Wednesday 30 May 2018

In The Loop! - First Impressions of the MFJ-1788

Was feeling pretty chuffed after repairing the MFJ-1788 'Super Loop' and couldn't wait to try it out! So for a couple of evenings of experimenting I put the loop in the garden on a 5ft pole held up by a heavy drive-on stand with 20m (65ft) of RG58 running into the shack.

Temporary test setup
I chose the easy option of using FT8 to do some testing, selecting the 30m FT8 frequency initially. I tuned the loop and was met with a cacophony of signals, far louder than my usual OCFD would receive. Working with around 30 Watts had a few contacts in a brief 30 minute session, including a nice one in Greece with SV1IW.

PSKReporter showing where I was spotted
What was striking was the lack of noise and just how tight the tuning was, indeed I had to tweak the tuning a couple of times during tests, a slight adjustment either way and the signals just disappeared. I checked out 40m and 20m as well with similar good results.

I had a few more sessions and a few days later I tried it out to receive the Shortwave Radiogram broadcast from Bulgaria on 9400kHz, this time as it was a broadcast band had to use my ear to do the tuning, adjusting till I heard a rise in 'noise' and signal.
I have now got the loop up on a rotator and mounted slightly higher up with a shorter length of RG213 (not on the video) it is still quite low and unfortunately is slightly shielded to the south by the neighbours metal roofed building,



I am very happy with the loop. Transmission wise it unsurprisingly doesn't seem a huge improvement over the OCFD on its resonant bands, it scores over the OCFD is on its 'non-resonant' bands such as 30m and 17m. But the massive improvement is in receiving, signals are stronger and noise is much lower, picking up some more distance signals even given the poor conditions.
The antenna cannot be said to be a pretty thing to have in the garden! The tuning is very particular, in the video I show the 'auto tuning' isn't ideal. It requires the radio to be putting out a signal into a mismatched load for what could be nearly a minute. Not good for the radio and is a source of QRM during this time, the usual technique of tuning slightly off a QSO frequency is more problematic due to the sharpness of the resonance. You can still tune off frequency and then tweak with the slow tune buttons to bring it in. I've noticed that on some of the higher bands it occasionally doesn't auto-tune because the 'dip' seems very short/sharp and the controller doesn't react in time and overshoots especially using low power settings.



I also have had issues trying to operate the radio remotely, I have tuned it up on an FT8 frequency in the morning and then later in the day logged in to try to make a few QSOs during a coffee-break to see the loop has drifted out of resonance. This can only be down to the loop getting warm in the summer sun.

I am still evaluating the antenna but am looking at making a better controller, over on AmateurRadio.com where this blog is syndicated, I have had a number of kind comments including one from Elwood Downey, WB0OEW who pointed me to his published design of a controller, using a similar method to what I was toying with. Thanks Elwood.

73 for now, more updates soon.

The antenna farm

Thursday 17 May 2018

In The Loop! - Repairing a MFJ-1788

Over the last few years I've seen various musings about magnetic loop antennas and their supposed efficiency and performance. I had even toyed with the idea of building one considering the eye-watering prices commercial ones are sold for! Sadly that idea had been added to the big pile of 'to-do' projects.

Earlier this year I spotted a posting on social media by a member of my club who had brought a MFJ-1788 second hand who was having issues with it not working. I had offered some advice on its repair and glibly offered to take it off his hands should he want to dispose of it. A few months later I got a message asking if I was still interested? I certainly was at the price he wanted.

I collect it at the club meeting a fortnight ago still with a 5ft pole still attached as a bonus! Thankfully it fitted in the car (just).

Loop as purchased, with pvc tape covering loop


Control Box

I downloaded the manual and schematic from the MFJ website and I saw there were conspicuous warnings about using an 'isolated' power supply both in the manual and on the back of the controller, with ominous warnings of damage if you didn't. Most shack supplies have the negative/black pin connected to the chassis/ground.

The controller had come to me supplied with a standard fused lead, you know the ones that come to connect your ATU/SWR meter lamp to the shack supply? Mmmm.. my spider sense was tingling!

I pulled out a small double-insulated 12V plug-in supply from my collection (you can spot them as they often have a plastic earth pin) and with nothing else connected powered up the controller. The meter lamp came on and some of the LEDS briefly flickered and heard a few clicks from the internal buzzer then nothing. Pressing buttons did nothing and then I caught the unmistakable scent of burning electronics!

Opening the box up on the bench quickly spotted the source, the regular had well and truly smoked, but since nothing had been connected the short must have been in the controller itself (and it was not the dodgy wires that look like they had been victims to a wayward soldering iron)


Inspecting the rest of the board it was clear that it had been repaired (badly) once before, the two main transistors/FETs (Q1/Q5) used to control the motor clearly showed signs of being replaced (misaligned and with tell-tale scorching from an hot air gun)


a SMD diode had been swapped and one of the 'fine tune' switches had been replaced, its removal had obviously been problematic taking with it some of the through hole copper and adjacent tracks which had been lifted/damaged.

 


Before opening up the box I had expected to see 'through hole' components and DIL logic ICs not for it to be all surface mounted. Undaunted I went through the parts list and decided to get replacements for all the semiconductor parts not knowing at this stage what I would have to replace. Getting two of each device in some cases five or ten of each due to minimum order quantity, the staggering cost was £8 ($11) including postage!

While waiting for the parts to arrive I decided to check out the actual antenna head. I removed the plastic covering and attempted to extract it from the pole..


First issue I had was the mounting. One bolt had seized, the nut inserts are held in quite soft aluminium so not surprisingly the insert came out out the mounting when trying to remove it. Using grips to hold the insert the bolt still refused to turn and in the end it sheared off with very little force! Obviously quality fittings these! 



The variable capacitor was suffering a common problem due to poor quality control, the fins had gone out of alignment. Each fin is a separate assembly held on a shaft with a nut to compress them together. It is a simple fix, just realigned and tightened the bolt. The limit switches and motor connection were okay (another common problem) Applying a voltage to the coax connector the motor turned fine, reversing the voltage reversed the direction of turn, the vanes going fully in and fully out before the limit switches activated.



The components arrived prompty and so I got the hot air gun out and set about repairing the board.


I started off replacing the regulator (and the damaged power track) and Q1/Q5 and we had the flickering of life. Pressing the AUTO TUNE button the LED lit and there was 9V across the antenna connector, pressing the other AUTO TUNE button had -9V across. This is why you need an isolated power supply since it reverses the voltage to control direction. The radio doesn't see any of this voltage due to the bias-t arrangement. However if the supply shared the shack ground you would short out the supply, but the current would go via all those delicate electronics! Which is what I think had happened.

It soon became clear that most devices had suffered damage and so I ended up replacing the regulator 78L12, Q1 (SN7002 FET), Q2 (MMBT3906), Q5 (MMBT3906) all the logic ICs CMOS 4001, 4011 and 4066 and the LM324 Quad op-amp chip. I also gave the board a good clean since was covered in grime and flux residue.

Once I was happy the controller was reassembled and connected to my FT-857D on minimum power and the loop antenna propped up against the side of the shack. Success! I was able to successful tune it as per the instructions on 40m and 10m, the two extremes of operation and could here the telltale rise in receive noise as it became resonant.

I have yet to put the loop up in situ for a proper evaluation but have refurbished an old TV rotator to mount it on. I have fitted some nice new quality bolts. The black insulating tape round the loop has been removed, preferring the silver look myself.

Waiting to be put on the pole

Shiny once again

Assuming it does performs and will find out this weekend, I have paid a total of £65 ($90) (£50 for the initial purchase the rest for replacement parts) which is an absolute bargain as to buy new it is expensive.. very expensive!

One of the main supplier in the UK, have it on sale for £699 ($940) but this includes their engineers inspecting and rebuilding each one before shipping! (Can only be due to poor build quality control and warranty claims) the RRP seems to be around £570 ($770)

It does amaze me how much this units costs new. While the loop construction is generally good, the mounting fittings appear to be poor and the issues with the capacitor and quality control are widely reported.

The components in the controller are not expensive, the switches, meter, box are the usual MFJ fare and the design is quite old (the PCB has copyright 1998 on the silk screen) the auto-tune process requires the radio to be putting RF into a mismatched load for up to a minute and even with low power and SWR protection this isn't good for the radio PA.

Perhaps most surprising is given the risk of damage due to incorrect use is why firstly the loop isn't supplied with a power supply anyway as they only cost a pittance and secondary why their isn't any protection built in? I intend fitting a simple 100mA fuse to offer some protection should a problem occurs.  

If the loop performs it may be another project to build a better controller, there are a few designs out there on the internet using Arduino and DDS devices to create auto-tuners. Has that pile of potential projects just grows taller?

73 for now.. and promise to post a bit more regularly