notes on incorrect and excessively high (ridiculous) power measurements when excessive drive and volting are applied and the case for one obvious cause of this anomaly. "When you apply more volts than 13.8 to 2879's there are no specs to show results, just because you say it doesnt exist doesnt make it so." oh there are plenty of maximum ratings provided in the manufacturers specifications that when read and properly understood more than explain why these transistor power output claims are nothing but smoke and mirrors. let's begin with the most revealing of the maximum ratings provided in the case of the 2879. Collector Power Dissipation: 250W: that's the collector INPUT they're talking about. now here's a question that needs answered. how can a device that's rated for 250W maximum input dissipation produce output levels of 400W + unless there's something going on that is not readily seen and understood? you plug in any efficiency percentage you want to there for any class of amplifier operation and multiply it by 250 and see what you get. furthermore, power gain (Gp) for the 2879 is 13db.. at 5W in that's 100W output or 40% of the maximum collector input power dissipation, assuming here for the moment that the gain of the device doesn't begin dropping off as power output is increased and that's a lie. as conduction of the device rises above the turn on "knee" and compression begins to take place the gain of the transistor begins to deteriorate, not actually a desirable trait in an amplifier being modulated with an asymmetrical signal source. @ 400W output that's 160% of the same. so we're supposed to believe that a transistor can produce more power than it consumes. i don't think so. haven't any of you wizards figured out yet why the power supplies required to generate these mythical power figures keep getting larger and larger? maximum input power is listed at 10W at 10WPEP input the output from a single device will be in excess of 130W but IMD levels will be trashed without the addition of suitable low pass filtering. coincidentally the output level is >50% of the maximum collector dissipation rating at this point and transistor Gp has fallen to almost 11 db.. beyond this point we begin to see energy developed that is NOT OCCURRING at the fundamental frequency. there's some of your maximum ratings. dbg = 10log(Pin/Pout) dbg = 10log(130/10) dbg = 10log(13.0) dbg = 10(1.113) dbg = 11.139 the largest factor contributing to the highly erroneous readings produced by the majority of hf wattmeters is created when energy begins to appear at frequencies 3, 5, 7 (and higher) times the fundamental operating frequency. the more the drive and operating voltage is increased the more pronounced the effect becomes. in addition to the sum addition of the energy present at these multiples of the operating frequency the problem is compounded by the wattmeter design itself. the line section is designed to sample a specific amount of energy at the fundamental operating frequency in a direct relationship to the physical length of the line section as that length correlates to a given fraction of a wavelength at the intended operating frequency. when energy begins appearing at the odd harmonics not only is the power added by the line section but it is also amplified because as frequency is increased the line section becomes longer with regard to the fractional wavelength represented by the line section initially at 27 mhz.. there fore the line section is "oversampling" the energy levels occurring at the frequencies previously mentioned. once these harmonics begin to occur the environment is ripe for a sort of controlled "oscillation", the harmonic energy present acting as the trigger. now in addition to the energy present at 27 mhz. you now have substantial levels of vhf and in some cases even uhf energy directly affecting the accuracy of the measuring device. if the line section is say 1/100th of a wavelength (typical) at 27, as energy appears at the odd harmonics at 81, 135 and 189 mhz. the line section becomes exceedingly many times longer in terms of wavelength and oversampling occurs. since at the higher harmonic frequencies the line section is sampling incrementally longer portions of a single wavelength this produces increasing ly higher readings and measurements. this is probably the largest contributing factor to the rid iculous power measurements. you can easily prove this to yourself by inputting a predetermined amount of vhf transmitter power into your hf wattmeter. pre-measure say 1 watt of vhf power using the proper vhf wattmeter and a dummy load and then after the output power is confirmed then feed it into your hf wattmeter set to an appropriate scale and see what happens. your hf wattmeter will indicate several times more power present than the 1 watt of power seen on the vhf wattmeter. the end result is that literally hundreds of watts of power (in this example) as indicated by the average hf wattmeter is NOT OCCURRING AT THE FUNDAMENTAL OPERATING FREQUENCY. unknowingly or otherwise, this deception is being created by excessive drive and operating voltages. (spectral purity of the exciter signal not withstanding) i have seen some suggestions bandied about that could limit these erroneous readings such as using line sections and slugs that are frequency restricted in their design so as to eliminate the oversampling problem by rejecting the energy occurring at odd harmonics above 54 mhz and the use of effective low pass filtering. that might end up costing some of you a lot more money. it's a lot simpler to just adhere to the transistor specifications and follow the manufacturers recommendations regarding operating voltages and input drive levels. it's easier all around if the IMD energy isn't generated in the first place. the bottom line here is that amplifier buyers are being sold a "bill of goods" that does not perform as represented and advertised. 2X2SC2879's will NOT PRODUCE 800W OF ANY KIND OF POWER AT THE FUNDAMENTAL OPERATING FREQUENCY. there's the disclaimer. adjusted line section target length = 4.34 inches, approximately 1/100 of a wavelength @ 27.185. look below to see how the line section becomes longer in terms of wavelengths as energy appears and is measured at odd harmonics at IMD3, IMD5 and IMD7 levels which are the direct result of overvolting and excessive drive. as you can see, at frequencies above 27Mhz. and successfully higher there is a much larger sampling of energy as the line section becomes longer in terms of fractions of a wavelength. this greatly exaggerates the real power occurring at the fundamental operating frequency. 1/100 of a wavelength @ 27Mhz. 1/50 of a wavelength @ 54Mhz. 1/37.5 of a wavelength @ 81Mhz. IMD3 1/25 of a wavelength @ 108Mhz. 1/21.875 of a wavelength @ 135Mhz. IMD5 1/18.75 of a wavelength @ 162Mhz. 1/15.625 of a wavelength @ 189Mhz. IMD7 1/12.5 of a wavelength @ 216Mhz. 1/11.71875 of a wavelength @ 243Mhz. ODD 1/10.9375 of a wavelength @ 270Mhz. 1/10.15625 of a wavelength @ 297Mhz. ODD 1/9.375 of a wavelength @ 324Mhz. 1/8.59375 of a wavelength @ 351Mhz. ODD 1/7.8125 of a wavelength @ 378Mhz. 1/7.03125 of a wavelength @ 405Mhz. ODD 1/6.25 of a wavelength @ 432Mhz. while the line section is actually slightly shorter (inch fractional) than 1/100 of a wavelength @ 27mhz. the proportions represented by the fractions regarding wavelength are accurate. the adjusted figure given at the top of the page is a direct result of: 984/FMhz. the correction was approximately +0.04....from actual measured length of the line section. i admit the measurement was made hastily but not enough of a difference exists to really matter. my purpose here is moreso to provide a concept for consideration and experimentation than it is to obtain absolute accuracy, albeit both are important. we have known for quite some time that in comparison to the fundamental power levels that these other "products" were relatively small by comparison but this does begin to explain the exaggerated power claims and why we're seeing all of this imaginary energy when referenced to the actual operating frequency and the manufacturer's specifications regarding the devices used in this equipment. i am by no means precluding the presence of other contributing factors as well and would like any input available along those lines. ref: http://www.firecommunications.com/trdbase/2sc2879.pdf 292 Radio Shop / FireCommunications Network email me @: http://www.firecommunications.com/site/contact.cgi a thorough discussion of this topic will begin soon here in the Technicians Exchange Forum: http://s10.invisionfree.com/Technicians_Exchange/