The linked Guidelines for tuning a transmitter to Class-E are useful for the general library of knowledge on the subject, but in this particular arena any mention of Class-E automatically means we are referring to the SSTran AMT5000, and the facts are quite different in that case.
Obviously the tutorial applies to high-power transmitters feedlines of undetermined impedance, in which case extreme care must be taken to avoid damage to vital components within the system.
PhilB has made it clear on numerous occasions that the AMT5000 is deliberately designed for the known and expected range of impedance that will be encountered with a short 3-meter antenna located in open space free of obstacles. In any case, the AMT5000 is never in danger of damage from mis-tuning because of the small power level involved.
Also, PhilB has explained over and over that he has engineered a reliable means of adjusting for Class-E mode as described in the Manual without the need for an oscilloscope, for the likely fact that many Part 15 users will not have access to such equipment.
I know PhilB's method is real-world, because it worked for me and continues everyday to outperform all my other transmitters.
We can reasonably realize that the author of the linked article did not have Part 15 in mind when presenting his paper.
The ALPB has all the answers.
Did you read all the tutorial? The was an explanation for tuning without an O-scope and since amateur radio utilizes power from milliwatts to kilowatts, one cannot make the assumption that the article for stictly for high power.
Hello John WDCX.
I did read the whole article, but found it in one turn contradictory and in another imprecise.
At the very opening sentence it clearly says: "The first time a class E rig is put into operation, you MUST use an oscilloscope to ensure everything is working correctly."
There is no room in the opening statement for a later exception.
In the later section TUNING WITHOUT AN OSCILLOSCOPE it repeats the same claim made at the beginning: "Once you have verified your transmitter operation WITH AN OSCILLOSCOPE, you should be able to tune using the drain current meter." That is contradictory, because it doesn't really tell us how to tune without an oscilloscope.
The final section, USING AN EFFICIENCY METER, is lacking in detail, and although it hints at the possibility of having a "one meter serves all solution," it doesn't actually say that an oscilloscope will thereafter NOT be needed for initial setup.
You are so picky.
PLEASE, all readers, download and save the linked article for your AMT5000 E-Class Library. It is a fair and legit article that can benefit from SPECIAL CASE disclosures, as I've tried to provide, making it compatible with the AMT5000 tuning instructions.
Left in its original form the link might (unwittingly) seem like a discrediting of the AMT5000 Manual.
The ALPB is working behind the scene to make it all good.
Yeah it was not my intent to discredit SSTRAN. I just ran across it and thought that itmight be of some benefit.
http://www.google.com/url?sa=t&rct=j&q=&esrc ="s&source=web&cd=1&ved=0CCgQFjAA&url=http%3A%2F%2Fwww.norcalqrp.org%2Ffiles%2Fclass_e_amplifiers.pdf&ei=8vlHU4vaLYLfsAS2yICQDQ&usg=AFQjCNFp7mTHYy14HZ4GGIc_WjXIphAVOg"
Nicely detailed information about the ups, downs, ins, outs, and back and forths of everything Class E. Perfect for the growing library collection on the subjct.
For me there is only one problem, the document shows a "goof proof" Class E circuit, which rules me out.
Thanks for posting the N7VE article. Many write-ups regarding class E circuits stress only the theoretical 100% efficiency yet this one presents the real world implementation advantages and problems.
The need for a "hard driving" input signal and the impedance transformation at the output are usually not mentioned in most discussions.
Setting the input power (100 mW for Part 15) also sets the output impedance of the amplifier so the designer is not free to change this and must either resort to an impedance transformation network and/or a transformer. When I designed my amplifier I found that impedance transformation using a network was not feasible and I included a transformer as did N7VE.
The author's comments about no tuning are correct in my experience in that once the operating power and frequency is set then the circuit will tolerate a wide range of load impedance (20 to 100 ohms in my case) with only a few percent loss in efficiency.
His 80 to 90% efficiency statement is realistic in my experience with this type of circuit and is consistent with what I measured. There are losses in the output network.
The problem of high drain voltage came up during my experiments and this needs to be watched. Other not so obvious problems are that FETs have a relatively high input capacitance and a very high drain to source output capacitance. Though these can be accommodated it makes sense to use a bipolar junction transistor as the output device to mitigate these. In theory a BJT will have less efficiency than a FET since the BJT has a non-zero vce(sat) whereas the FET presents a small drain to source resistance yet I found no measurable difference in efficiency at 100 mW input between a FET and BJT circuit. Using a BJT overcomes the high voltage problem (high voltage and fast BJTs are available) as well as the high device output capacitance (which allows the tuning shunt capacitance to be set by a fixed passive component.) FETS are an instinctive choice for switching to ground but BJTs also perform this function very well.
N7VE is to be complimented for making this circuit work at several times the part15 AM frequency since parasitics can cause problems. This is a good reference article for those working with high efficiency RF amplifiers.
Neil
That article makes a Class E transmitter sound like a complicated and hazardous undertaking. He's talking about a 5 W transmitter for the 7 MHz and 14 MHz ham bands using ultra-cheap transistors. Ham transmitters are designed to drive a 50 ohm load connected to a self-resonant antenna or an external antenna tuner that presents a 50 ohm load while resonating the antenna. This transmitter has a low Q output network consisting of a very low-Q resonant circuit to satisfy the Class E requirement and additional low pass filtering.
The configuration for a MW 100mW transmitter feeding a 3 meter antenna is a LOT simpler, primarily due to the much lower power and the much higher Q of the output network. All of his points under the heading "Problems with Class E QRP Amps" quite simply do not apply at all to a Class E Part 15 MW transmitter.
A Part 15 Class E amplifier only requires a few components for excellent performance, stability and reliability. Take a look at this article discussing a 1 W Class E "lowFER" beacon transmitter: http://www.k3pgp.org/Notebook/Wd5cvg/Classetx/classetx.htm This article is much more appropriate for understanding a Part 15 Class E transmitter.
One thing that caught my eye is the graph labeled "Figure 5 - Output Power and Efficiency vs. Lcoil Adjustment". You can click on the graph to bring up a more readable view. I have done similar plots for the AMT5000 and my results are consistent with Figure 5. In my plots, curve B corresponds to the meter readings at the "RF INPUT CURRENT" test points. Note that the current peaks to the left (higher frequency side) of the high efficiency region. For the AMT5000, the transistor efficiency is about 82%. at the peak current point. The AMT5000 tuning procedure calls for first adjusting the trimmer to peak input current and then adjusting the trimmer one full turn clockwise (about .7 pF additional) to reduce the tuned frequency slightly to move into the high efficiency region where the transistor efficiency is 96% - 98%. Note that the high efficiency region is rather broad, so the required slight decrease in frequency is not a critical adjustment. Overall transmitter output efficiency is lower due to coil loss, but the change in overall efficiency exactly tracks the change in transistor efficiency.
I keep seeing people whispering down the lane that tuning the AMT5000 is difficult. What's so hard about adjusting a trimmer to get a peak meter reading and then turning the screw one turn clockwise?
Just as PhilB says, it is not difficult to properly tune the AMT5000 into a conventional part 15 AM antenna. I did it. Many others have done it.
Of all those who've built, tested, and reported on their AMT5000s, all of them, to a man, have had high praise for the experience.
The only debunkers are always characters who do not actually have an AMT5000, hissing from the sidelines without disclosing their true motives, which could be anything from promoting another product or simply serving as a disciple to a third party debunker.
Get over it. The AMT5000 is the leader in Part 15 History and as long as I'm around, that's the way it is.
Thank you, Phil, for the link to the article.
The maximum power transfer theorem is often misapplied without understanding that, though maximum power TRANSFER occurs, the efficiency is 50%. In this case the author acknowledges this and the addition of the resistor does stabilize the circuit since this tends to approach a constant current source for the final. The negative feedback mentioned is the real advantage at the price of a drop in overall efficiency.
A question remains for me though and this is how does one measure the "transistor efficiency"? My assumption is that this is done by measuring the DC drain current (average) and the DC drain to source voltage (average) with the product being the DC input power and then dividing this into the output power. Next question is where in such a circuit does one measure the output power? It seems the author does this by measuring the RMS current through the loading coil and calculating the "load" power using the Rloss + Rrad numbers. Since the loading coil R is included in the Rloss this method places the Rcoil in the load rather than in the source. Since the desired performance measure is the power delivered to the antenna system including the Rcoil as part of the transmitter rather than the load would give a more useful indication of the amplifier efficiency than just the "transistor efficiency".
For example, one transmitter with a high loss coil and another with a low loss coil could have the same efficiency (that is, the same output power) if the coil loss is counted as part of the load yet the power delivered to the radiator will differ (assuming the same ground loss for both).
By convention the output power would be measured at the load which would be after the loading coil and the coil loss would reduce the efficiency quite a bit but the result would be better in terms of comparing different amplifiers with internal antenna loading coils.
Those familiar with farming know of the Nebraska Test for tractors which resulted from the disparity between engine horsepower and the power available to do the work. The Test measures the PTO and drawbar horsepower which is the power which does the work and which is always less than the engine horsepower. I see a similarity here.
Neil
Over at another neutral site, I got the originator of the Challenge to effectively admit that they likely did not tune the AMT5000 correctly. He still is blaming the documentation, and defending the test on the basis that from the beginning, he intended to use transmitters as delivered, and set up according to the documentation provided. That is all fine and good, to a point, but it certainly shows that the Challenge results have to be taken with a grain of salt. I certainly would not use the results if I wanted to choose the transmitter which would give me OPTIMUM field strength, even if I did have to work a little at it and do some more research.
And, of course, as I pointed out over there, if they had used the AMT5000 as it was delivered from the manufacturer, they would be testing a bunch of unconnected electronic components.
... it certainly shows that the Challenge results have to be taken with a grain of salt. I certainly would not use the results if I wanted to choose the transmitter which would give me OPTIMUM field strength, ...
Just to note that the ratings shown in the DeFelice Challenge report were based on the relative fields produced by the unmodulated carriers of those systems.
Even if the radiating length of the antenna system, the carrier frequency, the d-c power input to the final r-f amplifier stage, the propagation path, interference conditions, earth conductivity, measuring equipment/settings, and the resistance of the r-f ground connection were the same for all of those system tests, that does not necessarily eliminate the possibility of setup errors in that Challenge sequence, and/or the possible biases of those who were involved in conducting and documenting that Challenge sequence.