MFJ989C

The MFJ989C Versatuner-V

The MFJ989C 'Versatuner V' T-match unit is nominally rated for 1.5KW PEP throughput when used with antennas of ³l/2 in length. It offers an extremely wide matching range and may be used with electrically short antennas if suitably de-rated. Chassis dimensions are 377 ´ 275 ´ 102mm (excluding cover).

The two air-capacitors are variable from 24-240pF, and are rated 6KV RMS. The unused turns of the air-cored roller inductor are shorted to ground, and an additional shorting contact is applied when less than half of the coil is in use (MFJ "Self-Resonance Killer™"). The output-side balun consists of 12 turns of silver-plated PTFE covered wire wound on two stacked Amidon FT240-K cores. A 50W dummy load resistor (next to the balun) is provided so that a linear-amplifier p-tank can be set-up prior to connection to the antenna matching network.

Modifications:
An L network is more efficient than the T-network but has a reduced matching range for given coil and capacitor values. A T-match antenna tuner can be modified to allow its use in one of the two possible L network (C-series, L-parallel) configurations by the addition of capacitor shorting switches [see refs 2 and 3 below]. This type of modification is particularly easy to implement in cases where the tuning capacitors allow 360° of rotation, but only 180° is required, because the shorting switches can be arranged to operate when the capacitor adjustment knobs are rotated to an unused part of the dial. This article relates to the MFJ989C, but the information given should prove useful to those interested in modifying other models.

If the subject of capacitor shorting switches is raised in the company of other radio operators, someone will inevitably offer the helpful advice that: "all you have to do is bend the corner of one of the plates". This however, is exactly how not to do it, as a simple analysis of the problem will show: Firstly, damaging the capacitor plates will encourage field-emission breakdown; and secondly, two pieces of aluminium scraping together is not generally recognised as the basis for a high-current switching system. The MFJ989C is rated for 1.5KW PEP throughput [see note 1], and provides the transmitter with a 50W load. Hence, using Joule's law (P=|I|²R), we may determine that the transmitter-side shorting contact should be rated for at least |I|=Ö(P/R)=5.5A; and even if we accept that the tuner must be derated when used with electrically short antennas, we should still rate the antenna-side shorting contact for a minimum load impedance of about 10W, i.e., |I|=Ö(1500/10)=12.2A. Thus, presuming that the modification is to be carried out using salvaged parts, we should be looking to recycle old switchgear with high-conductivity contacts rated for at least 20A.

The photograph below gives an overview of the modified MFJ989C.

Shorting switches have been fitted to the rear ends of the capacitors, the object of the exercise being to short directly across the capacitor connections. It would have been easier to fit the switch at the front of the capacitor in each case, but this would have resulted in a current loop of just over twice the length of the capacitor, with capacitance across it, i.e., a parallel resonant circuit in series with the signal path. The fronts of the capacitors have instead been fitted with end-stops, which restrict rotaton to about 240°. The anti-clockwise end-stop gives a positive registration for the point at which the shorting switch is closed (about 55° past maximum capacitance); and the clockwise end-stop restricts rotation to about 5° past minimum capacitance, thereby preventing the switch rotor from interfering with the connection to the capacitor rotor. The provision of end-stops also allows an additional capacitor rotor connection to be made by means of a flexible copper braid (a direct rotor connection, bypassing the existing wiping contact, generally reduces the ESR of an air-variable capacitor by about a factor of two). The end-stop arrangement is shown below left. Lateral disturbance of the long screw on meeting the stop pillars gave it a tendency to become loose, and so a lock-nut was added and a single drop of shellac varnish (button polish) was applied to the screw threads prior to final adjustment. The right-most pillar on the antenna-side capacitor had to be shortened prior to final assembly, to avoid interference with the power meter, and the collar was moved back accordingly (see overview photograph above).

End stops

Rear of unmodified capacitor

The rear of an unmodified capacitor is shown above right. Notice that a copper leaf-spring is fitted to apply side-thrust to the rotor bearing. The purpose of this spring is not immediately obvious, especially since no electrical connection is made to it; but since the rotor connection is made via the rear bearing (and its grease), it may be that that the sideways force reduces the capacitor ESR and prevents overheating of the bearing. The rotor connection system is, in any case, non-optimal; and for the modification, the leaf spring is removed to make way for the switch mechanism, and the bearing contact is bypassed by a copper braid attached directly to the rotor.

Two views of a capacitor after modification are shown below. The switch rotor is made from a strip of soft copper about 1.2mm thick, and the stator is a knife-switch contact salvaged from an old transmitter band-changer turret (No historic radio artefacts were harmed during the execution of this work. The transmitter in question was bought pre-ruined at a radio rally). Similar contacts can be salvaged from old circuit-breakers and heavy-duty switchgear.

Max. capacitance setting.

Shorted. (Contact shown inset).

In order to fit the switch rotor; the original 1/4" UNC stiff-nut (nut with nylon insert) at the back of the capacitor and the slip-collar for the side-thrust spring are removed, the front and rear bearing thrust surfaces are coated with molybdenum-disulphide grease, and a 1/4" UNC thin-nut is threaded onto the shaft to restore the rotor-blades of the capacitor to their mid-position. The switch rotor is then locked between the thin-nut and the stiff-nut; the whole assembly being adjusted to gap the capacitor correctly and set the switch rotor to engage with its stator contact when the capacitor is at the anti-clockwise end-stop. The two nuts must be locked together reasonably tightly; and to avoid stressing the end-stops, a thin spanner (wrench) should be used to hold the thin-nut while tightening the outer nut. Since the capacitor main-shaft is aluminium, there is a danger of thread-stripping if excessive force is used. The switch stator contact is mounted on a small aluminium bracket, which is attached to the capacitor back-plate by means of nylon M3 screws and nuts. Electrical connection to the capacitor-rotor is effected by fitting a solder-tag on to the switch-rotor; a Neoprene (Hellerman) sleeve being fitted over the point where the braid is soldered to the tag, in order to provide strain relief. Final assembly is shown below, the original electrical connection to the back bearing being re-made, and the braid being soldered to the connecting wire (joint covered by Neoprene sleeve) in such a way as to allow a short loop which can coil and uncoil as the capacitor is rotated.

Capacitors after re-installation.

Notes and References:
[1] The nominal power ratings of antenna tuners of North American origin are generally advertised to be twice the actual PEP rating. The idea is that the instantaneous DC input of a linear amplifier is about twice the PEP output (i.e., class AB amplifiers are about 50% efficient) and so the tuner power rating is the maximum DC input of the associated amplifier. This is a thoroughly silly system, and forces people to make qualfying statements such as: "The tuner is rated for 3KW DC input" (manual tuners do not have a DC input, except perhaps for the panel lamps). We should note, of course, that a tuner doesn't have a meaningful power rating on its own, it depends entirely on the antenna system.

[2] "Understanding the T-tuner (C-L-C) Transmatch" William E Sabin, W0IYH, QEX, Dec. 1997, p13-21.
Suggests that shorting out C1 and C2 will improve efficiency in some situations.

[3] "Save Your Tuner for Two Pence", Tony Preedy, G3LNP, Rad Com, May 2000, p20-25.
The "two pence" in question are copper cams used to make shorting switches for the capacitors in a commercial T-match tuner. Article contains much useful information on ATU design and coil losses, with graphs of Q vs frequency for various roller inductors [unfortunately, working out which graph corresponds to which coil is left as a puzzle for the reader, but the solution appears to be a=1, b=2, e=5. If this is correct, graph e is for the inductor from an MFJ 989C tuner, and shows a rather poor Q (~100 at 4MHz, ~40 at 29MHz)]. Author demonstrates that lowest losses occur in the T-match when one capacitor is set to its maximum value, and consequently gives correct procedure for adjustment. Goes on to describe shorting-switch modification, noting that in addition to reducing losses, it permits higher power on higher frequencies, and extends the upper frequency tuning limit for a given range of load impedances. In practice, notes that there is sometimes insufficient capacitance (in the commercial tuners modified) to effect a match with the L-configuration at lower frequencies, and the intermediate transformation (T-network) is required.

MFJ Enterprises website .