Retune radio to ham bandConvert to duplex operationBring required I/O out to rear connector

Transmitter:

J103-	Tune:	For:
1	T101	peak
1	T102	dip
1	T103	peak
3	T104	peak
3	T105	dip
4	T106	peak
4	T107	dip
7	T108	peak
7	C155	dip
6	C157	peak
6	C167	dip
2	C171, C175	peak
1	T101	peak
3	T102, T103, T104	peak
4	T105, T106	peak
7	T107, T108	peak
6	C155, C157	peak
2	C167, C171, C175	peak

At this point the manual says to basically repeat the steps above shaded in green. I find this unnecessary, but if you really want every stage peaked to the last millivolt, go ahead.

Receiver:

J601-	Tune:	For:
3	C406	peak
	C411, C416	same pos. at C046
	C306, C307, C308	full CCW
4	C411, C416, C411	peak (twice on C411)
4	C406, C411, C416	peak
4	C306	any change
7	C307, C306	peak twice
7	C308	dip
7	C306	peak
1	C301 thru C305	peak or best quieting
1	A303-C2	peak or best quieting
(P903-1)	L603	peak
(P903-1)	R603	1 V RMS

Steps shaded in green are performed with a 50 ohm signal generator connected to the radio's antenna input. If you aren't sure about the output impedance of your signal source, just use a 6 dB or greater 50 ohm pad on the output. Due to the tremendous selectivity of the helical resonators, you will probably need a lot of signal at first (millivolts). If you have trouble detecting a signal to tune up on, first make sure your crystal isn't way off frequency by moving the sig. gen. +/- 30 kHz or so off of your center frequency. Then try presetting C301-5 to within about 2 turns of bottom, as this is near the resonance point for the 440-450 MHz range.  If that doesn't work you can try injecting signal through the little holes in the helical assembly provided for just this purpose (the manual recommends using them but I find them unnecessary). Steps shaded in blue are done with a full-quieting 1 kHz modulated signal at 3 kHz deviation. If you don't have access to an FM sig. gen. you can adjust L603 by transmitting a DTMF tone with another radio. When you're finished tuning you should get 0.35 microvolts or better for 12 dB SINAD, or 0.5 microvolts for 20 dB quieting. These instructions assume your radio does not have the UHS bipolar preamp. If it does have it I recommend you bypass it & get yourself a good stable GaAsFET or MOSFET amplifier from Angle Linear unless you're REALLY trying to pinch pennies. The UHS preamp does work, but it's noise figure is going to be around 4.5 dB plus I've noticed problems with RF coupling to the outside of the coax leading into the preamp. This can lead to internal desense problems. Also bipolar preamps are known for having a lower 1 dB compression point than the FET amps, meaning you'll be more susceptable to IMD.
 Duplex conversion
There are four steps to perform to convert the MVP to duplex operation.  Duplex operation means the radio will transmit and receive simultaneously. Unlike ham transceivers, the MVP is able to do this all on the same band; this is part of what makes it such a nice radio for repeater use.  The four steps are as follows.
 

Strap the RX OSC line to the 10 V bus to keep the receiver on while the transmitter is keyed. Solder a wire on the bottom of the system audio squelch board as seen in the following pictures:

Cut the RX MUTE line to keep the receiver squelch working while the radio is transmitting.  The cut trace is shown in the following picture:

A little work is required to bring the CTCSS decode signal out of the MVP's "Channel Guard" (CG) board, as it normally doesn't provide the radio with a discrete decode logic signal. Instead, it pull the "RX MUTE" line low whenever no CTCSS tone is being decoded. If your repeater controller doesn't have an input for CTCSS decode, you can leave the RX MUTE connected to the receiver & not bother bringing any logic out. But since 99% of the controllers currently available do have a CTCSS decode input, you will probably want to cut the RX MUTE line here as well (see the already-cut wire in the picture). The wire just below & to the left of the RX MUTE wire also needs to be cut. This is the PTT line, which needs to be disconnected from the CG board in order to keep the decoder in decode mode while the transmitter is on. Also solder a shorting strap across CR3; this will effectively turn the "CG DISABLE" line into a "CG DECODE" output, going to 0 V when decoding & 10 V when inactive. In my MVP, the CG DISABLE was already wired to pin 8 of the system connector. Now how easy can that be!

Now that the radio can receive and transmit at the same time, we need to bring a second RF connection out for either the receiver or transmitter. The way the MVP is designed, it turns out that it's easier to bring the receiver input out on the added coaxial connection. There are several possible approaches; mine is shown in the pictures. I simply drilled a hole in the back of the radio & ran a piece of RG-142 from the back over to the RCA jack on the helical resonator assembly (or UHS preamp, if you're using one). I actually prefer to use RG-223 for this cable because it's more flexible. Be sure to keep the coax away from the center pin of the SO-239 connector to prevent excessive TX RF from coupling to the shield of the new RX feedline.

The following table shows the pinouts for this connector as defined by G.E., and my revised pinouts to accommodate the added I/O lines:

J1-	G.E. definition	Revised definition
1	+12 V (16 ga. red)	+12 V (16 ga. red AND 22 ga. red)
2	gnd. (thin solid wire to chassis)	gnd. (no change)
3	spkr. hi	spkr. hi (no change)
4	spkr. lo	CG hi
5	+12 V (22 ga. red)	PTT
6	gnd. (thin solid wire to chassis)	vol./sq. hi (discriminator)
7	spare; NC	mic. hi
8	CG disable (green)	RUS (squelch logic)
9	spare; NC	NC (no change)
10	spare; NC	CTCSS decode

Overheating of the transmitter power amplifier
The MVP's transmitter is not designed for continuous duty, so it will overheat unless it is run at reduced power or supplemental cooling is provided. I use a muffin fan positioned to blow directly on the back area of the 20 watt radio or heatsink of the 35 watt model. You can also simply add a heatsink to the 20 watt radio & not use a fan, but I recommend the fan in either case. If you're happy running the radio at 10 watts or less, then you can probably get away with not adding a fan or heatsink.TX/LO intermodulation problem
A strange problem was discovered a few years ago by someone on the repeater owners' remailer when he tried to duplex a G.E. radio that was set up for 448.625 RX/443.625 TX. This problem appears to affect a range of frequencies, as another ham had trouble duplexing his G.E. radio at 448.500 RX/443.500 TX. To this date the specific cause of this problem has yet to be found; those who have been affected by this problem have resolved it by using two separate radios for TX & RX.

Another problem was discovered a few years ago when one MVP repeater received an unusually high amount of IMD that included the output TX as one of the mix products. It was discovered that the "howling" would occur whenever a strong signal (~-60 to -70 dBm) at one of two specific frequencies was input to the RX while the TX was enabled. The RX's "soft spots" depended on the TX & RX frequencies in the following manner:F_SS = F_RX ± [(F_RXOSC X 28) - F_TX]

Where F_SS = receiver's "soft spot", the frequency of spurious reception when the TX is on (2 frequencies due to the ± term), F_RX = nominal RX frequency, F_RXOSC = RX LO oscillator base frequency (equal to (F_RX - 11.2) / 27 if using low-side LO, or (F_RX + 11.2) / 27 if using high-side LO), & F_TX = the TX frequency.

As an example, for an MVP repeater operating on 442.000 MHz RX & 447.000 MHz TX, the spurious RX responses would be at 441.756 & 442.244 MHz (rounded to the nearest kHz).

The easiest solution to this problem is to switch to high-side LO.  In the above example, this would push the "soft spots" out to almost 23 MHz away from the nominal RX frequency!  Since the root of the problem lies in intermodulation with undesired harmonics of the fundamental LO, another check is still required by replacing 28 with 26 in the "soft spot" formula.  However, this still yields spurious RX frequencies that are over 10 MHz away, so the problem is effectively cured.

Since discovering this problem, I've always specified "high-side LO" when ordering RX crystals for all my G.E. radios (I've found this problem in Mastr IIs Mastr Exec IIs as well). Before discovering the cause of the problem, a friend of mine removed his receiver & put it in another mobile case. Strangely, this had NO EFFECT on the severity of the spurious responses.  He ended up installing the receiver in an RF-tight box with EMI feedthroughs; this of course eliminated the problem completely.

Thanks to Mike Steiner KD6LVP & Bill Wood W6FXJ for their research in discovering & solving the G.E. mobile TX-LO IMD problem.

Article by Bob Dengler, NO6B Copyright October 2001.
HTML Copyright October 14, 2001 Kevin K. Custer, W3KKC All Rights Reserved.