RF Column #45 - August 1995 - Doug Lung Copyright (c) 1995 H. Douglas Lung ALL RIGHTS RESERVED TOPICS: Acrodyne Au60D 60KW Diacrode transmitter - Comprehensive tests! Avoiding transient problems in transmitter remote controls "Cheap Remote" enhancements using the Micro-485 RS-232 serial port isolation UPS's and Surge Protectors Update on the ITS-830 1 KW Solid state UHF transmitter Announcements: RF Page debuts, disks discontinued ---------------------------------------------------------------------- Just as time was running out for finishing this column, I heard Acrodyne finally had their 60 KW Diacrode transmitter ready for testing. Bill Hayes and Dane Ericksen invited me to join them while they evaluated the transmitter. I'll let you know how it turned out. I continue to get requests for diagrams and software for the "Cheap Remote" I wrote about some time ago. Since then I've expanded the remote control and added a number of enhancements. This month I'll reveal some techniques I used in my new Blue Earth Micro-485 based remote control circuits to avoid problems with control line transients and offer some tips on making remote control installations more bullet- proof. You might want to check it out if you use a computer based remote control and occasionally experience some weird control behavior. Finally, I launched the "RF Page" on the World Wide Web June 26th. If you are looking for back issues of this column or other material related to broadcast RF, check it out.. As you may have gathered from my June column, I was both excited and a bit skeptical of Thomson's 60 KW (combined visual / 10% aural service) Diacrode tube. Some of that skepticism remains, but after seeing the Acrodyne Au60-D transmitter in operation with the Diacrode at Acrodyne's plant in Blue Bell, Pennsylvania I'm much more confident the tube will work. It was a bit of a shock to see a transmitter no larger than Acrodyne's first 5 KW transmitter putting out 60 KW! First I'll tell you the good news. We verified the transmitter was capable of putting out 60 KW peak of sync visual and 6 KW aural power. The bad news was the tube was unable to produce a peak visual power of 60 KW for more than a few seconds with APL's less than 30 before shutting off with a screen current overload. This problem grew worse as the testing proceeded and appeared to be related to some sort of screen grid to control grid leakage. Bill Barrow, who, by the way, is now back at Acrodyne working on their high power transmitters, said this problem developed after about a hundred hours of use and they felt it was not a design problem with the tube. I had to finish this column before Acrodyne had a chance to get another tube working. Since the tube had no problem making peak sync power as long as we kept the APL's high enough, we were able to take some real performance measurements on the transmitter. The linearity and intermodulation correctors are similar to those used in the 1 KW solid state TLU-1KE transmitter I reported on in June and the performance was similar as well. Please keep in mind the results I'm about to give you are from Au60-D transmitter serial number 1 using a tube which clearly had some problems. As Acrodyne and Thomson learn from these first tests I expect to see the numbers improve. One test used to evaluate the performance of common-amplification transmitter systems involves measuring the in-band intermodulation product between the aural subcarrier at 4.5 MHz. and the chroma subcarrier at 3.58 MHz. (for NTSC systems). The resulting 920 KHz. intermodulation product will appear as a interference in the picture. Comark claims that in-band modulation products will be at least 60 dB below peak visual carrier in their high power common-amplification I.O.T. transmitters. Acrodyne managed to reduce the average 920 KHz. intermod in the Au60-D to 58 dB below peak visual carrier as measured using an HP spectrum analyzer at 100 KHz. resolution bandwidth. There are several different ways of measuring intermodulation and they will give different results. Both Acrodyne and I.T.S. engineers told me they prefer using a modulated ramp, since it tests intermodulation over the full range of video levels. The test used for measuring intermodulation for C.C.I.R. standards uses three separate carriers. The visual carrier is set 8 dB below the peak visual carrier power. This corresponds to average visual power. The chroma carrier is set 17 dB below peak visual carrier power and the aural carrier is set 10 dB down from the peak. It is easy to optimize intermodulation using this test because the carriers are fixed in level. In the real world, with different levels, results will vary. Dr. Tim Hulick (the designer of the Au60-D) showed me how to set up these levels using a old Tektronix 147A test generator. The 147A will accept an external signal which it adds to the internally generated video. The 147A's subcarrier output is attenuated and fed back to this input. By setting the 147A's to produce a pedestal at 55 percent APL and setting the chroma at 80 IRE units peak to peak, the result is a set of carrier levels matching those requires for the C.C.I.R. tests. One thing I learned from these measurements is that when intermodulation levels drop as low as we were seeing here the resolution bandwidth of the spectrum analyzer will have a significant affect on the reading. For example, when the HP analyzer with 100 KHz. resolution bandwidth showed a -58 dB intermod product, a Tektronix 2710 with 300 KHz. bandwidth showed between -54 and -55 dB. Another test used to evaluate common-amplification transmitters is the amount of spurious aural carrier deviation produced by the video sync pulses. I moderated a session on this topic at the SBE convention in Miami in 1993. At that time many of the common-amplification transmitters sold had a significant amount of spurious horizontal sync frequency harmonics present on the FM baseband signal. Most of these transmitters were used in low power TV applications and were not transmitting stereo. Problems caused by the spurious deviation were few. When broadcasters starting using common-amplification transmitters for high power levels and demanded cleaner audio for stereo, Comark responded first by developing a correction circuit to reduce this spurious deviation and most other manufacturers followed to some extent. Dane Ericksen developed a technique for measuring this spurious modulation using the baseband output of a Tektronix 1450 operating in the split audio carrier mode and a spectrum analyzer usable down to 10 KHz., which includes the Tektronix 2710. Since the Tektronix 1450 demodulator is now history and the replacement 1350 doesn't have split audio carrier demodulation available, an H.P. modulation analyzer was used for these tests. We found the Acrodyne Au60-D's spurious aural modulation varied with the power level. At the optimum power level, which was approximately 90%, the spurious modulation from sync spurs less than 100 KHz. from the aural carrier was 2.8%. The spurious modulation including spurs above 100 KHz. was 7.2%. Dane was satisfied with this performance, but, like me, would like to see them duplicated at 100 percent power. Since the results were worse at both higher and lower power levels, we suspected that the corrector was optimized at 90% power and could be re-optimized at 100%. I also checked the more traditional performance measurements -- differential gain, differential phase and low frequency luminance non-linearity. Differential gain was less than four percent peak and differential phase was within two degrees. Low frequency luminance non-linearity was very close to ten percent. This is similar to what I was on my Boston Acrodyne 1 KW solid state transmitter. When the linearity circuits on Acrodyne's exciter are optimized for low intermodulation products, the low frequency linearity suffers. Acrodyne is working on this. ICPM was measured at 5.4 degrees peak to peak on the VM-700A. One of the difficulties Acrodyne faces in correcting the Au60-D non-linearity is a difference in the correction curves required for the Class AB solid state driver stage and the Class AB tube output stage. While the transmitter was operating at 60 KW peak visual power and 6 KW aural I took down some of the meter readings. Those of you familiar with klystrons might find them interesting. The plate had 8500 volts on it and was drawing 9 amps average with a modulated ramp as the video source. The screen grid was drawing 200 mA at 695 volts and the control grid showed a positive 50 mA at negative 113 volts. (The grid readings are not normal as mentioned earlier.) The filament meters showed 8.9 volts at 180 amps. The cooling system seemed to be working well - the power amplifier inlet temperature was 37 degrees C and the outlet temperature was 44 degrees. It's too early to give you my verdict on the transmitter. From what Bill Hayes, Dane Ericksen and I saw at Acrodyne there is no longer any doubt the transmitter can meet current broadcast performance standards. What needs to be tested more is the tube itself. Can it handle 60 KW peak visual power and 6 KW aural power day in and day out without fading? I've proposed after Acrodyne obtains a new tube they try to maintain the transmitter at sync and 75 percent blanking level at 100% power for a 24 hour period. If the transmitter is stable under that load for that amount of time, it should survive fine in the field under normal conditions. Acrodyne seemed willing to undertake such a test. If and when it happens, I'll let you know the result. A few years ago I developed a remote control circuit to monitor our various LPTV stations around the country. The first design used the Blue Earth Micro-440 micro- controller and a simple interface circuit consisting of some op amps and opto-couplers for isolation. I published the latest revision of the circuit for that interface in this column two years ago. Since that time Blue Earth Research has introduced their Micro- 485 controller, which includes a more accurate A/D converter (12 bits versus 8 bits on the Micro-440) and also adds an 8255 I/O port circuit. The 8255 IC is a product of the personal computer industry. Until the advent of higher level ASIC chips and high speed parallel ports, you could find this chip used to provide the parallel printer port in most PC's. I was a bit hesitant to use the 8255 in a transmitter control circuit because of its known tendency to latch up when subjected to transients. While not common, I've seen examples of weird behavior of other remote control systems that use the 8255 after heavy lightning storms. The usual solution is to power down the unit and bring it back up, which forces the 8255 to reset. My first solution was to create a pseudo 8255 using standard 74HC373 octal latches, an idea I have to admit I stole from an old EDN magazine Design Idea. I would have stayed with that design, but after reviewing the spec's on the Blue Earth Micro-485, I found that unlike the Micro-440 it did not provide access to the CPU bus. My final design ended up as a hybrid, using the Micro-485's 8255 output to drive multiple 74HC373's which are used for command control and internal multiplexer control. I also used 74HC373's to store status information which the 8255 could grab at the same time it was sending command data. Because of these latches, I can reset the 8255 regularly without worrying about it switching the transmitter off. I can even remove the Micro-485 from the circuit without changing the command outputs. Being a firm believer in Murphy's law, I added opto-isolators to each command output and status input. I have one of these new remotes working on top the Prudential Tower in Boston and so far (knock on wood) no problems. The real test will come after my second Micro-485 based unit goes on line on a hill top outside Abilene Texas, under a tall tower at the end of the power line. Here are some tips to avoid weird control problems with your remote control system. You may find some of these in the manuals included with the remotes. First, if the remote control outputs and inputs are not opto-isolated, consider building up a circuit to do that. Use a separate power supply with plenty of isolation from the remote control to supply any voltages you need to drive transistors or status LED's on the isolated side of the coupler. You will still have analog connections to the remote control, so don't ignore grounding. Rack screws can't be depended on. When using opto-couplers for reading status lights make sure that the opto-coupler's LED isn't being lit by leakage currents in the system. I've had problems with this in some transmitters. I've used zener diodes, regular diodes and shunt resistors to keep this current out of the coupler. If you experience intermittent remote control problems, resist the temptation to try to solve them by hanging capacitors and RC networks across the inputs. Often you will only succeed in changing the problem from one that happens on a regular basis to one that happens on an infrequent basis at the worst time. Most remote control systems I've seen do not react well to power line glitches. Computer uninterruptible power supply systems (UPS) are so cheap that I now automatically put one on every remote control installation. One side benefit is that if the communications circuit stays up, you will have control even if the AC power dies. Nothing is more frustrating than seeing the transmitter drop off the air and having no way to check what happened. Many newer high power transmitters come with the remote control integrated with the transmitter control logic. This eliminates many problems but creates a few scary new ones. First, if you have to bring a phone line into the transmitter for the remote control connection you are also creating the potential for a lightning strike on the phone line to come into your transmitter control circuitry. The best telephone line surge protector I've used is made by PolyPhaser. Like any surge protector, it is only as good as the ground to which it is connected. Second, if you have a serial cable going into the transmitter from a modem, the protection is only as good as the protection you can give the modem. In this case, however, there is a very easy, very effective additional level of protection available. Several computer accessory companies now sell optically isolated "short haul" modems for connecting RS-232 serial lines between buildings and power services. I've used the Black Box ME-800A modems with good success in Miami. Before installing these modems at our uplink site I saw one Compaq computer completely toasted and more burned out RS-232 interface IC's than I can remember. The problem was caused by lighting hitting a nearby STL tower or the uplink dish, raising the ground potential at that location above that in the computer room. Since installing the ME-800A modems that circuit has had no interruptions. One update on last month's column -- I.T.S. informed me they have corrected the ITS- 830's sync level drift problem I mentioned last month. The production version of the revised board should be available by the time you read this. Also, based on what I saw during the intermodulation testing at Acrodyne, I may have overstated the in-band modulation product on the ITS-830. The Acrodyne TLU-1KE measurements were done using an HP analyzer at 100 KHz. bandwidth while my I.T.S. measurements where done using a Tektronix 2712 at 300 KHz. bandwidth. As a result, the I.T.S. unit's performance was probably closer to that of the Acrodyne unit. Note, however, that the I.T.S. operates at only 5% aural power while the Acrodyne TLU-1KE operates at 10%. Regarding my comment about the ITS-830's collector voltage, I.T.S. said that metered voltage is measured at the power supply and not at the amplifier transistors. Voltage drops in the wiring and circuitry to the transistors reduce it by a few volts. I'm now on the Internet World Wide Web! Point your browser to http://www.gate.net/~dlung/rf.html to reach the RF Page. I'm constantly updating it as time permits. All the files I mentioned last month are available through links from this page as well as back issues of my columns. By the time you read this I should have columns through 1993 on line. The listing includes the topics discussed in each column. Although I do have pictures on the RF Page, the columns are in ASCII text format and easily readable with any browser, even text based only ones like Lynx. I plan to convert the ASCII columns into HTML so you can jump to referenced columns, figures or sites easily. Until I have the columns in HTML format I will have to grab them from Cybergate's FTP server, which has a limit of twenty users. If you get a busy or error message when you try to grab a column, keep trying. Usually it doesn't take more than half a dozen tries to get through. While I may offer teasers or surveys for future columns, do not look for current columns on-line. They will only be made available after the printed version has been out for at least a month. Future plans for the RF Page include adding background information and updates on items presented here, information from RF manufacturers, reader surveys and links to other sites I think you will find useful. If you have a site that might interest my readers, drop me a note at dlung@gate.net and I'll consider adding a link to it. If you are involved with video broadcast RF, whether satellite, uplink, downlink, microwave, VHF, UHF or the equipment to control, test or monitor it, and have a short news item, product announcement or useful material (i.e. free programs or tutorials) that would benefit other readers, I'd like to include it. I do not plan on offering advertising at this time, so any material I publish on my page will have to have some practical, educational or news value beyond selling products or services. Obviously, there is some flexibility here. For example, a catalog listing of RF transmission line from a manufacturer that includes line attenuation and other planning information is clearly commercial, but also very useful, so I'd probably use it. I'm paying the costs of running the Web page myself and intend to continue to do so as long as the costs stay low enough. One reason is it gives a lot more freedom to express my opinions. It also doesn't interfere with what allows TV Technology to come to you every month at no charge -- paid advertising. On another topic, I `m no longer offering to mail disks of programs I've written and mentioned in these columns or copies of old columns. The reason is with my travel and work schedule it has been taking me months to catch up with the requests and I know I've disappointed some people. On line access is so cheap and universal that anyone should be able to obtain this information from either my FTP site (ftp.gate.net in the directory /pub/users/dlung) or my RF Page on the Web. Compuserve, Prodigy and America On Line all offer Internet access and often have free time offers. Netcom's Netcruiser account is only $20 a month and includes more time than you should ever use in that fee. Local Internet providers offer inexpensive access in most cities -- ask at one of your local computer stores. Most schools and many libraries now have Internet connections. I'm receiving Internet email from readers as far away as Indonesia. Now that my articles and programs are so widely available, you'll find it much faster to download the information you want from the Internet than to wait for me to find time to ship you a disk. I welcome your comments, even if you don't have an on-line account. My email address is dlung@gate.net. Compuserve users can use my Compuserve PPN - 70255,460. You can also fax me at 305-884-9661 or phone me after 6 PM eastern time (when things quiet down a bit) at 305-884-9664. Both numbers are at the Miami Telemundo office, so expect a delay in a response if I'm traveling. My mail service address is 2265 Westwood Blvd., Suite 553, Los Angeles, CA 90064. If time is critical contact me for a local address. Copyright (c) 1995 H. Douglas Lung ALL RIGHTS RESERVED