MFJ948 to S-match conversion (3)

 My last post about the MFJ948 T-match tuner conversion to S-match tuner dates from February this year. It contained a fault which was quickly seen by PA0VAJ Jan. I asked the readers if they could discover the fault but it seems no one was really interested or it was too difficult to find. Or readers might have expected I would write about it in a short time. But actually I had other things on my mind and the tuner was put in a carton box for later experimenting.

The fault

The MFJ989 to S-match conversion diagram from PA0FRI's page contains a fault. It is there since 2018 and I don't think anyone has noticed. Probabely no one else has tried to convert a T-match to S-match yet. The coil and capacitor(s) need to be on both sides of the two 8 winding sections. In this diagram the coil is mounted over the 16 winding part.

The strange thing is that the same page does show the correct diagram at the top. I didn't see the difference at first till Jan told me the wiring was incorrect. So, I changed it to correct things.

The problem

After changing the wiring testing showed a little better SWR but still most frequenties couldn't be tuned exactly to 50 Ohm. The problem is the switchable coil. PA0FRI noted this as well and so I changed the wiring again to make a quasi roller inductor. This is done by connecting a varco in series with the coil. Unfortunately the PA0FRI site contains another fault which was easy to see...

Obviously I guess readers of this blog will see this fault instantly. 

The result

Much better! I'm finally able to tune 1:1 on most bands from 3,5-30MHz. Except 15m, for that band I need to switch extra capacity over te main capacitor C2. For 80m it helps as well to get a slightly better SWR. For 60m I need to connect the antenna over de coil, it will not tune when connected over the capacitor. Tuning is pretty difficult since the components in the MFJ-948 are small. I remember it was difficult to tune when it still was a T-match tuner as well. Without a antenna analyzer like the MFJ-948 I use it is almost impossible to find the right settings.

For archive purposes this is my list of settings: 

S-match test

Over L                            Over C

Band	C1	L	C2		C1	L	C2
80*	8	K	5,5		6	L	1,5
60	3,75	K	0		np	np	np
40	2	J	12		-3,5	L	-3,5
30	1	I	-5		-3	F	-2
20	0,5	H	-3		-1	F	-1
17	2	E	-3		0	E	-2
15*	10,5	L	7		np	np	np
12	1	I	9,5		2,5	E	-2
10	9,5	H	-1		I	C	-4

np=not possible

* with 300pF C parallel over C2

Now, the S-match tuner seems to be excellent to balance your transmission line. But is it efficient? I think that depends on what components you use. The MFJ-948 components are not that great...

To determine the quality of the build and compare it with my Palstar AT1KM I should build a RF antennacurrent meter first. More about that later...

The finished project: