For operation around 1600 KHz
Change these parts:
Change C23 from 820pF to 560pF.
C5 replace with 1800pF fixed capacitor, polycarbonate 1800J.
Remove L4, L5, L6, L7 and replace with a single toroid coil, 15 uH, 34 turns wound on a T106-2 iron core.
Set all S5 switch position to OFF.
-- Mike
Interesting modification which you describe as converting the transmitter to drive a 50 ohm antenna. You do not state that this converts the output impedance to 50 ohms which would not be a good approach so I assume this is not what you did but rather modified the output network to accept a 50 ohm load..
Question, what 50 ohm antenna would be available for Part 15 operation? For example, I use a 3 meter vertical with a base loading coil over ground radials and the impedance varies greatly due to weather. Usually the feedpoint impedance at the coil input is close to 30 ohms. Is the toroid you added performing the function of a loading coil? Is the actual antenna a 3 meter wire? Just trying to understand this better.
Neil
I suppose the Isotron might be 50ohm.
The modifcations are from SStran. I am guessing one could use a 75 meter Ham Stick syle antenna with a home brew coil and still comply with the 3 meter rule.
http://www.craigwilliams.com/radio/mobile/160meter/index.html
The part which caught my interest was adding the toroid because it appears in series with the antenna output which suggests that this functions as the loading coil with a 3 meter wire antenna. This is not a 50 ohm system so I asked some questions to make sure I wasn't misunderstanding or overlooking something.
Maybe Mike will clarify this.
Neil
From the opening post, "...toroid coil, 15 uH, 34 turns wound on a T106-2 iron core...."
A ~9-1/2 foot high x 3/4" OD, base-driven, base-loaded monopole with its base within 9" of the earth needs an inductance of about 290 µH to resonate it at 1600 kHz.
Maybe Mike also can elaborate on the use of the 15 µH coil in the proposed modification.
______
Impedance has a "real" term (R) and an imaginary term (± j). For example, the impedance of a perfect 50 ohm r-f load is expressed as 50 ± j 0 ohms.
The real part* of the feedpoint impedance of this antenna system consists of the radiation resistance of the monopole plus the r-f loss in the system connection to r-f ground plus the r-f loss in the loading coil needed to achieve resonance.
Typically that real part can range from about 30 ohms to over 100 ohms, however only the radiation resistance (about 0.1 ohm) produces the radiation of radio waves.
At a given operating frequency, the feedpoint reactance (j term) at the base of that monopole depends only on its length, and OD.
The function of the coil is not to match the transmitter to the real part of the feedpoint impedance (which it can't do), but to offset the capacitive reactance (-j) of this electrically short antenna system.
*only the real part is able to dissipate/radiate r-f energy
It looks like its a low pass pi net
The changes I describe are to modify the pi-network output circuit. The original (unmodified) pi-network serves the purpose of an impedance transformer to match the 820 ohm output impedance of the AMT3000 RF amplifier stage to the very high impedance of a 3 meter long antenna. My intent was to modify the pi-network so that the transmitter could drive a 50 ohm antenna rather than a very short high impedance antenna. The toroid coil is not part of the antenna but rather a replacement of the four coils (L4 - L7) that are part of the pi-network.
I'm still fine-tuning the modifications and may also need to make modifications to increase the current capability of the RF amplifier stage.
Thanks for clarification, that makes sense.
Neil
That output stage is really 820 ohms?
Yes, the output stage is 820 ohms. The output amplifier stage is configured as a differential amplifier. This differential amplifier is essentially a pair of BJT common emitter amplifiers. The output impedance of a common emitter (CE) amp is set by the collector resistor (R18 820 ohms). If this was a balanced diffferential amplifier (both outputs being used) then there would be a second 820 ohm resistor on the collector of Q4 and the output resistance would be 1640 ohms (2Rc). However, only one side of the amp (Q5) is used to drive the pi-network, thus 820 ohms output impedance.
I have some things about the AMT3000 to consider. I have not analysed the output circuitry but recall reading (a while ago on a FIDO board which is long lost) that when this transmitter is used with a base coil loaded antenna that the antenna is tuned slightly off resonance to present an inductive component to the transmitter which in conjunction with the internal capacitors and the antenna parasitic capacitance provides an impedance transformation of the 3000 or so capacitive reactance of the radiator to 820 ohms.
Pi networks are used for matching and perhaps the residual L combined with the Cs on each side of the loading coil performs this function.
If this is true than there would not be a need to include the toroid to match 50 ohms and then use a loading coil to produce a 50 ohm load especially since the extra inductor wires add significant loss.
If there is another 50 ohm antenna available then this modification could be used possibly to advantage.
Neil
The based-loaded antenna often described for use with the AMT3000 is a specialized design that does take advantage of the output stage capacitors of the AMT3000. However, like all base-loaded antennas, the efficiency is low and the base coil losses are high. plan is to use a center-loaded antenna with the AMT3000. A center-loaded antenna has a higher efficiency compared to the based-loaded antenna but has a feedpoint impedance of about 35 to 50 ohms. Therefore, modifying the AMT3000 output stage to drive a 50 ohm load is required.
From Reply 11 "... thus 820 ohms output impedance. ..."
The feedpoint impedance at the base of a ground-mounted, 3-meter, vertical monopole antenna system at the upper end of the AM broadcast band ranges from about 30 -j 3000 ohms to about 100 -j 3000 ohms -- mainly depending on the losses in the connection of that antenna system to r-f ground.
A loading coil could resonate such a system. If it did, then the load impedance at the coil input would be on the order of 30 ± j0 ohms to 100 ± j0 ohms.
A transmitter expecting/modified to drive a load impedance of 50 ± j0 ohms then could be driving a load having rather high SWR, and that system would radiate relatively less of the output power of the transmitter.
It all comes down to maximizing the current flow in the vertical portion of the antenna. Higher current in the antenna results in a stronger radiated signal. A center loaded antenna has considerably more current flow compared to a base loaded antenna; the current in the bottom portion (below the coil) is close to maximum possible (and similar to the current flow of a full size 1/4 wavelength antenna. The base loaded antenna exhibits no such current flow. An illustration of this is at this link: 
As far as SWR losses; a 50 ohm transmitter driving an antenna impedance in the range of 30 to 100 ohms would result in a maximum SWR of 2:1 which is a loss of about 11.1%. This is perfectly acceptable given the significant increase in antenna efficiency. The antenna can be further adjusted to achive the matching 50 ohm impedance resulting in a 1:1 SWR.