I sometimes get the distinct impression that the corner office doesn’t understand what it takes to keep a radio station on the air and in good repair. It is most often the problems or “issues” that tend to get the most attention. The things that are working well tend to get ignored. After all, how often do you hear a news report about the airliner that landed safely.
Lightning strike, TV tower
When lightning strikes the tower and knocks the transmitter off the air causing major damage and expensive repairs, that is a problem. When lightning strikes the tower and nothing happens, no problem. What is the difference between those two situations?
Grounding strap, FM transmitter site
If the generator starts and runs during every power outage and has done so for the last five years straight, it is obviously a reliable unit, does it need all that maintenance?
Caterpillar 75 KW diesel GENSET
Money spent on preventing undesirable outcomes can be difficult to quantify as disasters and events that do not happen are ill defined. It is difficult to quantify the “amount saved” on something that didn’t or won’t occur. Using past situations is good start, but that only covers a fraction of possible outcomes. In order to invest money wisely, one has to look at the probabilities. If there is an unlimited budget, then the probability exercise should be minimal, however, there is very seldom an unlimited budget.
For example, how much does a back up STL system cost vs the risk of being off the air while the main STL system is being repaired? How often do failures occur, when are they likely to occur and for how long are all good questions. Is there an alternative to a full back up like an IP CODEC? Such a solution would cover all aspects of the STL system including antennas, transmission line, transmitters and receivers.
There are certain FM stations north of here that have neither RADOMES or antenna heaters. Once every two years or so, the antenna ices up and the transmitter folds back due to VSWR. How much of an impact to listeners notice when this happens? If it happened more often, say two to three times a year, would it be wise to invest in some type of deicing equipment?
What is the ownership and management opinion on off air conditions? I have often heard tell “Oh, its only the AM, we don’t mind if it goes off the air.” That is, until it actually goes off the air, then it is a big problem.
Based on my and others experiences, these are the things that will happen at an average transmitter site:
The electric will go off at least once per year for several hours.
The main transmitter will fail at least once every two years.
Lightning will strike the tower at least once per year.
The STL system will fail, at unknown intervals.
At studio sites, these things will occur:
The file server will crash depending on the operating system
The telephone lines and or T-1 service, internet service, ISDN etc will go out
The electric power will go out for several hours
The satellite dish will fail once every two to three years
If there is a tower, it will get struck by lightning
Other site specific things can occur like floods, blizzards, earthquakes, fire, etc.
Money spent on backup systems for those items is good insurance. Not only will the station stay on the air, the on call engineer’s phone will ring less often, which, if you are the on call engineer, should make you happy.
If a full backup is not available, a second transmitter for example, having a good stock of spare parts on hand can mean the difference between an early evening and an all nighter. Keeping good maintenance logs and well documented repair records can point out trends and give a good basis for ordering spare parts.
Repair trends are important. If the same part seems to be going bad over and over, it is time to dig deeper and find the cause of failure.
The old adage “An once of prevention is worth a pound of cure,” still holds true.
Right after Tropical Storm Irene, it was noted that the STL signal strength at the WHUD transmitter site was low. Normally it was 300+ µV, now reading around 100 µV, which is a problem. Upon further investigation, it was revealed that the STL transmitter on the intermediate hop had higher than normal reflected power.
Time to call the tower crew.
The STL transmit antenna for WHUD’s STL (WPOU464) hop is a Scala Paraflector (PR-950), mounted at the 280 foot level on this tower:
Scala PR-950 on a guyed tower
The fact that it happened after a major storm and the transmitter was showing higher than normal reflected power indicates a problem with either the antenna or the jumper between the 7/8″ Cablewave coax and the N connector on the antenna. A measurement with a spectrum analyzer shows very high return loss:
WHUD STL antenna return loss
This shows distance to fault 413 feet, with a return loss of -7.4 dB. That distance is either near or at the antenna and -7.4 dB indicates a lot of reflected power. We had the tower climber take apart the jumper connections and terminate the jumper with a known good 50 ohm load. The return loss did not change. We then had him swap out jumpers and reconnect to the antenna. That did the trick:
WHUD STL antenna with new jumper
Much better, most of the power is now being radiated by the antenna, the VSWR is 1.02:1. The impedance bump at 51 feet is a sharp bend in the coax where it is attached to an ice bridge. Reconnecting the transmission line to the transmitter and turning it on confirms that all is normal again. The problem with the jumper was found in one of the connectors, it was full of water.
Water contaminated Andrew flexwell connector
I cut away the boot, water had entered the connector from the back because waterproofing and tape was not applied all the way to the coax. This was installed in 1998 when the station moved from Peekskill to their current location in the town of Fishkill. The fact that it happened now in the nice weather when Mt. Beacon is still accessible and not in the middle of winter means the radio gods are smiling on us.
Not that I am a glutton for punishment or anything, but I enjoy troubleshooting. There is a certain satisfaction in the analytical aspect of tracking down a problem and fixing it, hopefully in a permanent fashion. Figuring out where a problem is requires a good bit of detective work;
Examining the clues; what happened before the failure, what are the fault indications, are there any external factors
Round up the usual suspects; a good maintenance log is helpful here to track re-occurring failures. If the failure cannot be attributed to an external source (such as power surge or lightning storm), what was the last thing that was changed or worked on.
Following the trail back to the origin; Often the first failed part found is a symptom, not the actual problem. It takes some skill in reading schematics and making sense of a failure to trace it back to the real problem.
It can sometimes be exciting, like turning on the 25 KV high voltage supply and have big blue flashes issue forth from the top of the transmitter. Sometimes it can be quite frustrating, like when the station owners refuse to spend money to fix a problem. Sometimes it can be dull, like fixing the same problem over and over again because of the previously stated money problem. It’s also disheartening when the problem was caused by the stupid DJ spilling soda in the console. Not that all DJ’s are stupid, just the ones that spill things into consoles.
The challenge of finding the root cause can often be enlightening. I have often discovered unrelated problems waiting in the wings while investigating the why of an outage. It is great to fix those things before they burn the house down, but this approach often goes unnoticed by the ownership or management. Lately, for some reason, an once of prevention goes un-noticed or unappreciated.
There is quite a bit of science to trouble shooting, but there is some combination of personal traits that make a good trouble shooter. These are:
Inquiring or curious disposition. It is fairly easy to get to the first failed module or part. Discovering the reasons behind the failure and or getting down to the component level takes a good deal more effort.
Patience. This goes with the second part above, it takes some stick-to-it-tive-ness to trace out the not readily apparent problem.
Good analytical skills. Often failures generate a cause and effect scenario. The effects can be startlingly distractive and mask the causes and the under lying problem.
Ability to view the large picture. This is critical to discover outside influences and other issues that are indirectly connected to the system or issue at hand.
Ability to analyse the system design. This requires the background training and experience to look at a circuit diagram and discover non-error tolerant systems. Sometimes these systems can be modified for better fault tolerance.
Poorly designed equipment is the bane of the broadcast engineer. Equipment manufactures can sometimes fail to follow two key principles: KISS and maintainability. KISS stands for Keep It Simple, Stupid. There is no better design criteria than the KISS principle. Adding layers of complexity increases the failure expectations. Maintenance can be something as simple as cleaning or changing air filters. Making maintenance tasks difficult almost ensures that they will not be done.
Bathtub design curve
Eventually, all things wear out. It also takes some large picture skills to know when it is time to replace equipment and that can vary greatly from situation to situation.
Mouser Electronics has created a mobile web edition to their online store. This is a handy tool for searching, cross referencing and ordering parts. Mouser has a large stock and they ship quickly. Time once was that you could run down to the local electronics store and get just about anything you needed. Even Radio Shack carried a fair amount of small parts, tools, connectors and so on. Since then, the local electronics shop has closed and Radio Shack inventory gets smaller every year. Using a large parts supply company like Mouser or Allied is necessary if any type of trouble shooting and repair is undertaken.
I like the Mouser Mobile site because there is no app to download and install. One simply points the web browser on any mobile device to mouser.com and it will automatically redirect to the mobile site. If that does not work, then m.mouser.com will. The mobile website is easy to browse around, and if needed, a quick call can be made by hitting the little phone icon. Here is more in a video:
A quick glance at almost any circuit board these days will show that almost all of the components are surface mount. They are small rectangles or squares that sit on top of the circuit board. This is different from the through hole components that were used for many years and are still found in older equipment. There are radio engineers that feel that surface mount components are too hard to work on, thus the boards are not repairable.
California Amplifiers C band block down converter
As with anything in the engineering field, there needs to be a cost/benefit analysis. Most computer components boards, things like NICs, modems, sound cards, VGA cards are very inexpensive and often times it would be more expensive to repair the board than it would to buy a new one. In other situations, however, local repair of circuit boards makes good sense and can be a good learning tool.
Consoles and transmitters offer some good opportunities for local repair, provided the schematics are available. SMT component trouble shooting is the same as through hole trouble shoots, except the components are smaller. That is where a magnifying glass comes in handy. I purchased a magnifying glass/light to work on SMT boards.
three diopter magnifying lamp
Soldering and unsoldering techniques are also different. A temperature controlled soldering iron with a small tip is important. I find the easiest way to unsolder a component is with solder wick. Once most of the solder has been wicked up, a brief touch of the iron and the component will come off. Small resistors and capacitors are fairly rugged, but should not be overheated. Semiconductor components such as diodes, transistors and ICs are susceptible to heat damage and Electro Static Discharge (EDS). A grounding wrist strap should always be used when handling semiconductor components. Soldering iron temperature should be enough to quickly melt the solder and heat the connection surface without overheating the SMT component. Lead free solder requires slightly higher temperatures than the traditional 60/40 rosin core.
Weller WD1 temperature controlled soldering station
A temperature controlled soldering station is a must. Too much heat will damage components and boards, too little will make soldering SMT an arduous task.
Soldering supplies
Other soldering supplies include liquid flux, desoldering wick, flux remover and 55/40/5 solder. The desoldering wick makes it easy to clean up an errant solder deposits and is the best way to desolder surface mount components. I have had limited success using a solder pump on surface mount boards. They do come in handy for RF MOSFETS, which have large tabs, often with liberal amounts of solder applied at the factory.
Soldering new components:
HF receiver preamp SMT board
A typical 0.1 uf capacitor is placed on the surface mount board and ready to be soldered. These components are all small, but I would characterize this as a medium sized one. There are some very small diodes, ICs and other devices that require the magnifying glass to identify pins and polarity.
The best way that I have found to solder components onto a surface mount board is to use a little bit of liquid flux on the board.
Using tweezers or small needle nose pliers, place the component.
Soldering 0.1 uf bypass cap on SMT board
Wet the end of the soldering iron with a little bit of solder.
Using the placing tool, hold the component in place and touch one of the pads with the soldering iron. This should tack the component in place. Solder the component to the other pad using the soldering iron and solder. Then come back and touch up the tacked side. I have found that 600 degrees F is a good temperature to quickly melt the solder, while not heating up the component too much.
HF receiver preamp with bypass capacitor soldered
Sorry I could not get pictures of the actually process, I don’t have enough hands to hold the soldering iron, hold the component down and take a picture.
HF preamp based on Norton 1N5109 design
The completed preamplifier. I have been calling this an HF preamp, because that is its intended use. In practice, this preamp should work well from 50 KHz up to about 75 MHz, with 3dB points at 30 KHz and 100 MHz.
Norton HF preamp Schematic
The Norton design is an inverse feed back and using the 1N5109 transistor, which has input and output impedance of 50 ohms, makes it simple to implement. In testing, I found this unit has about 10-11 dB of gain with about 4 dB of noise. The use of SMT makes the design stable and I didn’t see any evidence of oscillations when testing it. More on the preamp here. I installed it out at the base of my K9AY antenna and it can be remotely turned on and off as needed. My main reason for wanting it is to overcome the 6.5 dB signal loss in the four port hybrid receiver coupler and transmission line I use. Truth be told, most of the time it is off.
Ever need to cross reference something? Or find a part fast? Get it ordered and on the way? No longer satisfied with the meager Radio Shack parts selection? Don’t have time to run to Radio Shack?
DigiKey has a mobile app for you. Its not that I like Digikey better than any of the other major parts dealers like Mouser, Newark, Allied, etc. They seem to be the first ones with a searchable mobile app and it seems to work well and it saves time. One of the things that I do not have very much of these days is time so anything that can move a project or repair along is welcome.
I also don’t have anything against Radio Shack, either, they have simply moved away from the parts business.
It’s the middle of the night and the phone is ringing. That is never good. The transmitter is off the air. You call the remote control and try to put the main transmitter back on the air. No good. The backup comes up, no problem. Shaking off the sluggishness, you get dressed and head out the door. The transmitter is about 30 miles away, but it’s the middle of the night, so there is no traffic. While driving, you are thinking of all the things that could be wrong. The blower motor was sounding a little loud last trip. The exciter has some reflected power. The PA tube is two and a half years old.
Upon arrival, there are several overload lights lit, including the driver plate. An investigation is in order. You turn everything off and open the doors. The trouble seems to be a bad IPA power supply. There are spares on the parts shelf, so you put one in. Put the transmitter into the dummy load. You turn on the filament and the transmitter comes to life again. Reset the overloads.
Broadcast Electronics FM35A transmitter ready to be turned on
Now you are standing there looking at the plate on button. Was it really only the IPA or was that just a symptom? Was there something else that took out the IPA power supply? What will happen when I press the plate on button? Will it come on normally or go BANG! I hate BANG! By the way, my tradition in a situation like this, if on a mountain top somewhere, I go outside and pee. I give the situation one more run through the mental checklist, then come back inside and press the button.
Broadcast Electronics FM35A transmitter high voltage on button
Please excuse the blurry picture, it is hard to take a picture of yourself turning on a transmitter…
Gone is the day when the radio station engineer had to trouble shoot down to the component level, often crawling in and out of transmitters to get at the suspected part. I for one, spent many a long night at a transmitter site chasing some weird combination of symptoms down to the $0.34 1N914 diode in the directional coupler (see previous post about the MW-50).
It is a skill set now mostly confined to manufactures’ repair departments, for which they charge a pretty penny. Nowadays, the technician simply slides out one module or circuit card and slides in another. If that doesn’t fix it, panic ensues. I know of several class C FM radio stations that are now relying on the computer guy to fix transmitters, because, you know, it’s cheaper.
To be fair, most engineers are contractors and many of those simply do not have the time to trouble shoot to the component level. So, they ship everything back to the factory then pass the cost on to their client.
Then of course, most circuit boards these days are surface mount systems, which are hard to work on if you don’t have the right tools. Normally an expensive temperature controlled soldering station is required, as well as a magnifying glass.
All of these things combine make circuit board work something to be outsourced. Unfortunately, a night spent trouble shooting was often a great learning experience. I have done some of my best work when my back was up against a wall and I was out of options.
I make the attempt to fix things locally, unless the transmitter or other item is under warranty or not having a spare/attempting to trouble shoot will take the station off the air. I think it is important to keep abreast of technology and keep my trouble shooting skills up to par. Besides, I find it gratifying that at least I can still fix things.
We have a Harris Z5-CD transmitter for one of our FM stations. Brand H is not my preferred make, however, it was already installed when we bought the station, so I have to live with it.
This particular site gets hit by lightning strikes often. Normally, it does not affect anything until the transmitter gets turned off for maintenance. Then, almost invariably, when turning the transmitter back on one of the modules will fail. Most often this is manifest when one of the two power supplies shut down causing the transmitter to run no more than 20% power.
The way this is trouble shot is to slide each module out and turn the transmitter back on. When the power supply stays on, the bad module has been located. A confirmation test is to check the MOSFET for a short circuit between Drain and Source. This short circuit condition puts a direct short on the power supply causing it to crow bar and turn off.
So, once the bad module has been located, and the spare module is installed in the transmitter, then what? Most engineers call Harris and ship the module back for repair. Most engineers don’t want to mess with unsoldering a surface mount MOSFET and soldering a new one in. I find it moderately entertaining to fix things myself, so I do not do what most engineers do.
NXP BLF177 MOSFETS
The MOSFET in this particular module is the BLF177, made by NXP. Harris will sell you one for quite a bit of money. You can also buy one from Mouser for about half the cost.
Harris FM Z series transmitter PA module with cover removed
Once the parts are obtained, the worst part of the entire job is unsoldering the old MOSFET. This takes some patience and skill. What I found works best is to melt some solder on the foil leads and get them good and hot. Since this MOSFET is already destroyed, we don’t have to worry about heat etc. The one thing you do not want to do it actually break the MOSFET open. That is because it contains beryllium oxide, a known carcinogen. Once all the solder is liquid, carefully pry the foil up with a small screw driver. There are several components that have to be moved to work on this.
Harris Z series PA module with MOSFETS removed
After the old MOSFET is removed, clean up the solder pad with a solder pump and solder wick. I like to use a little liquid flux on the solder wick, it makes things go faster.
Once all the old solder is cleaned off the solder pads, I brush a light coat of liquid flux in the pad. Again, this makes things go faster.
Harris Z series FM transmitter module new MOSFETs waiting to be soldered
The new MOSFETS are very sensitive to static discharge, so I always use a static drain wrist band when handling. I place both MOSFETs on to the circuit board. I then solder them on using as little heat as possible from the soldering iron. Again, the MOSFETs are sensitive to heat and one can easily be destroyed if it gets too hot.
Harris Z FM series PA module repaired
This is the module with the new MOSFETs soldered in. I use defluxing compound to remove all the extra flux. Once it cools off, I test the new module with a DVM:
Harris Z series FM PA circuit board under test, resistance is 3.3 Mohm
If the MOSFETS are good, they will have an internal resistance of around 3.3 MΩ. If the module is bad the MOSFETS will read only a few ohms if shorted:
Harris Z series FM PA module under test, DVM reads 1.6 ohms
That is how you do it. I think Harris charges $775.00 per module to repair. I fixed this one for $240.00, but that is not the reason I did it. I did it for the fun that was in it.
Congress shall make no law respecting an establishment of religion, or prohibiting the free exercise thereof; or abridging the freedom of speech, or of the press; or the right of the people peaceably to assemble, and to petition the Government for a redress of grievances.
~1st amendment to the United States Constitution
Any society that would give up a little liberty to gain a little security will deserve neither and lose both.
~Benjamin Franklin
...radio was discovered, and not invented, and that these frequencies and principles were always in existence long before man was aware of them. Therefore, no one owns them. They are there as free as sunlight, which is a higher frequency form of the same energy.
~Alan Weiner
Everyone has the right to freedom of opinion and expression; this right includes the freedom to hold opinions without interference and to seek, receive and impart information and ideas through any media and regardless of frontiers
~Universal Declaration Of Human Rights, Article 19
Open Mic