Hi Jules,
Don't use the 3.3V on the LO boards, I think its intended as an output from the regulator. Using the 5V input gives a better result - sounds like you have fixed it.
Mike
Are my constellations any good?
Re: Are my constellations any good?
Mike,
Hopefully so. Though it would be nice to screw the last few dB oput of the system. The suggested MER should be 26dB and I am seeing 23dB on 300ksym.
I ended up taking the 5V reg off the LO filter board and feeding it and the LO synth from the modulator board reg. Though to keep this from getting too toasty, I feed that from an 8V switching reg. Which also saves me a bit of current, important as I will only ever be operating portable.
Dave suggested that there may be an issue with the active filter, for the lower rates. This being my 3dB difference between 300 and 333 kSym. So I may have a bit of tweaking there to do still.
Now to start building PA's. (And antennas, as my usual load is resistive one on the bench.)
Jules
Hopefully so. Though it would be nice to screw the last few dB oput of the system. The suggested MER should be 26dB and I am seeing 23dB on 300ksym.
I ended up taking the 5V reg off the LO filter board and feeding it and the LO synth from the modulator board reg. Though to keep this from getting too toasty, I feed that from an 8V switching reg. Which also saves me a bit of current, important as I will only ever be operating portable.
Dave suggested that there may be an issue with the active filter, for the lower rates. This being my 3dB difference between 300 and 333 kSym. So I may have a bit of tweaking there to do still.
Now to start building PA's. (And antennas, as my usual load is resistive one on the bench.)
Jules
Re: Are my constellations any good?
Hi Jules
The "low MER" issue appears to be caused by the way the MiniTioune software indicates the MER from different tuners, and appears unlikely to affect weak signal performance. We first noticed this issue a few months ago, and I did some tests mixing the amplified output of a noise source with the attenuated output of either a Portsdown or DATV Express at 333 KS. The weak signal performance of all the transmitter/tuner combinations was identical despite the lower reported "no noise" MER for the Portsdown when viewed on the Serit tuner.
In my first test, I adjusted the Portsdown DATV signal level to give 8dB MER indicated on a Sharp tuner with the broadband noise on. With the same signal, the Serit tuner indicated 8dB MER and had the same level of Viterbi errors. I then switched the noise source off and the Sharp MER rose to 25dB, but the Serit only indicated 19dB. Signal levels were high enough not to be a factor.
I then repeated the test with a signal at the same level from DATV Express, and both MERs were 8dB with noise and 26dB without noise.
There is clearly some interaction between the tight filtering on the Portsdown TX and the measurement point (in time relative to symbol transition) of MER in the Serit tuner that is not apparent in the Sharp Tuner. I did some further tests with wider filtering on the Portsdown; the Serit indicated "no noise" MER approached that of the DATV Express as the filter was widened.
I am confident that your Portsdown would indicate 26 dB on a Sharp tuner if you are getting 23dB on a Serit.
Dave
The "low MER" issue appears to be caused by the way the MiniTioune software indicates the MER from different tuners, and appears unlikely to affect weak signal performance. We first noticed this issue a few months ago, and I did some tests mixing the amplified output of a noise source with the attenuated output of either a Portsdown or DATV Express at 333 KS. The weak signal performance of all the transmitter/tuner combinations was identical despite the lower reported "no noise" MER for the Portsdown when viewed on the Serit tuner.
In my first test, I adjusted the Portsdown DATV signal level to give 8dB MER indicated on a Sharp tuner with the broadband noise on. With the same signal, the Serit tuner indicated 8dB MER and had the same level of Viterbi errors. I then switched the noise source off and the Sharp MER rose to 25dB, but the Serit only indicated 19dB. Signal levels were high enough not to be a factor.
I then repeated the test with a signal at the same level from DATV Express, and both MERs were 8dB with noise and 26dB without noise.
There is clearly some interaction between the tight filtering on the Portsdown TX and the measurement point (in time relative to symbol transition) of MER in the Serit tuner that is not apparent in the Sharp Tuner. I did some further tests with wider filtering on the Portsdown; the Serit indicated "no noise" MER approached that of the DATV Express as the filter was widened.
I am confident that your Portsdown would indicate 26 dB on a Sharp tuner if you are getting 23dB on a Serit.
Dave
Re: Are my constellations any good?
Hi Dave,
I suspect some inter-symbol interference is arising from the Portsdown filters. Perhaps we are a little too aggressive on the bandwidth on the portsdown, or maybe the Express is too wide, but if its good enough, I say keep it.
Mike
I suspect some inter-symbol interference is arising from the Portsdown filters. Perhaps we are a little too aggressive on the bandwidth on the portsdown, or maybe the Express is too wide, but if its good enough, I say keep it.
Mike
Re: Are my constellations any good?
That is interesting Dave.
It's possible that I am pretty close to 'best effort' then. Which is reasuring.
Jules
It's possible that I am pretty close to 'best effort' then. Which is reasuring.
Jules