Excellent! That is good news. I must admit, I am curious to know what was wrong with the original controller. As for changing the feedline, don't bother, the difference in attenuation is insignificant. In order for there to be a noticeable difference in the signal strength at the receive end, you would need at least a 3dB improvement, which is half an S-point, and even that would probably not be all that noticeable. The other thing that feedline loss does is mask the antenna's real SWR, tending to make the system seem better than it really is. But again, this would only be noticeable for a significant change of feedline loss. For example, RG-8X is specified as 0.9dB/100' @ 10MHz, and 2.1dB/100' @ 50MHz, RG-213 is 0.62dB/100' @ 10MHz, and 1.5dB/100' @ 50MHz. At 10MHz therefore the difference is 0.3dB and if you read my previous post you will understand when I say that fractions of a dB are of no consequence. Even though the loop does not cover 50MHz, the difference there is only 0.6dB, again, insignificant, so leave well alone, there is no benefit to changing. Remember the engineering advice, "Don't Fix What Ain't Broke!" Anyway, I am glad that your problem is resolved, we had fun doing some detective work, and I'm always glad to try and help. Now go and have some fun!

73, Adrian, 5B4AIY

Hi Adrian,

Well, I received the new ATU 2.0 this afternoon and I have good news! The controller now finds my Baby Loop during the Auto Detect Routine which would thrill me to no end if that was all. But, I can now make nice large jumps in frequency without the dreaded SWR or Motor faults.

My testing is still in the preliminary phase but it sure is a lot better performance than it was before. Still have a bit of a noise hump on 40m but it's pretty flat on 20m so not sure what that is about. Like I said before, I had those before with my MFJ Loop so perhaps it's a local thing.

I'm still thinking of upgrading my feed line to RG213, but looked up the specs and really would not gain much in loss, plus it's not as flexible as RG8U.

I want to thank you again for your help with this, you have been a big help for sure!

73 Jim K4ECP

OK, I have now made some measurements of my controller. Using a precision 50 ohm dummy load known to be better than 1.05:1 from DC through to 3GHz, this was connected with a very short piece of coax to the ANT socket of the controller, and, using the TEST facility the SWR was measured. At all frequencies from 6,600kHz to 9,000kHz it measured 1.0:1. From 10,000kHz to21,000kHz it was 1.1:1, and from 22,000kHz to 29,800kHz it was 1.2:1. This shows the intrinsic accuracy of the C-M controller's SWR bridge. The same test was then repeated using my scruffy MFJ 100W dummy load, and from 6,600kHz to 9,000kHz it was 1.1:1, from 10,000kHz to 29,800kHz it was 1.2:1. I then inserted a known good 20m length of Ultraflex-7 and repeated the tests, with no change, showing that the cable was good. Yesterday I measured the output power of my controller and it was +11dBm +/-0.5dB over the entire frequency range using my Siglent spectrum analyser. Bear in mind that the C-M controller does not need to actually 'measure' the SWR - in all probability it simply uses the reflection bridge's reflected voltage, the SWR is calculated and displayed merely as a convenience for the user. At least, if I were designing it that's what I would do. This then should give you a baseline as to the accuracy of the measurements you are making. As I said, the ACCURATE, and I stress ACCURATE and REPEATABLE measurement of RF power is a challenging task, and for amateur equipment if you are within +/-10% you are doing well. As you go higher in frequency, the challenge increases. Even professional power meters are not usually specified to better than +/-0.5dB. As my old professor in university was fond of saying with a dismissive wave of his hand, "There's no such thing as a fraction of a dB in RF measurements! You'll be lucky to get the same measurement 2 days in a row!" Add to that the fact that most digital instruments indicate their results to an excessive degree of precision, and it's no wonder there is a certain amount of confusion. I certainly always had to remind my technicians that just because it displays the result to 3 decimal places does not mean it is accurate to 3 decimal places! Accuracy and resolution are two different things. OK, I've had my morning rant, I feel so much better now.

Adrian, 5B4AIY

Hi, Jim,

Yes, there are some aspects of your feedline sweep that bear further investigation. As a (now retired) professional satellite communications systems engineer I have to say that accurate and repeatable measurement of RF power, and SWR are some of the more 'challenging' things to deal with. If your feedline is new, has properly fitted connectors, and, most important, is completely free from sharp kinks, then there is little reason to suspect it. If you compare your findings with the sweep I posted, you will see that in my case my feedline is flat, the SWR is less that 1.05:1 everywhere, but, this was using a precision 50 ohm dummy load guaranteed flat to 3GHz. So your irregularities might well be due to the MFJ load itself, at least if its anything like my MFJ dummy load. Equally, if you made these measurements using the TEST facility of the C-M controller as I suggested, then we are assuming that it is reasonably accurate as well. You can cross-check your readings by using your Flex radio as both the signal source and the measurement. If you get essentially the same results then it is probably the dummy load, after all, the MFJ dummy load is not a precision device. You can further check this by using a short piece of coax and connecting it directly to the Flex and making some measurements. You do not have to repeat all the frequencies, just those that are a bit on the high side for this test. If your Flex shows the same results using a short length of coax, then it is probably the MFJ load. You could also put the MFJ dummy load directly on the ANT connector of the C-M controller and check it at a few frequencies to see what the intrinsic accuracy of its internal SWR bridge is. I'll put my precision dummy load on my controller and make a few measurements for comparison and get back to you. I have to say, this is an interesting exercise. More to follow.....

Adrian, 5B4AIY

C-M finally agreed that the ATU is probably at fault after seeing the sweep that I sent them. So I am just waiting to hear from the vendor that I purchased it from. I also did the feed line sweep and it was not as flat as I expected. However, the only 50 ohm load I had was an MFJ...However I think it's acceptable -

I have a 50 ohm load and will run the same test today to check the feed line

73 Jim K4ECP

Hi, Jim,

Yes, your SWR is not unreasonable. If you turn it into a graph:-

then you can see that apart from the 2:1 point the rest is OK. True, it is not as flat as I would have liked, but it is not unworkable. As for the frequencies between about 18 and 21 MHz, yes, I can see why the loop might have a problem there, but I have seen worse, and the loop has worked. As I said previously, provided that far from resonance the SWR is high and constant, then the tuning algorithm will easily locate the resonance point. I presume that these results were obtained using the Ciro-Mazzoni controller. What is more revealing is measuring the SWR with a dummy load at the end, which takes into account the feedline. Although this is best done with a swept-frequency analyser, it is possible to do it with the C-M controller. Simply terminate the loop end of the feedline with a good 50 ohm dummy load, and then using the TEST mode enter the various frequencies and observe the resultant SWR. Tedious, to say the least, because in order to enter the next frequency you have to press the BACKSPACE key, and then press and hold the / key to re-enter the test mode, but it can be done.

Adrian, 5B4AIY

Hi Adrian,

Tech Support requested that I do a frequency sweep with the loop in it's present location which is attached. I don't see any issue with the loop, feed line or location that is causing an SWR issue, do you???

Hi Adrian,

Great information thanks...while looking at the display while it's tuning I can see where some of that action is occurring. It seems happy with any frequency in the 40m band as it tunes 7MHz frequencies quickly and accurately, Anything other than that, it seems to have a problem. Sometimes 20m is fine and other times trying to tune any 20m frequency is a real bitch. And pretty much forget anything higher than 18MHz.

I see what you are saying about feedline issues...however I did replace my feedline just a week or so ago with no change in my situation. I had the exact same issue with the new feedline as I did the feedline with the splice. I keep going back to the spurious nature of the ATU. Just like the feedline not being flat confusing the alogorithm, I think a dirty generator could possibly cause the same thing. Undoubtedly you are much more familiar with these systems than I and it seems that your work has to do with something along these lines.

73 Jim

Hi, Jim,

Although Ciro-Mazzoni have not shared their source code with me, there are only a small number of ways this sort of thing can work. I have written the software for similar systems, and typically when far from resonance the algorithm expects the SWR to be both high and constant. In this state the actuator would be driven at maximum speed and the SWR constantly monitored until a reduction is observed. At this point the algorithm would switch to the fine tune mode where the actuator would be stepped in small increments. At each step the SWR is measured and compared to the previous value. If it is less, then another step is made in the same direction. Eventually either the measured SWR is the same as the previous step or there is an increase (a reversal) which means that you have reached the resonance point. In the case of the C-M loop the frequency is also switched either side of the resonance point presumably to ensure that you have reached the optimum point. The algorithm can be confused if the feed line is not flat. If the algorithm detects that the SWR increases at some point then it assumes that you have passed the resonance point and will now reverse the motor direction and search for a reduction in SWR in the opposite direction. Eventually the actuator will be driven into its electrical limit, the motor current will reduce to zero which will be detected by the controller and the famous 'MOTOR FAIL' error message delivered and the controller shuts off. So again, if your system constantly seems to be driven in the wrong direction, despite the motor sense being correct, then this points to some sort of feedline impedance discontinuity. This can only really be measured or displayed using a graphic antenna analyser or a vector network analyser. What you do is disconnect the loop from the feedline and terminate it with a good 50 ohm dummy load. Then measure the SWR at various frequencies and verify that it is 1:1 everywhere in this condition. With a swept frequency network analyser this is simple to do.

This is a plot of my Ultraflex feedline when terminated in 50 ohms, using a RigExpert analyser, and, as you can see it is nice and flat everywhere. Also attached is my overall SWR.

Adrian,

I just re-read your original post, and you were having the same issue that I am that when making a large frequency move, it drives to the minimum position and throws the SWR error. You found a feedline problem and it was solved. I also had a splice, however my coupler was pristine...clean and tight. You are using RG213 and I am using RG8U. My run is 125' and my run is also new as I just replaced it less than 2 weeks ago.

C-M is aware of what I'm using for feedline and has not said that they think it's an issue. I would be willing to spend the extra $$$ for a 213 run if it would help.

What I don't understand is why it would drive to the opposite direction than it needs to go in. Is the controller that stupid that it does not know which direction it needs to drive??? I can't believe that it is - So something must glitch it somehow to make it do what it is doing - I would really like them to explain to me why it's doing what it's doing...

Hi Adrian,

Just checked manually and the polarity of the motor wiring is correct - 7 open and 1 close. It just seems so odd that when I command it to auto tune from 20m to 10m instead of loop "opening" it goes to the minimum capacitor position. I can manually tune any 10m frequency with excellent SWR. I sent C-M all the ATU 2.0 pictures with all the spurs and have been asking them do they think it's at fault...nothing yet.

Hi, Jim,

OK, if you have not managed to get it to auto-detect, then the controller will not have determined the correct 'sense' of the actuator wiring. You will need to make sure that it is connected in the correct sense manually. Go to the TEST facility, and press the 9 key and observe which way the actuator moves. If it opens the antenna then you have to correct sense, if it tries to close the antenna, then simply reverse the motor connections. The keys are, 9 - fast open, 8 - medium speed open, 7 - slow open, 3 - fast close, 2 - medium speed close, and 1 - slow close.

I must apologise for not getting you to check your firmware revision earlier, I had assumed that you had the latest version as you had just purchased it. It simply goes to show, Never Assume - It Makes An Ass Out Of U & Me!, and, I fell into the trap. My apologies.

Adrian, 5B4AIY

Hi Adrian,

I heard back from tech support and they asked which version of software I was using in the ATU saying that 4.05 was buggy. It came with 4.05 so I upgraded to 4.06 and there was some improvement...It still fails Auto Detect but it does try now to tune from 20m to 10m! Before it would just immediately throw the SWR error, now it actually displays the tuning screen on the Flex and the motor drives the capacitor but in the WRONG direction. It drives it closed instead of further open - and the of course throws the SWR error - At least there is SOME forward movement -

73 Jim K4ECP

Here's another odd one, this was displayed on the Flex while I was in TEST Mode manually tuning 24.930

Hi Jim,

Yuk! That looks really dirty. Attached are my tuning signal and band using the same frequencies as you

AS you can see, my tuning signal is quite clean, but notice the small spur at about 14.360, and another at about 14.240, compare that with the same thing without the tuning signal.

As you can see, there is quite a difference between your generator's output signal and mine. I really think you have a controller problem. I have examined the tuning signal with my Tekronix scope, and it is essentially a square waveform. Similarly, when examined with my Siglent spectrum analyser, it is fairly clean. I'll do some tests and post a screen shot of its spectral purity.

73, Adrian 5B4AIY

Hi Adrian,

OK, unplugged PS and noise hump still there, also had an MFJ Magnetic Loop before this one and had a hump with that one as well. Must be something to do with my setup I guess - Thanks for the thought though! Oh, C-M supplied a 2 amp supply.

73 Jim

OK...will explore that. Here is pic of controller cross fed into Flex, looks nasty to me -

Hi, Jim,

Overnight a thought fell into my head, probably because there was nothing to stop it. My setup is different than your because I am not using a C-M 24V PSU. I found that they were very noisy, and so I designed a linear 24V/2A PSU, and that's what I use. To see if it is the PSU, tune as before, but then disconnect the PSU or switch it off at the mains, and see if your 'noise bump' disappears. Also, check the rating, some loops were supplied with a 24V/1A supply, but that cannot always supply the peak current pulse, and later antennas are supplied with a 2A supply. I would recommend a 24V/2A linear supply to eliminate all possibility of noise. At my QTH here in the mountains the background noise level is around S2/S3 on 20m as there is absolutely no industry or other sources of electrical noise, it's very quiet. Even 40m after dark is quiet! The 24V PSU does not need to be regulated, as the 24V is only used by the H-Bridge motor controller IC and the actuator. There is an internal regulator the drops the voltage down to that required by the electronics. In fact the actuator's motor will operate down to very much lower voltages, so precise regulation is unimportant. As long as it can supply the initial peak current pulse and has an output voltage of at least 20V it will work fine, and will be free of noise.

73, Adrian, 5B4AIY

I also cross fed generator into the Flex and it looked a lot more spurious than yours did, but I was in SSB and not CW so will also try in CW to see how it looks.