The AM Forum

Hi all,

Background- Having put together recently a cheap SDR HF receiver kit, I started noticing birdies and noise on several HF bands including one that was smack dab in the middle of the big 3885 window. After researching and sniffing around, I identified the problem as spurs from my wireless router (not in the shack) mounted high in the center of the house in the living room. I tried all kinds of RF beads, bypassing etc. with negative results. So, without even thinking I replaced the router with a newer one (N type) and got exactly the same results..

OK.. time to rethink (that's what we do right?).. So I took some time off from the problem and started operating on the air with my trusty Collins 32V-1 and R-388 combo just this afternoon. Have you ever noticed, (thanks Murphy) that bad stuff happens in little bursts.. A bad day with the Electro-Gods is a bad day; well almost..

As per my usual routine, I load up into a dummy load and set the Mod monitor, audio gear etc. and get generally ready to transmit on the set frequency. I will monitor myself off an RF detector to make sure all is proper BUT on this day I started hearing a crackling noise (way down but noticeable) that sounded like a cap or high voltage component on the way out. This is a completely reworked 32V with newer caps and mods.. NOT TODAY, not after the other problem, errrgh.. As most of you would do, I went through the audio chain first, monitoring it directly hoping for a simple connection problem, but NO crackling, clean. So, I switched transmitters and .. wait the SAME crackling!! Ok, was it a monitoring problem? Perhaps a leaky diode in the detector?? I had another detector and the SAME results. The plot thickens..

I have many TR relays in the shack, with a myriad of coax switches routing to various transmitters and receivers and started to troubleshoot .. With some luck while transmitting and monitoring, I was able to identify a 3 ft length of RG-213 that was the piece used as a common feed to the main coax switch..and with a little shaking and bending, the “crackling” went away..

Result- The point is that the birdies on the SDR also vanished! The rectification that the low-level arcing produced in that coax jumper was the common problem with both transmitting and arcing, and the lack of proper shield for the wireless router's birdie manifestation in the SDR.

So this is the question to the group.. How many of you are using “indoor” pieces of coax that could be as old as some of your vintage gear, and maybe it's time to simplify our switching systems to a less convenient direct patch panel with NEW coax..

73,
Jeff

Try 600 ohm 1:1 transformers on the I/Q audio to the sound card, you have a common problem. Radio shack sold dual transformers. I'm told they work well.

Frank,

Thanks for the input but the spurs are gone as in my post after identifying the coax jumper in the switching path at my station:

"Result- The point is that the birdies on the SDR also vanished! The rectification that the low-level arcing produced in that coax jumper was the common problem with both transmitting and arcing, and the lack of proper shield for the wireless router's birdie manifestation in the SDR."

Thanks for the info.

I have in the back of my mind a plan to some day make a UHF jack matrix to use as a patch panel and make up a bunch of short RG213 jumpers and do my mix and match of tx and rx and antennas with that.  I firstly need the tx and rx to mix and match.  I'd just mount 20 or 30 UHF jacks on an aluminum rack panel.  Maybe someone makes one drilled out already.

Rob

Jeff,

I'm glad you tracked down the problem and ended up solving two problems with one fix!

To answer your question, I have several "ancient" patch cables that are definitely past their prime. One of them caused some issues when I was running a software power calibration routine with Power SDR. On 20 meters the Flex began hauling heavy current and could only make 70 watts maximum. It was one of those head scratching events because it didn't make sense since I could easily make 100 watts out into the antenna. Being suspicious, I tracked the path back from the output connector of the rig through the metering couplers to the dummy load, and that's when I saw the culprit. It was an 8 foot length of gummy looking RG-8 with the infamous brand name of "Tandy" in faded blue letters on the outer jacket. That chunk of pseudo-coax had to be more than 40 years old. When I chopped it up, the foam was all yellowy, the 75% shield was nearly black, and when it was flexed back and forth it made crunchy sounds. Nasty!

I also have some dubious ancient coax connectors with bake-lite type center material on a few patch cables that really should be chucked out. I've been grabbing half a dozen silver plated Amphenol connectors at each flea market and slowly making up new cables with the end of a roll of Belden 8267 that a neighboring ham handed me for helping him move his station.

That's good advice to take a critical look at the inside cabling and re-think the routing setup. At least a few disasters might be averted by cleaning that stuff up!

Rob W1AEX

Jeff, your advice is spot-on.  I got a box of BNC jumper cables at a hamfest and had to try 3 or 4 for the jumper between the HP 3325A synth and the GPT-750 before I found one that didn't leak like hell.  Also had some issues with jumpers when using a network analyzer, the short/open/50 ohm cal procedure would produce strange results until I found one jumper that had been physically damaged.

Consider RG-316DS, RG-142, RG-400, RG-393 and Andrew Superflex for interconnection cables. In this age of 'digital trash everywhere' stopping ingress is a significant issue. Double or solid shielding makes a noticeable difference.   

Thanks for sharing.

Once I was working on an old AM TX and it had an internally shorted short RG-8 cable below the chassis.

The cable's dielectric had been completely melted and the cable became a UHF stub.

Someone must have tried loading up into thin air.

Be wary of shortness.

Jeff,
         I had a similar problem just a few days back. The S/N ratio of my SX-28 (which has always been a dependable workhorse) went to hell in a handbasket. Also had scratchy arcy noises when listening to myself transmit, but everything looked clean on the scope.

I didn't have these problems if I switched to another receiver. I scratched my head for a while and started looking around. Eventially, I found the problem to be the coax and RCA connector that carried the antenna over to the 28 was oxidized.

Every time that I burn a PL-259 off of the end of a piece of jumper coass, I just throw the whole thing away and make up a new jumper. (this has happened quite a few times)

With a burned up connector on the end of a coass, I have seen some really wierd quirks happen. I do mean some strange symptoms.

Jeff
et al.

I agree. 1x2 switches in series seemed (up until recently) like a convenient way to select among multiple tranmitters, receivers, etc. But, not only do I have a whole lot more places to look for a faulty jumper cable, or a switch or connector that isn't making contact... I find that it often takes longer for me to remember all of the switches that need to be reset (and to get them all set correctly) than it would to make up the new configuration with direct connections.

I'm putting the idea of a patch panel high on my list of radio projects.

I'm thinking about/wondering whether the patch panel should be designed with all of the connector grounds attached to a common metal mounting plate... or whether the connector grounds should be insulated from each other to minimize prospective ground loops. I'm also thinking about UHF connectors v. BNC.

A problem I had a few months back was a jumper with a PL-259 plug that wasn't making good contact to the jumper's shield. The missing shield current was picked up by other cables, but the broken shield problem created a lot of RF in the shack. It was a pain in the XXX to find the problem because (as you pointed out) I had to troubleshoot all of the cables and connections that were in series, one by one, by opening up the connections.

I also had a problem with a Dow-Key relay... some time ago... whose SO-239 connectors weren't making good contact with the center pins of the PL-259 plugs on my jumper cables. I solved that problem by rolling up narrow strips of aluminum foil up into  "miniature musket balls", and using a piece of #12 wire to "load" them into the SO-239's. Inserting the PL-259's compressed the "musket balls" to make a good contact.

Stu

A TDR (Time Domain Reflectometer) can be useful in looking at everything from short patch cables and connectors through long runs of coax.  A Tektronix 7S12 with its pulse and sampling heads came installed in a 7904A mainframe I bought several years ago and it is the most useful "unsought" piece of test gear I have acquired.  The most unusual use to date was finding a break in around 400 feet of AC feed cable to a friends deep well pump.  The AC cable looked very much like a balanced feeder and the the TDR can be calibrated for a variety of different velocity factors in addition to standard presets.  The break was found within 1 foot of the calculated distance which was very good and saved him a lot of exploratory digging.

My first run-in with weird feed problems occurred as a brand new general back in 1975.  I built a two element quad and found the SWR was very high.  After multiple checks of my soldered connections, coaxial matching sections, etc. I found the problem was the double female UHF barrel I was temporarily using between the feedline and the individual element feeders while I was waiting for a coax switch to arrive.  The barrel measured around 20 ohms between center and ground which didn't do a lot for performance.  I wasted several hours checking all of my work thinking I would find a mistake there instead of suspecting a new component.

Rob- K5UJ- Not sure if drilled out boxes are readily available, but I am already searching for a source of HIGH quality UHF chassis mount connectors, and a suitable aluminum enclosure to wall mount 2 patch bays around the shack... This project will be at the top of the list in the static months ahead..

Rob-W1AEX- Interesting on the “Tandy” piece that basically had suffered coaxial Rigor Mortis with decades of moisture working its way in.. We all have a stash of jumpers some having sat coiled up for too long with out proof of performance until strange things start happening!

John JN- Indeed those pesky BNC cables that I have a load of inherited thanks to CBS.. I remember having an average of one in five that were OK ..god knows how long those puppies were in storage..Hope all is well!!

JAY-W1VD- All of those cables have GREAT shielding and thanks for the list! I will indeed upgrade all interconnects when this project gets going.. BTW, if you truly analyze the series connected relays, switches, couplers, etc. it gets sloppy and doesn't help the RF trash from getting in!

Dave-W2VW- Shortness? (so many things I COULD say) :)  Interesting on that trusty short jumper probably looked fine on the outside? Longer into matched loads would be better..

Frank (The SLAB BACON) KB3AHE- I can see you now after that tedious restoration of the SX-28 (with some choice expletives) thinking oh no not the 28!! That's very interesting on the frying noise heard while monitoring but NOT in the transmit path. It is all the more reason for a true discreet and isolated patch panel that would instantly help at least identify those kinds of problems! Hope life is ok my friend!

STU-AB2EZ- One thing that's absolutely true is the secret code ring we all have to make the switches necessary to make all the equipment sing.. How many (ok be honest) of us have forgotten to throw that cherished correct combination and loaded into absolutely NOTHING.. ( short UHF stubs).. The series connected path of all the coax switches (I have counted 8 in the upstairs shack), the coaxial relays (at least 6), mismatched coax types, and it gets RF drunk and certainly susceptible to external RF sources and could become some kind of big diode as a whole..  I am going to make as mentioned earlier a true isolated (no common ground) panel. The artifacts (if any) and potential ground loops are gone, There are a bunch of ways to do that but I am inclined to use an metal enclosure but use heavy nylon washers as insulators from the chassis for the connectors. I also believe that UHF makes interfacing easier so that's the connector of choice for me.. One last note is that I have decided to consolidate ALL the antenna relay switching in this box as well. How many failing Dow keys must we burnish the contacts or add aluminum balls to?? A central HIGH quality RF relay could be part of this matrix..

Rodger-WQ9E- Thanks much for the info on the TDR! I looked that and the mainframe up on the web and that would be the way to go with quick troubleshooting and analysis but NOT cheap!! 
Also very interesting on the female barrel.. It's the little things!

Jeff
et al.

There is a UHF Female-Female bulkhead connector that might be the right choice for making a patch panel. Using it implies two (2) UHF -type connections back-to-back... but it would be easy to rearrange the cables that come into the back of the patch panel.

See, for example:

The Radio Works catalog, page 84, RF connector part #406 (UG-363).

http://www.radioworks.com/

Also: http://www.universal-radio.com/catalog/antsup/ug363.html

Stu

Stu,

Stuffing aluminum foil balls into the connector hole, probably not a good idea because of electrolytic corrosion, especially with a bit of moisture and some RF current flowing... quick fix, sure. Long term, I dunno.

                         _-_-bear

Interesting thread, since I'm planning the construction of a patch panel for my shack. The mess of switches required to be able to connect all the possible permutations is just more than I care to deal with.

MFJ has a couple of patch panels, but they use UHF barrel connectors, which is more PL-259s than I care to make up. So, I've ordered in a pile of Amphenol 83-1R chassis mount SO-239s from Mouser.

Like Stu, I wondered about the common ground situation. But, I've decided to run a common bus bar down the middle between the two rows and just solder the grounds up right. It'll be a great contribution to a good single point ground system in the shack. The four-hole mount S0-239 makes for a lot of holes to drill, but that's what a drill press is for.

I've got the 1/8" aluminum plate so the only thing left to do is just start drilling. And making up the patch cables.

Bird switches and type N connectors quality coax, done
You just can't avoid BNC cables I try to use RG223 or RG400 when I do anything with RF. RG58 is junk

I very much prefer BNCs in the shack, at least for receiving and low/med power apps.  Much better shielding, no chance of getting cross-threaded, and positive contact.

I agree with John on the use of BNC connectors where possible.  I also find them easier to install than PL-259 connectors.  I have a bunch of receivers set up on a multi-coupler and since BNC is used for the coupler I used that (with proper adapters) for all of the receiver connections. 

BNC connectors also stand up better to frequent connect/disconnect cycles; one of the reasons they are so commonly used for T&M equipment.

When I'm working at -130 dBM on 3.885 I can hear strong AM signals due to leakage around shields in a 6 foot hunk of RG58. yes BNC is a standard but they anr not the greatest. Yup much better than PL259 junk

Wow! I wanna live where you live. If your noise floor is -135 dBm on 3.885, you must have a very quiet location!

The noise floor between the signal generator and receiver.
Leaky coax I hear strong signals

TNC is a better choice than BNC for patch cables where repeatable leakage performance is concerned - it's a threaded connector vs. bayonet. Many flea market sellers offer them at a good price since they don't normally attract much attention. TNC connectors on RG400/RG142 is about as good as it gets short of high zoot purpose built cables.   

Nice thing about TNC connectors is they have the same guts as a BNC so very easy to convert.

OK. So on the air you would never hear this stuff. Why worry about it then?

just bakuz

It's good to know the limits of your measurement system. When I was working in the EMI field years ago, we always used double-shielded coax to minimize or hopefully eliminate low level signal corruption.

Even RG223 with BNC connectors is a problem for me at work.