of a work space, without even needing to move the equipment, or pull out the supporting hardware for that equipment.  In short, you bring out everything from the device to a panel.  This allows you to patch inputs to outputs, outputs to inputs, etc., all by just adding a jumper cable.

I got tired of pulling out my station desk everytime I wanted to change something, or add a SWR meter, etc.  so I decided to build a patch panel for the station.

Overview

I have made a decision to not do anything halfway anymore, so when I do something, I will do it correctly, or not at all…  Far too many ham stations look like something exploded in the rear of the operations desk area, as did mine, before the patch panel.  This sort of installation promotes problems when rewiring, and I wanted no more issues along that line, whenever I added, or removed something, hence, the patch panel idea.

The Build

I am a believer in laying things out prior to doing anything…  So I built a to scale work image in my software, scaled all the items to size, and started moving things around on the desktop to see what I might come up with…  I got the image on the left, as a preliminary starting point.  Nothing ever comes to fruition as expected, so I was not married to this layout, I just wanted to be sure I was close to what I would get.  I first needed to look at the equipment and make a list of all the ports available to me, RX IN, RF OUT, etc.  I came up with a much larger list than I expected, so I am glad I took the time to make one.  I next set about designing a patch panel that had all the items on it, in some sort of order.  I decided to order things by signal flow, why signal flow, because it lets me have those cool lines between patch ports, showing flow right on the patch panel.

Layout

With those two decisions, I was able to lay out a patch panel in one of those online development thingys the places that will cut the panels provide.  I looked around at who provided on line design, followed by drilling, and lettering…  I decided to use Front Panel Express, as they had a fairly easy design tool called, oddly enough, Front panel designer!  There was a bit of a learning curve, and it took about a week for me to get comfortable with the designer tool.  The big plus for me, was that it generated a file I could just email to Front Panel Express, and they could drill directly from that file…  The downside of this is that if you get it wrong, you still have to pay for the panel…  Given the cost of custom CNC machined panels, one gets very careful in everything.  So the next step was to pick the BNC connectors I would use.  I do not suffer from Amature Radio Cheapness Disease, (ARCD), but by that same token I do not need the most expensive item on earth for my projects either…  I took some time, and looked around  for a good quality double female chassis mount connector, that was decent, yet affordable.  I had been using American Radio Supply for my connectors for years, so I looked to them for connectors.  I ended up with the design you see at the left for the panel.  Many little items to be careful of…  For instance, if you look at the type of BNC connector you will see it is a D type, meaning that the connector is NOT round, but in the shape of a D.  I wanted all of my connector ears to be lined up in exactly the same orientation.  Why?  Because I am OCD about that sort of thing.  I could have saved a few bucks and purchased round connectors, but decided to get D type.  That of course then forced a change in teh panel to a D hole for each connector.  Then there is the labeling…  How far below the connector does the label go to not be covered by the connector flat mounting plate?  How far below that to read the label, how big to make the font, what color, etc.  Suffice to say, there was a LOT of little details involved in planning a panel that will be created via CNC.  The penalty for not getting it correct the first time, was having to throw out the panel, and redo it, at around 200 bucks per try…  This was a good motivator on my part for getting it right the first time.  The result was the panel shown on the left.  In retrospect I might have chosen slightly different names for a few items like “IF Monitor”.  That probably should have been “IF Scope’.  But, for 200 bucks to correct that, I can tolerate this sort of thing…

Next was to create the mass of wires needed to connect 36 connectors.  I opted for using a good quality RG-8X in the panel for ease of construction.  I decided to use DX Engineering cable, as I have had very good luck with their products, and have used their cable for year.  I bought a 500 foot roll of cable, as I expected to also be constructing a Coax entry panel, and lightning protector  as well.

Mounting

I next needed a desk to mount all of this in, so I set about laying out the desk shown at the top of this article.  That took a while to do, as I am not good with things made of wood.  In fact, I have stayed away from wood tools, as the can cut body parts off, where a soldering iron can’t!  So I found a fellow that liked wood build wooden things, and consulted him.  We came up what you see on the left.  That took some time to complete, as he built it in his spare time.  But that also saved a lot of money…  Next was teh cable layout.  I tend to dislike crimp on connectors, so I elected to solder each one.  This was sixty or seventy BNC connectors, and few few dozen PL-259 connectors…  Soldering coax connectors does not scare me as it does so many hams, I think because at an early age I decided to master this…  However nothing prepared me for soldering BNC connectors…  I bit the bullet and decided to use crimp on shield, and soldered center connectors…  This was not difficult as I used roson on every connector.    I like using roson, it makes soldering anything simple, and clean.  If you have not tried it, do!  You won’t be sorry!  Anyway, back to the mass of cables, that of course then triggered how would I label them.  I have tried all sorts of methods but a good friend, WA7TUV, (RIP), once many years ago, showed me how he labeled cables.  He used printed labels, and heat shrink tubing.  I decided to do the same.

Labeling

That resulted in some really good looking labels, that have passed the test of time, and heat…  See the image to the left.  So I look my list of rear connectors, and got my coax cutters out, and started cutting and laying in cables for the K3 to the patch panel.  I have never been a fan of cables so tight one can not pull the equipment out, so I added an S curve to all the cables both at the device, and in the patch panel cavity behind the panel.  I used hundreds of cable ties to keep each cable bundle in place, and together, and folded into the S curve shown later…  I soon realized that I needed a different method of labeling cables, as I did not know where the connectors would end up.  Each run of cable was slightly different in length.  So I decided to remove my nice labeling, and replaced them with temporary labels made from price tags placed on each cable.  That way, I could lay out the cables, adjust them as needed, hold them in place with wire ties, and then cut them to the desired lengths.  After that I would relabel and then terminate each cable.  That resulted in the image you see at the left.  The next task was to relabel and terminate each cable.  This took a LOT of time.  Something like a week.

My PL-259 success rate was around 97 or 98 percent, so I only had to redo one PL-259 connector.  I wish my soldered BNC rate were that good.  I had ordered maybe 10% more connector wise than I thought I would need.  So off to the order desk again for American Radio Supply…  Another 30 or so connectors ordered, as I wanted both spares, and expected to blow up a few more connectors…  I had to be very careful in this, as each time I blew up a connector, the cable got a bit shorter, and messed up how the panel would look when finished.  After a few practice attempts I got good at BNC connectors.  I ended up using toothpicks to insert rosen into the little hole in the center connector of the BNC.  That turned out to be the trick for making solder go where I wanted it to go…  I unfortunately did not take any photos of the solder jobs for the BNC, but I did for the PL-259s…  I also had to relabel my cables again before putting on the BNC or PL-259 connectors.  Once that was done, it was time to wire the mess of 30 some odd cables up to the patch panel, and see how things fit and if my idea of an S curved cable set would actually allow me to work on the patch panel, by pulling it out and extending the S curve of cables.   IU set about testing each cable.  I ohmed out every single cable both center to center and shield to shield.  That took a day.  I found one or two connections I redid as a result of this testing.  That done, I set about mounting each connectors into each hole on the patch panel.  That took a lot of time, but at last it was done, and to my surprise they all fit perfectly, and with enough space to allow reading the writing on the panel as well.  The D on each was correctly placed, so each of teh ears on teach BNC connector lined up as they should.  I next connected each cable to the correct connector on the patch panel.  I pulled back the cabling, one cable at a time, and threaded each cable so as to avoid tangles, and in general adjusted things until I had them as I liked them for spacing and layout within the cavity in the desk.  I pulled back entire cables mass adding in the S curve, and then pressed the cable mass into the desk cavity, building in that S curve I talked about a bit ago, I next wire tied the entire bundle in place, and let it sit overnight to give the cable bundle a memory of how I wanted it.  I then added wire ties every 6 or so inches to hold the cables relative to each other in place.  Once that was finished, I dressed each cable out the hole in the back, and into the Panduit.

Panduit

Oh my God– if you have never used Panduit please go out right now, buy some,  Next find something to dress up and use it.  The stuff is God’s gift to people that lay out cables…  I had intended on using only two and maybe three runs of panduit, but no…  After seeing how well it worked for the coax, I bought another round of panduit, this time, smaller for AC and DC power…  So, the result of this was to put panduit everywhere on the back of teh desk, I used it for running every cable now in all projects.

Cable heating at high power

A few folks have mentioned they think using BNC, and RG8X cable is not the right path…  I worried about this prior to building the patch panel and did some research and a few tests to make sure 8X and BNC would take 500 watts without issues.  Here is a FLIR photo of the patch panel in operation.  This is after about 10 minutes of FT8 operation into a dummy load at 500 watts.  As you can see neither the cable, nor the connectors are getting warm.  If you look to the left side of the photo, you will see two vertically stacked Powerwerx power supplies, the top one is running the radio, and the bottom one is running a tuner and a watt meter.  You can see that the cable, and connectors are no warmer than the power supplies, and in fact are far cooler than the face of the supplies.

Final

I bolted the patch panel in place, and have not needed to remove it from that day forward.  The desk has been in place for around a year now, and I have not needed to mess with anything.  I have added part to the station layout, but to do so, I just connected them to the existing jumpers, add a patch to the patch panel, and was done…  No more do I have to pull out the desk to change something, add an antenna, replace an item, etc.  The final look and location for the patch panel is on my operating desk, as shown below.  Click the image to the left, to enlarge it if you are interested.