Amateur Radio Control Shifts

| For radio control on 6 meters using PPM -- what is the amount of | frequency shift, and which direction is it (please don't just say | 'positive' or 'negative' -- I need to know which direction is 'pulse' | and which direction is 'not pulse').

and may be of some assistance, or at least there will be some pretty pictures.

Positive shift means that the frequency increases during a pulse, and negative means it decreases. I don't think the exact amount of the frequency change really matters, but I think it's normally around 1.5 KHz to 2.5 KHz.

Or, to make it really simple, for positive shift, `pulse' is about 2 KHz higher than `not pulse', and reverse it for negative shift. And modern 6m stuff is positive shift.

| Yes, I know there's no standard -- what do manufacturer's do these days?

From

--

SPECIAL NOTE CONCERNING HAM BAND TRANSMITTERS: All current transmitters use positive shift modulation. When you order for current transmitters, your invoice will list the purchased item as 805FM50V2FJ or 805FM53V2FJ. This receiver will work with all recently manufactured transmitters. If you have an ACE transmitter or an old Futaba (prior to around 1990), you will require an ACE version. Your invoice will list the 805FM50V2ACE or 805FM53V2ACE. Please be advised, it is unclear when Futaba changed from negative to positive shift modulation. Unfortunately, unless you have access to test equipment, you may not know if your old Futaba requires negative or positive shift.

... so it looks like there IS a standard now, at least on the six meter band stuff. I've heard some say that this isn't true, that brand X 6m RX didn't workt with brand Y 6m TX, but details were never really given.

That was useful, but I forgot to ask:

Is the space (no pulse) frequency the nominal frequency, or is it (more sensibly IMHO) 1/2 the shift below -- or at least _some_ amount below the nominal?

If I were designing such a rig I would have the space frequency (off, no pulse, whatever) be 1/2 of the shift _below_ the nominal frequency, and the mark frequency be 1/2 of the shift _above_. I may shade the space frequency to be a bit closer to the nominal frequency to balance out the spectrum, but I doubt that I'd stick it right onto the nominal frequency.

Thanks again.

| That was useful, but I forgot to ask: | | Is the space (no pulse) frequency the nominal frequency, or is it (more | sensibly IMHO) 1/2 the shift below -- or at least _some_ amount below | the nominal?

Do not know.

The shift for the FM (frequency shift keying) is up to +1500 and -1500 about the nominal frequency. This is true for 72 MHz or 50 MHz. This is used on 53 MHz but is not mandated on this band since the channels are

100 KHz wide.

Dan Thomps> Doug McLaren wrote:

Thanks Dan. It has to be fairly close in to meet the spectral requirements, of course -- I just didn't know what the relationship was between nominal and reality.

You got my curiosity up so I dragged in the ol' service monitor to see what my 9C was doing. All of the pix were with the freq error meter on the 1.5KHz range and the scope scale was on the +/-1.5KHz range.

The first was using a FP-TP-FM module on 08 (50.960MHz):

Modulation is +/-1500Hz. The freq error meter is showing about 1KHz low, but I am assuming it is the service monitor adjusting to the non-symetrical deviation.

The second pic is with a FT-TP-FM module on 60 (72.990MHz):

Inverted from the 6 meter.

The third pic is with the TP-FSM syntesized module set to channel 60:

Same as the crystal module except the modulation is just a tad lower.

FWIW, David - WD4JKH

That pretty much says it all. I guessed that +/-1500 was the number to shoot for -- apparently your TX (Futaba?) agrees with me.

With 2kHz shifts I'm not sure that this is particularly relevant. Most of the receivers probably have bandwidths of 10kHz or more.

Ideally the center frequency would fall in the center of the passband of the receiver, following your "1/2 the shift below" if the center frequency were truly accurately calibrated. And the receiver bandwidth would be simlar to the FSK spacing. But things have always been much looser than this.

1kc at 50MHz is 20 ppm, and 30 or 40 years ago when I did 6M remote control I'm pretty sure that most of the crystals would've truly struggled to meet this spec. Some of the transmitters used LC circuits for tuning (I am not kidding!) and receive bandwidths were as wide as 100kHz or more. But that kind of slop was going away as the tube transmitters disappeared :-).

Of course the Gonset portable sets (transmitter and regen receiver both tuned only by LC's) set truly abysmal standards for stabilities and bandwidths. Maybe I'm being too pessimistic in extrapolating their specs to today!

Tim.

Current spec is 20kHz channels, so you have to be better than that.

You are correct that I should expect significant offsets, however -- I should have been thinking in those terms.

| Current spec is 20kHz channels, so you have to be better than that.

Current spec is _10_ KHz channels. Sure, our (72 MHz) channels are 20 KHz apart, but that's because there's channels for other things (pagers, industrial R/C, wireless microphones) in between our channels.

The hazelnuts (or something) are blooming in Oregon, and I'm getting the early spring allergies.

I had my head stuck in the 6m band, which gives exclusive 20kHz channels. Yes, 72MHz RC is 10kHz.

I think I need to start tattooing this to my forearms.

Carons's Rule has the bandwidth, defined as 98-99% of the energy passed

for FSK modulation (Digital Signal Processing in Communication Systems Frerking)--

BW ~ 2 * (fd + fm)

where fd is freq deviation, and fm is highest freq of present in th modulation.

From the preceeding posts the deviation appears to be in the 3 - 6 Kh range. The modulation freq would have to do with the puls filtering/shaping. My Futaba TXAS w R127DF rcvr shows the positiv going pulses at 0.505 ms wide, and rounded, which I would eye-ball a having the highest freq component around 1-2 Khz. So the bandwidt would approximately be-- worst: 2*(6+2) = 16 best: 2*(3+1) = 8

I'm currently building an rf deck from Gordon Anderson'

pc board, this topic is of high interest. I too, was wondering if there was any standard. I suppose th proof-of-the-pudding approach would be to adjust the homebrew xmitte to match the stock Futaba TXAS transmitter as viewed on the R127D rcvr

Don, W4D

-- deh

----------------------------------------------------------------------- deh6's Profile:

this thread: