What I meant is the difference when actually measuring into that pure 50 ohm load vs crunching it out in numbers. Variances always comes into play due to tolerances between one pure 50 ohm load and another, even if that tolerance is a fraction.
Curious...why does Phil not chime in here and provide that useful information directly? I am sure you would be able to convey it accurately..but what better way than for all here to hear it from the very individual who designed those transmitters?
Why enclose it all in an email to just one individual?
Even posting that info on the SSTran website would be useful.
Just my thoughts on that part.
RFB
hes very busy filling orders and designing and figuring costs for the amt5k external case.
"hes very busy filling orders and designing and figuring costs for the amt5k external case."
I might believe that if there wasn't enough time to send the 1 email to 1 individual in the first place when that time spent doing so was put into posting on the forum answering the Op question. Sort of tips the hand IMO.
Anyway not to get off thread topic....I put in the spotlight to this issue Bruce's approach, bypass all the internal stuff meant for a short piece of wire and replace with a network meant for a 3 meter loaded whip or stick.
Should make a neat weekend project or at least an evening project.
RFB
hi rfb,
in the interest of getting back on topic i am going to post phil's email. i hope he wont mind me repasting this. if he does then I AM VERY SORRY and can delete it.
The base loaded vertical antenna shown at
http://www.sstran.com/pages/COMMON/sstran_buildant.html is functionally
identical to the 160 meter base loaded mobile antenna described in the ARRL
Antenna Handbook. To connect to this antenna, the AMT3000 needs to be
modified slightly as descried at
http://www.sstran.com/pages/COMMON/sstran_amt3000mods.html.For the following, the base loaded antenna described in the provided link is
assumed. Other loaded short vertical antennas from the ARRL handbook, such
as center loaded, top hat, and screwdriver antennas are compatible.The inductance of the loading coil is chosen so the inductive reactance is
equal to, and opposite to the capacitive reactance of the antenna element
resulting in resonance at the operating frequency.The capacitance of the antenna element is constant (independent of
frequency) and depends on its length and diameter. The capacitance of a 102"
CB whip antenna is about 24 pf and the capacitance of the 106", 1/2" copper
pipe is about 30 pf. There is a formula for this in the handbook.The Q of a base loaded antenna is very high, so careful tuning to the
precise resonant peak is mandatory. Tuning is done by first selecting the
best coil tap and then fine tuning by adjusting the length of the antenna
element. If the loading coil is a continuously variable inductor, the
antenna length need not be adjusted.The actual feed point resistance of the tuned antenna depends mostly on the
coil resistance (about 10 ohms) and the ground resistance. The ground
resistance is the most widely varying part. It can range from 10 ohms for a
good ground radial system to 50 ohms for an antenna with no radials and a
basic ground rod. The feed point resistance is the sum of the coil
resistance and the ground resistance yielding a range of 20 ohms to 60 ohms.The AMT3000 performs at its optimum output level when is sees an 800 ohm
load. If the load resistance is higher, the RF output waveform will start
clipping on modulation peaks. If the load resistance is lower, the amplitude
of the RF output will be lower resulting in lower power output.To match the transmitter to the base loaded antenna, an L-network is
introduced by simply placing a capacitor from the transmitter antenna output
to ground. This is done internally when the modifications are implemented.
Then the transmitter simply connects directly to the loading coil. When all
four S5 switches are set to ON, the internal inductors are all bypassed and
C5 and C23 are connected in parallel to ground. Since C23 is a trimmer,
there is a small adjustment range, but the range isn't really adequate to
cover the possible feed point resistance range of 20 ohms to 60 ohms.
Perfect match occurs when 13 VDC is measured across the test points T1 and
T2 with no modulation. For any particular feed point resistance, the L-net
can be adjusted by varying the value of capacitor C23. The recommended value
of 560 pf is good for fairly decent ground systems (at least a few radials).
If the max voltage is less than 13 VDC with the trimmer cap at minimum, C23
can be replaced with a slightly lower value, like 470 pf. I always
recommend that if 13 VDC cannot be achieved, it is better to improve the
ground radial system rather than reducing the value of C23 since this will
provide the greatest range increase.Hope this helps. Let me know if you have further questions.
Phil Bolyn
SSTRAN
Hi dietnews!
Great! Now there is more info to add to the record books.
Side note: If blame is warranted..point it my way...I'm used to it and shrug it off with a "whatever".
RFB
so it looks like when you do the base loaded antenna mod you are creating an L-Network that matches 800r to 30r. the network consists of the loading coil and a combo of C5/C23.
when i get a chance i am going to do a couple simple experiments. first i am going to remove C5/C23 and then check the Pout using an 800ohm dummy load. then i am going to install a simple L-network to match 800r to 50r. i'll check the Pout with this matching network as well.
i'll probably do that next weekend. i was planning to do it this weekend but i got behind on a lot of other work and haven't had a chance to mess with this. i'll post info and photos after i have completed the experiment.
Yes..basically what I've been saying as well as what Bruce did on his unit...create a new matching network for a loaded antenna system or 50 ohm antenna to replace the matching network already inside the unit meant for short pieces of wire.
RFB
Deleted per poster's request. See his post below.
Note that you should have been able to edit your post but deleting a post is prohibited because doing so deletes all comments below the post.
Neil
That should be a lot more than 30mW...RMS not PEP. Interesting.
RFB
hi. i just tried to edit my previous post but the site isnt letting me. i would appreciate if a mod could edit or delete my previous post. here is what i wanted to replace my previous post with:
note: my preliminary report on using an 820ohm load on the output of the pi-network is totally meaningless. i did it just out of curiousity to see what would happen but it says nothing about the transmitter's performance.
after that test i bypassed the pi-network and saw 2.6Vrms into an 820ohm load (8.2mW) with an unmodulated carrier. with a modulated carrier it is producing 30mW which seems reasonable.
phil has told me that i might have some other problems with my tests. he gave me some detailed info on what the output should look like with everything past C22 bypassed (which i have already done) and R18 removed. tomorrow i will remove R18 and see what happens.
Don't know about you or anyone else but getting 30mW modulated to me is NOT reasonable at all! Getting 70 or 80mW with an UNMODULATED carrier IS reasonable.
Yes perhaps it is something in your testing (wonder what that could be), or there is a fault somewhere in that transmitter (possibly the problem), or it doesn't output what it is supposed to output (doubtful but until proven otherwise cannot be ruled out). One of those three will be the explanation.
Were you monitoring the finals voltage and current drain during the testing? Is the scope or the probe or the resistor load loading down the output stages..when it should not be?
Of course to validate your tests and results, the testing will have to be duplicated by another individual with an AMT3K using the same testing techniques and whatever butchering..err bypassing is being done to the unit to REACH for that expected number.
Here's a quick thing to check...is your scope probe on 1x or 10x?!!!
That one really get's em angry when they discover that's what was giving them incorrect measurements! I've done that too sometimes and I even laugh at myself for it! It's amazing how a slightly dirty scope probe multiplier switch can muck up measurements!
IMO..that unit should be outputting what it claims to be outputting, and the standard used to measure AM transmitter RMS is with an unmodulated carrier. At the very least that unit should spit out 50mW unmodulated...at the VERY LEAST and that is probably worst case scenario...unless your results will now be declared the worse case scenario, which is cheating and no different from the NRC re-adjusting the safe limits on the radioactive danger levels chart.
Eyebrows raising left and right now I bet.
RFB
two things:
1. unless we know the class of operation is C,D, or E, 30mW from ~100mW is not unreasonable. 30% efficiency is on the high end for Class A and a little on the low end for Class AB but not unreasonable.
2. i think there are definitely problems in my testing procedure. for me the main purpose of these tests is to educate myself on how to test a transmitter's functionality and how the amt-3000 operates.
on that note, my test procedure can use a lot of improvement. i am not using a proper 10x probe currently. i have alligator clips hooked to C22 and ground on the amt-3000 which are then hooked to the leads of my load (820ohm resistor). a bnc->alligator clip test lead is hooked to my oscilloscope and across the load as well.
i normally would use a 10x probe but mine broke recently. i am actually buying a new one later today so i will report back any changes in my results when i get a chance to use it.
i also removed R18, as per phil's instruction and now i am seeing 6.2Vrms into the 820ohm load with a modulated carrier. that would be 46mW which would actually be about 50% efficiency.
"i am seeing 6.2Vrms into the 820ohm load with a modulated carrier. that would be 46mW which would actually be about 50% efficiency."
So that would equal about 11.5mW unmodulated. That isnt no 50 percent efficiency.
Since your learning how to measure an AM transmitter, might as well measure it like any other AM transmitter is for mean power, ie unmodulated, then you can factor the efficiency for RMS. Then measure for PEP efficiency, which you have done.
But that isnt no 50 percent efficiency off the bat. Your getting UP TO that point by modulation, not by mean RMS carrier power.
But its all turning out to be quite interesting as your going along with the measuring.
RFB
This is not to feed a rumor mill and is only for a point of interest but years back I simulated the AMT3000 output stage using an 800 ohm load and including R18 and the simulated output power was not much different than dietnews reports.
Having said this, don't place much stock in this since the simulation program I used was rather primitive and no actual physical measurement was made by me.
Neil
sorry i have been slow with updates. my initial poor output results appear to have been a result of not removing R18 (which is really important when using an external matching network and not using any matching network at all) and my use of a 1x probe.
according to phil, R18 is used to contral the max output level when using the internal pi-network and should be removed if the pi-network is not being used.
using a proper 10x probe on the output side of C22 and the entire pi-network bypassed this is what the carrier looks like in an 820ohm load: 
note: the output is square because there is no matching network.
4.8Vrms into 820ohms equates to a 28mW unmodulated carrier.
phil described the output of C22 with no matching network as being a square wave with 5.75Vrms (40mW into 820ohms). i think my output is somewhat low because of stray capacitance in my test setup--i think this because the output wave is not totally square.
phil also said that he actually as the amt-3000 setup for 90mW input into the final transistor to ensure that the transmitter passes fcc inspection. this would mean that his described 40mW output would indicate 44% efficiency and my (possibly still flawed test results) of 28mW indicate 31% efficiency. these results are pretty typically for an AB amplifier.
i dont have time to explain the results with my 800->50ohm matching network but i have it working and it i am seeing similar power output into a 50ohm load. i'll make another post describing those results when i have the time.