[1000mp] S-meter and S/N [was Real S9]
Tod-ID
tod at k0to.us
Mon Mar 26 16:08:22 EST 2007
In the absence of signal generators and the like, one can be reasonably
certain that the signal to noise is as good as it is going to get if when
you attach the antenna the RX noise increases. The major value of the
S-meter is reassurance that something is being received and when pressed by
the person you are talking to you have a basis for saying something is
different when they ask. As a measurement tool it is simply not accurate
other than to give binary data {louder/weaker}.
If you inspect
http://www.k0to.us/HAM/FT1000MP/S-Meter/S-meter%20Measurements.htm
on my web site you can see a plot of the S-meter changes as a function of
signal input for the range S0 to S9. The curves for IF gains of 10,11 and 12
are essentially the same with the end points displaced to conform to the IF
gain changes. In my FT1000MP IF gain = 11 is the closest approximation to
the "magic" S9 value of -73 dBm [ 50 uV]. The FT1000MP S-meter [mine at
least] is a lot closer to 3 dB per S unit than 6 dB per S unit.
Although not included in the plots, the S-meter curves for IF gain = 13 and
14 have the same shape. As an experiment I set the IF gain for 14. At that
setting the receiver audio noise still increases when I attach an antenna,
but I can hear some sound with no antenna attached. The antenna is
determining how 'quiet' the receiver is and I think there is little point in
reducing IF gain to simply make it quiet when no antenna is attached. There
may be merit in reducing the IF gain so that stages are not overdriven by
loud signals.
For those without good measurement instruments or short on time, probably
simply adjusting the IF gain so that the increase in noise with the antenna
connected is only very slightly more than when it is disconnected will be
sufficient. That should keep the setting from being too high and while
simultaneously assuring that the antenna noise is the limiting thing for
S/N.
Tod, K0TO
> -----Original Message-----
> From: Mike Schatzberg [mailto:cherokeehillfarm at earthlink.net]
> Sent: Monday, March 26, 2007 8:35 AM
> To: All about Yaesu 1000mp
> Subject: Re: [1000mp] Real S9
>
> Hello Chuck:
>
> What we were speaking of here is resetting the IF gain after
> the Inrad Roofing filter mod, not the IF Amp mod.
>
> Improved signal to noise ratio is achieved here due to the
> narrower first IF filter and the few dB of net gain resulting
> from the surplus gain remaining after the filter insertion
> losses are eliminated via the module amplifier.
> This is an additional advantage of the roofing filter mod,
> but not the primary objective. The original design of the
> module amplifier was to be able to compensate for the losses
> of the narrower crystal filter. Over time, the manufacturing
> process for the filter apparently became more refined, and
> the filter insertion losses deminished. I wrote to Inrad
> myself suggesting that their instructions for installation
> might be modified to take advantage of this development.
> They did indeed change their instructions after it became
> evident that newer filters had lower losses.
>
> This change in the filters does more or less let the roofing
> filter mod parallel the performance of their earlier IF Amp
> mod. Many people have swapped out the older mod for the
> newer roofing filter.
>
> I have worked here on the bench with the new filter
> installed, examining signal to noise ratio, noise floor with
> MDS, and the S meter calibration.
> While each radio is different due to component selection and
> overall set up, my personal conclusion has been that the
> signal to noise ratio is optimized generally within a
> decrease of 1 to 2 integers of the original factory settings,
> and that the S meter readings will not be significantly
> compromized for linearity within this range.
>
> While the S meter readings at S9 (50 uV input) may be correct
> for small reductions in IF gain settings, my measurements
> indicate that too low a setting will result in loss of
> linearity below S9. You can then perform the S meter
> calibration which is a two step process; first setting VR8006
> for a one dot deflection with a +11dBu input signal, and then
> setting VR8007 for
> S9 + 60 dB with a +100 dBu input signal. Yaesu does not
> calibrate the meter at S9, please see page 3-3 of the
> technical suppliment for this information.
> Please note that the IPO lamp is off for these adjustments,
> meaning the preamp is operating.
>
> Regarding the accuracy of other receivers S meters, I would
> say that there are two issues at work here. One source of
> error always seems to be whether the S meter is calibrated
> with the preamp on or off, this makes for a 2 S unit change
> alone generally. Secondly, some manufacturers really haven't
> adopted the Collins standard of S9 equal to 50 uV input, with
> one S unit being worth about 6 dB. Their S meters are not
> only incongruent with this standard, but they are also non linear.
>
> You just might want to actually measure what you have created
> once you have made a larger change to the IF gain to optimize
> signal to noise. I will not even venture to guess what might
> also change in terms of dynamic range and BDR with larger
> excursions from the original factory settings.
>
> Experience still says that downward adjustment of the gain in
> menu 9-1 by one to two units will produce acceptible results
> across the board. You will also see that the two receivers
> remain in approximate correlation at S9.
>
> 73 and Happy DXing,
>
> Mike
> W2AJI
>
>
>
>
>
>
>
> Chuck wrote:
>
> Hi, Mike,
>
> Oh, I have already installed the mod, and after that, the
> roofing filter.
> Both times I followed the Inrad suggestions, lowering the
> 9-1 settings by
> 1-2 numbers (in my case, the roofing filter didn't change
> that amount of compensation); and both times the resulting
> s-meter readings agreed generally with the sub-rcvr. However,
> I also used my ears and several weak sigs to establish what I
> sensed as the best weak-signal s/n ratio. This also agreed
> (within 1 number) with the Inrad suggestions. The S-meter
> STILL seems quite stingy (both main and sub) compared with
> other radios. My point is that, once I had done this, I
> satisfied myself that further messing with the IF gain (which
> would not improve the S/N ratio, but only make all sigs
> louder, including the noise) would be a bad idea, and that I
> should live with a stingy meter. Either Yaesu's calibration
> is in error, or we all have been misled by other radios'
> overly generous meters. The 50 uV generator is a great way to
> give us all a consistent S-9 standard. But Pete's suggestion
> to increase the IF gain until the s-meter reads S-9 at 50 uV
> just reverses all the good we got out of the IF mod.
> Remember, the reason for the mod was to balance the gain
> distribution by reducing the 70 MHz IF and using the Inrad
> amp. to compensate upstream (I think I remember that
> correctly). I think it's erroneous to assess the RCVR's
> performance by looking at the S-meter. We all installed the
> mod to make the radio quieter (without losing sensitivity).
> Just leave the IF alone, at it's best S/N gain and
> RECALIBRATE ONLY THE METER using the 50uv generator as a
> standard! If I can find the meter calibration adjustment in
> the radio, I'll share it.
>
> Yes, if you're referring to the BHI noise reduction system,
> I've already put that on my wish list!
>
> Chuck
>
>
> ----- Original Message -----
> From: "Pete Smith" <pete.n4zr at gmail.com>
> To: "All about Yaesu 1000mp" <1000mp at mailman.qth.net>
> Sent: Monday, March 26, 2007 7:19 AM
> Subject: Re: [1000mp] Real S9
>
>
> > Hi Chuck - well, my logic was that the only thing I had
> changed was the IF
> > gain setting in menu 9-1, and the radio not only sounded a
> little dead,
> the
> > main RX definitely had less sensitivity than the sub-RX.
> Given that, I
> > felt that getting the S-meter back to reading more or less
> the same on
> both
> > with a 50 uV signal would indicate that I had restored the
> IF gain to
> > something like spec. The XG-1 also has a 1 uV output, and
> Elecraft also
> > makes a highly accurate step attenuator kit if you want to
> go farther.
> >
> > 73, Pete N4ZR
> >
> > At 10:44 PM 3/25/2007, you wrote:
> > >Gee, Pete, I'm not trying to start a war here, but it
> seems to me that
> the
> > >gain distribution (affected by the I/F gain setting)
> should be set for
> > >max. S/N, not diddled with as a means to correct an S-meter
> > >mal-calibration. On the other hand, suggestions have been made that
> (after
> > >any Inrad mods, etc.) the I/F gain be adjusted until the main RCVR
> S-meter
> > >indicates the same as the Sub-RCVR reading. That would
> seem to be a way
> of
> > >returning the overall I/F gain to its factory setting.
> This seems fine as
> > >long as one assumes that the sub-RCVR setting is
> consistent with the best
> > >main RCVR S/N ratio: a bad assumption, IMHO.
> > >
> > >Why wouldn't it make more sense to (somehow) establish the
> best S/N ratio
> > >by adjusting the I/F gain and THEN use the 50 uV standard
> to set the
> > >S-meter sensitivity? I've played a little with the I/F
> menu setting using
> > >barely readable sigs to reach a (subjective) optimal
> setting. That turned
> > >out to be fairly consistent with Inrad's suggested
> compensation when
> > >installing their I/F mod kit. My S-meter seems stingy,
> too, but it seems
> a
> > >bad trade to give up weak signal performance (and lose the "hiss
> > >reduction" we paid for in the Inrad kit) just to get the
> S-meter right
> > >when there's probably a dedicated S-meter sensitivity adjustment
> somewhere.
> > >
> > >What am I missing here?
> > >
> > >73,
> > >Chuck, N4NM
> > >(Who's wishing he had one of those VG-1 kits, too :>)
> > >
> > >
> > >
> > >----- Original Message ----- From: "Pete Smith"
> <pete.n4zr at gmail.com>
> > >To: "All about Yaesu 1000mp" <1000mp at mailman.qth.net>
> > >Sent: Sunday, March 25, 2007 2:58 PM
> > >Subject: [1000mp] Real S9
> > >
> > >
> > >>I just finished building the Elecraft XG-1 signal generator.
> > >>
> > >>So why am I talking about it here? Because It puts out a
> calibrated 50
> > >>uv on 7040. After I built it, the first radio I hooked
> it to was my
> > >>TS-930. S9 right on the button. Then I connected it to
> my Mark 5. S6!
> > >>
> > >>I had thought for a long time that the Mk 5 sounded
> "dead" compared with
> > >>the 930, but put it down to DSP, etc.
> > >>Not long after I got the MK5, I put in the Inrad IF mod,
> and then a
> > >>couple of years later I substituted the roofing filter.
> The radio also
> > >>has cascaded 400 Hz Inrad CW filters. In the course of
> doing all that,
> I
> > >>guess I had turned the IF gain down too far - anyway,
> using the XG-1, I
> > >>was able to reset it to ~S9 at 50 uV - certainly sounds
> like the band is
> > >>more open!
> > >>
> > >>I expect the VG-1 is going to prove very useful, anytime
> I want to check
> > >>the performance of my antenna switching or bandpass
> filtering - and for
> > >>$40, I can't imagine a cheaper way to do that.
> > >>
> > >>______________________________________________________________
> > >>1000mp mailing list
> > >>Home: http://mailman.qth.net/mailman/listinfo/1000mp
> > >>Help: http://mailman.qth.net/mmfaq.html
> > >>Post: mailto:1000mp at mailman.qth.net
> > >>
> > >>
> > >>
> > >>--
> > >>No virus found in this incoming message.
> > >>Checked by AVG Free Edition.
> > >>Version: 7.5.446 / Virus Database: 268.18.18/733 - Release Date:
> > >>3/25/2007 11:07 AM
> > >
> > >______________________________________________________________
> > >1000mp mailing list
> > >Home: http://mailman.qth.net/mailman/listinfo/1000mp
> > >Help: http://mailman.qth.net/mmfaq.html
> > >Post: mailto:1000mp at mailman.qth.net
> >
> > ______________________________________________________________
> > 1000mp mailing list
> > Home: http://mailman.qth.net/mailman/listinfo/1000mp
> > Help: http://mailman.qth.net/mmfaq.html
> > Post: mailto:1000mp at mailman.qth.net
>
> ______________________________________________________________
> 1000mp mailing list
> Home: http://mailman.qth.net/mailman/listinfo/1000mp
> Help: http://mailman.qth.net/mmfaq.html
> Post: mailto:1000mp at mailman.qth.net
>
>
More information about the 1000mp
mailing list