Category: Magnetic loop antenna

Well, the replacement stepper controller chips arrived early this morning. Didn’t take too long to replace them and connect everything up for a test.

The good news is that after a slight adjustment of VR2 the stepper appeared to be controlled by the rotary encoder. It made the right noises.

Try as I may, I wasn’t able to carry out the initial setup procedure as outlined by Loftur. I couldn’t find a peak in the noise.

After applying a portable SWR analyser directly to the loop when I thought I was at the lowest frequency point the loop would tune, I discovered there was a dip around 23 MHz! I was nearer the other extreme of what I think this loop’s range should be. So I’m thinking I may have the stepper motors wired wrongly with the result that the stepper goes in the opposite direction to what the controller thinks it should be.

After switching one pair of wires at the controller end, the stepper made a terrible noise as it hit the end of the capacitor’s travel. Checking it tentatively it didn’t seem to distinguish between clockwise and counterclockwise movement of the encoder. Whatever you did, the stepper turned in the same direction.

Time for further closer inspection of the wiring around the newly re-oriented common mode chokes and their associated bypass capacitors. This kind of behaviour may have something to do with a missing or failed connection in this part of the circuit.

Stepper only appears to turn CCW which turns the VVC CW. I’m now hoping the glitch is pretty obvious in this part of the board or associated wiring.

Finally found some time to re-orientate the two transformers. I decided to remove the eight bypass 10nF capacitors to give my soldering iron more wriggle room to remove the transformers. The solder on the transformer pads needed some encouragement – in the form of more molten solder – to loosen up. This allowed me to use solder wick to take away most of what was holding them in place. The corners of the transformer are pretty fragile and would not tolerate much heat. Wasn’t too hard to lift them. Much more complicated was cleaning up the holes where the eight capacitors had been. I managed to destroy one of the pads. Decided to stop destroying the PCB and to re-install the bypass capacitors on the other side of the board.

I was actually able to use most of the capacitors I had removed and checked every connection a couple of times at least, especially a couple of the tiny pads for the caps that are near the larger choke pads.

Soldered the transformers back in properly this time, again checking every step over and over again.

Taking a few deep breaths now and having a break before reassembling the whole device and connecting it to the radio and the loop for the moment of truth and to find out if the A4975 stepper driver ICs can come back to life after having both their two outputs accidentally connected together! Normally they are connected across one of the stepper motor windings.

——

Well, because there is still no response from the stepper motor, the answer, unfortunately, is no, they probably need replacement. At least this time they’re not red hot. I have a pair of A4975 chips due to arrive sometime on Monday. Essentially $4.90 a pop, no shipping charge.

And reading the datasheet on the A4975 I see on page 10 mention of the thermal protection circuitry that shuts off the output transistors when the junction temperature reaches +165 degrees C. “This is intended only to protect the device from failures due to excessive junction temperatures and should not imply that output short circuits are permitted.” And I reckon the tiny choke winding is close enough to a short circuit.

At least we are still at three out of four. Looking forward to being able to control the stepper to tune the vacuum variable capacitor and calibrating the controller, and perhaps opening myself up to a new level of complications. At least the well-commented code should be a smart guide.

In the meantime – getting way ahead of myself now – I’m thinking about drilling two holes in the loop to take the stepper wiring, possibly via a length of CAT 5 cable. People seem to think the twisted pairs work well with stepper windings. Not sure what type of connector to use in VVC part and lower down alongside the feed loop SO-237.

I also need to get a couple of plastic pipe ends to fit the PVC tube cover Henrik gave me to weatherproof the stepper and the tuning capacitor.

Then I need to think about a rotator and maybe a tripod roof mounting. And a way of remotely controlling the rotator. Anthony K3NG most probably.

At least it looks like it might be ready to take away to Tooraweenah for a field test in March!!!

Maybe my mistake has been not to refer more directly to the actual code. Over the last 24 hours I have been steadily working through the last two years of email conversations. Loftur keeps the group informed of each new version of the firmware and the newly added features. Sometimes he jumps the gun, but then releases a newer version shortly afterwards.

Well, well well! I think I have discovered why the final part of the loop controller is not working and perhaps why the A4975 stepper motor driver ICs are running so hot. I have just been doing some continuity testing working back from the stepper motor connector, K1 and noticed that there is continuity between C11 and C12 and also between C15 and C16, and there is no continuity between the legs of say C16 and C14 connected T2 etc, all of which implies that I have oriented the common mode chokes, T1 and T2 exactly 90 degrees out from where they should be – I think and I hope… They are both surface mount with four points of contact(!!!!), so there might be some interesting re-working to be done. I actually have one spare if I bugger one up. Also I may need to replace the eight 10nF ceramic bypass caps nearby. But this is progress of a kind.

I wonder what made me install them this way. Surely not simply the orientation of the writing on the package!?!?!?, or the oblong shaped pads on the PCB?!?!?!?! Two visual cues to be misled by!

I suppose I should be optimistic about the transformers being okay if they still show continuity… There’s a big lesson here about using the schematic during the construction and not simply populating the board. And maybe even being curious enough to actually turn the component over to see its magic revealed and be inspired to orient it correctly.

And now of course when I look back to page 8 of the BOM and building instructions I can clearly see the correct orientation of the transformers. Oh me oh my!

And I have just noticed during my rapid review of the emails on the Loop Controller list that on 8 May last year, while advising someone about testing the sense of the stepper winding connections, Loftur dropped this comment, “Also, have a careful look at the circuitry around the A4975 stepper controllers. One fairly common mistake is to mount a common mode choke incorrectly.” I’ll say!

After a long break I have finally got back to completing the Magnetic Loop Controller designed by Loftur Jonasson that I started almost two years ago now. I am using his original PCB version from June 2014 which indicates how long I have been intending to build this device!

I noticed a small SWR board kit based on the kitsandparts kit offered via the Loop Controller list by Frederik ON5IA.

I ordered a couple, but while I was waiting I checked where I had got up to with my original build.

I completed my original SWR bridge board based on the kit from kitsandparts.com. I worked out I could use the same larger gauge bus wire (as used for the single turn in each ToroidCalc1 transformer) to securely and rigidly mount the SWR board to the inside of the BNC sockets. I discovered later that one of the joints to a BNC centre was intermittent, ie not actually soldered!

With the Controller apparently complete, I connected a 4-pin socket to the four wires coming from the stepper motor on the loop and attached the stepper to the VVC with the shaft couplers. I also spent quite a bit of time confirming the correct way to connect the main PCB to the SWR board, to the serial port, to the Stepper motor as well as all the other connectors and controls mounted on the front and back panels of the enclosure.

While the Teensy appears to be working as expected there was no apparent connection with the radio or the loop!!! So far the score is just one out of maybe four! (I suppose my four targets are 1. Teensy controller, 2. Serial connection with the radio, 3. Connection to the Stepper motor and 4. SWR Autotune.)

I did notice scratchy noise due to the dodgy non-joint on one of the BNC sockets. Also the Controller would occasionally just stop and reboot, especially if the enclosure was tilted slightly!

I checked continuity of the earth/-ve line. Seemed intermittent. Noticed I had used nylon stand-offs to mount the PCB. But there was a connection to the metal enclosure between the PCB ground plane via the heatsink tab of the 5 volt regulator. Decided to install metal stand offs. Needed added height of one washer to match original height for the USB socket to match the hole in the rear panel. Also confirmed that the -ve connected directly to the enclosure via the power socket mount.

Properly soldering the antenna lead eliminated the scratchy noise from the radio.

I had last installed firmware version 3.04. I updated the Arduino IDE and re-installed the Teensyduino libraries and then installed the latest version 3.08 of the Loop Controller firmware. No actual difference with connection to radio or loop.

Reading over the last couple of years of conversations on the Loop Controller Yahoo group I came across one post where someone had experienced problems connecting to their KX3. Loftur replied with a check about the way the TXD and RXD is connected to the serial lead. Mine was wired opposite to his description, as I may have confused the order on the serial header and the serial socket. Flipped the TXD and RXD connections and immediately I was able to see the frequency readout on the LCD update after a change on the KX3. Satisfying! A review of the schematic confirmed I had misread it earlier.

This was with the Elecraft K3 Auto setting on the Controller. The controller responds to changes of frequency on the KX3 but only after there is a change. And the update of the information in the frequency display feels a little laggy and as if the update periods are longish. Switching over to the polling method Elecraft K3 Poll the update is much faster and smoother. It also recognises band changes before a change in frequency. (Perhaps I need to change the KX3 setting for AutoInf back to its default). I am using a standard stereo patch cable between the serial port on the controller and the Acc1 socket on the KX3.

One step at a time! Of my four targets then, it looks like we may have achieved two!!! Now just SWR & Stepper to go.

I suppose I should be able to test the SWR board. I have a dummy load connected to the controller ANT BNC. Perhaps I could even calibrate the Power and SWR???? I have checked the ML.h file in the firmware package and read the section in the BOM and Building Instructions pdf (page 13) about the settings to enable the Power/SWR and Autotune options. #define PSWR_AUTOTUNE had been set to 0, so I set it to 1 and verified the program and re-programmed the Teensy. All pretty speedy. And it works, but because frequency ranges have not been set yet it seems to read most frequencies as out of range.

Definitely making some significant progress today!!!

Now to attack the issue with the lack of connection between the controller and the stepper motor.

I examined the solder joints of the two common mode chokes and tested them both for connection and continuity. I have also noticed that the two stepper driver chips are running very hot, way too hot to keep your finger on them for more than a touch. Also there’s probably a static discharge risk there too. Given their cost is only about $7 for these two ICs, it might be worthwhile replacing them, while taking proper antistatic precautions. It would be good if I could discover any other reason/s there is no linkage to the stepper and more importantly why they get so hot so quickly. I hooked up another NEMA 17 stepper motor directly to the board and there was again no movement, but the chips immediately heated up. With the stepper disconnected the chips still heat up.

Time to read more of the manual. That’s the only thing lacking here – a comprehensive manual pitched at dummies like me. I suspect a lot of the wisdom about using this device is waiting to be discovered in the email discussion group. I am slowly working my way backwards over the last almost 2 years I have been subscribed to the list and so far, I’m up to conversation #214 of 265 conversations, that is I have 214 to go! It would be great if you search the conversations directly from the group page… must see why I can’t do that…

One of the most important documents for anyone who wants to know what makes a magnetic loop tick is Leigh Turner VK5KLT’s “An Overview of the Underestimated Magnetic Loop HF Antenna” which can be found on his club website.

Midway through building my version of Loftur Jónasson – TF3LJ / VE2LJX‘s Automatic Loop Controller, I came across Leigh Turner’s impassioned plea to consider this noise bridge antenna tuning design mentioned on page 32 of the “Overview”. As a concluding note VK5KLT states that he considers “The perceived need for a slick and salubrious auto-controller for properly tuning an MLA is oftentimes overrated and exaggerated, IMHO”.

He argued that elaborate microcontroller aided automatic loop tuning circuits are unnecessary and people should consider using this more covert and considerate approach. I think the bridge could be an excellent idea and a simpler way of staying in tune as you change frequency for all sorts of antennas. For a magnetic loop, it still requires a way to remotely adjust the tuning capacitor.

“The circuit goes inline between the rig and the antenna and sends a gated broadband noise signal to the antenna using a directional coupler and a noise bridge. You just listen on the desired operating frequency and watch your RX S-meter for a sharp dip whilst adjusting the loop tuning capacitor.

You simply remotely tune the loop with the aid of the receiver S-meter while you are on the wanted frequency without keying up and TX power output. This makes tuning a breeze without having to move off frequency and have the TX put out any RF power.”

VK5KLT mentioned the MFJ-212 Matchmaker that uses this same approach (and which is still on the MFJ catalog at US$99.95) and also referenced ZL3KB’s April 2001 RadCom article (pp17-21) as an easy and more economical way to replicate the same functionality.

“The distinguishing merit of the novel gated coupler/noise bridge loop tuning method is it’s completely passive and covert in operation; you don’t transmit any TX power whatsoever to attain an optimal loop tune setting. The technique makes for fast, QRM free, safe and easy QSY shifts and netting a frequency.” Leigh Turner adds that it’s even simpler if you use a pan adaptor or a modern SDR receiver as you can see the sharp null on the screen of the band scope display.

Kelvin Barnsdale ZL3KB’s RadCom article describes building and using the wide band noise bridge as a silent antenna tuning indicator. These four pages include circuit, PCB design and layout and full details of BOM and balun/transformer construction.

RadCom-200104-pp17-21

On 14 May 2001 ZL3KB published a 4-page follow-up pdf document “Instructions for Antenna tuning Noise Bridge” with info supplementary to the RadCom article about construction and operation. This article has an updated circuit and parts layout and refers to an issue B of the PCB. The new board includes places for the LED and dropping resistor R14, and pads for the two 100Ω load resistors and the two diodes D2 & D3.

This is the updated circuit with some updated values.

This indicates parts placement with the updated PCB.

Here is the foil side of the updated PCB

I contacted Kelvin Barnsdale and was lucky enough to obtain the PCB above.