Phase Noise
Phase Noise
Hi Guys,
I have a question with regard to the effect on received signal strength of using a synthethised LO (Elcom) as compared to, for example a highly efficient and modern LO such as marketed by Kuhne. My two 47Ghz transverters each have a kuhne mixer/IF amp unit, and each use an FT-817 as IF rig, but one has a Kuhne LO, while the other uses an Elcom. We noticed a marked difference in received sig strength - the Kuhne unit receives, eg an S6 SSB or FM signal, whereas the other received S0. Can this difference be explained by phase noise differences between my two oscillators??
(I guess I'll need to buy another Kuhne LO for maximum performance...)
Dan
I have a question with regard to the effect on received signal strength of using a synthethised LO (Elcom) as compared to, for example a highly efficient and modern LO such as marketed by Kuhne. My two 47Ghz transverters each have a kuhne mixer/IF amp unit, and each use an FT-817 as IF rig, but one has a Kuhne LO, while the other uses an Elcom. We noticed a marked difference in received sig strength - the Kuhne unit receives, eg an S6 SSB or FM signal, whereas the other received S0. Can this difference be explained by phase noise differences between my two oscillators??
(I guess I'll need to buy another Kuhne LO for maximum performance...)
Dan
Re: Phase Noise
That's a good question Dan, however you may want to undertake some further troubleshooting before heading down a phase noise (PN) path as your first option. Have you tried swapping LO's on the mixer modules to see if you have the same performance in both mixers using the Kuhne LO? Not sure whether the 47Ghz units came as kits(?) and whether you soldered the mixer diodes by hand(?), it's reasonably well documented that performance differences are noticeable among various built mixers at 47 and above, the process of mounting the diodes themselves can contribute varying mixer performance unit to unit.
As for the practical perception of phase noise, you'll notice the effect of a LO with high phase noise as an increasing level of noise as you tune onto a signal (carrier), it can be rather marked. The signal you're listening for can in worst case be swamped and this may have an impact on S meter readings, however you'll notice noise contributing to the S meter indication as well(?).
I'm assuming the Elcom synths are running around 12Ghz(?), the phase noise probably isn't all that bad in the base unit and considering you only have to double to 23Ghz odd for injection into the 47Ghz mixer (they run as sub-harmonic) you shouldn't experience such a large degradation in phase noise performance, remember it's 20LogN (N being the multiplication) added to the phase noise degradation of the LO. I'm assuming the Kuhne LO is xtal based and as such has probably quite good PN performance compared to synthesized LO's?
To give you an example of poor PN performance and the practical perception of it in a microwave Rx sense... I have an early (model 1152) JWM PLO from the USA, one of it's selectable frequencies is 1136Mhz and I'd wanted to use it in a 10Ghz transverter feeding 10224Mhz (x9) as the LO injection. The basic PN rating of that model JWM PLO was around -90dBc. The resulting PN is calculated to be (90-19) approx. -71dBc and this resulted in the receiver effect that I described above. That is, as you tune onto a 10Ghz carrier the noise level rises quite markedly against the carrier signal itself, that's the practical effect of applying a LO signal to that particular mixer (DEMI 3cm transverter) having a PN performance as high as -71dBc.
There's been a number of discussions and articles describing just what is a practically acceptable level of PN performance for microwave LO's... this is just one (practical) example from my experience. Another is the now popular PLO's coming out of Israel for around US$45 each, these are great value. They can be ordered pre-programmed anywhere from 600 - 2700Mhz. If you take one programmed for 2556Mhz it'll ship with a PN of around -88dBc, from my experience. To use it as the basis of a 10224Mhz LO you'll need to x4 and the resultant PN will be of the order -75dBc (calculated). I've not tried it, but some ops are using these in 10Ghz transverters and clearly -75dBc is an acceptable level of PN performance? Interesting... I'll let you know my experience when I've tried it.
I hope that helps, it'll be interesting to understand what you find?
73, Ted
As for the practical perception of phase noise, you'll notice the effect of a LO with high phase noise as an increasing level of noise as you tune onto a signal (carrier), it can be rather marked. The signal you're listening for can in worst case be swamped and this may have an impact on S meter readings, however you'll notice noise contributing to the S meter indication as well(?).
I'm assuming the Elcom synths are running around 12Ghz(?), the phase noise probably isn't all that bad in the base unit and considering you only have to double to 23Ghz odd for injection into the 47Ghz mixer (they run as sub-harmonic) you shouldn't experience such a large degradation in phase noise performance, remember it's 20LogN (N being the multiplication) added to the phase noise degradation of the LO. I'm assuming the Kuhne LO is xtal based and as such has probably quite good PN performance compared to synthesized LO's?
To give you an example of poor PN performance and the practical perception of it in a microwave Rx sense... I have an early (model 1152) JWM PLO from the USA, one of it's selectable frequencies is 1136Mhz and I'd wanted to use it in a 10Ghz transverter feeding 10224Mhz (x9) as the LO injection. The basic PN rating of that model JWM PLO was around -90dBc. The resulting PN is calculated to be (90-19) approx. -71dBc and this resulted in the receiver effect that I described above. That is, as you tune onto a 10Ghz carrier the noise level rises quite markedly against the carrier signal itself, that's the practical effect of applying a LO signal to that particular mixer (DEMI 3cm transverter) having a PN performance as high as -71dBc.
There's been a number of discussions and articles describing just what is a practically acceptable level of PN performance for microwave LO's... this is just one (practical) example from my experience. Another is the now popular PLO's coming out of Israel for around US$45 each, these are great value. They can be ordered pre-programmed anywhere from 600 - 2700Mhz. If you take one programmed for 2556Mhz it'll ship with a PN of around -88dBc, from my experience. To use it as the basis of a 10224Mhz LO you'll need to x4 and the resultant PN will be of the order -75dBc (calculated). I've not tried it, but some ops are using these in 10Ghz transverters and clearly -75dBc is an acceptable level of PN performance? Interesting... I'll let you know my experience when I've tried it.
I hope that helps, it'll be interesting to understand what you find?
73, Ted
Re: Phase Noise
Dan
absolutely it does.
over / under driving the LO could also be the same issue. The Kuhne LO's are only 1mW out.
some surplus telecoms device could be way more than that if they are used in distributed systems.
a good article on phase noise vs MDS is here:
http://www.w1ghz.org/new/Phase_Noise_and_MDS.pdf
and I have to disagree -75 dBc phase noise is not acceptable for a performance systems ... calculating that out thats 2-3 dB loss in MDS signal
its a get you going PN!
to me that could be a make or break QSO!
the Kuhne PLL's are about as good as you will find on the market just now.
worth every penny! all my tvtrs are loaded with them.
have fun on 47g and good to see some activity down under on that band
rgds
Simon
absolutely it does.
over / under driving the LO could also be the same issue. The Kuhne LO's are only 1mW out.
some surplus telecoms device could be way more than that if they are used in distributed systems.
a good article on phase noise vs MDS is here:
http://www.w1ghz.org/new/Phase_Noise_and_MDS.pdf
and I have to disagree -75 dBc phase noise is not acceptable for a performance systems ... calculating that out thats 2-3 dB loss in MDS signal
its a get you going PN!
to me that could be a make or break QSO!
the Kuhne PLL's are about as good as you will find on the market just now.
worth every penny! all my tvtrs are loaded with them.
have fun on 47g and good to see some activity down under on that band
rgds
Simon
vk2gg wrote:Hi Guys,
I have a question with regard to the effect on received signal strength of using a synthethised LO (Elcom) as compared to, for example a highly efficient and modern LO such as marketed by Kuhne. My two 47Ghz transverters each have a kuhne mixer/IF amp unit, and each use an FT-817 as IF rig, but one has a Kuhne LO, while the other uses an Elcom. We noticed a marked difference in received sig strength - the Kuhne unit receives, eg an S6 SSB or FM signal, whereas the other received S0. Can this difference be explained by phase noise differences between my two oscillators??
(I guess I'll need to buy another Kuhne LO for maximum performance...)
Dan
Re: Phase Noise
Many thanks, Ted and Simon.
I have swapped over the two Mixer modules - they are both pre-built, not kit ones. There did appear to be some improvement in received signal strength, but that was over a relatively short path, some 8km. The 32km result was that SSB offered some improvement in copy over FM, but that is to be expected. SSB copy was 59 on the weakest end, (ie with the Elcom), but with no sig showing on the meter, Vs S6 on the other end, with the Crystal Osc LO (Kuhne). Interesting that noise would result in an increased signal level (on the 817's meter). I will consider another Kuhne LO. (what's 200 Euros anyway?)
We'll see when we attempt 60 km, hopefully before the humidity imcreases too much.
Stay tuned!
cheers,
Dan
I have swapped over the two Mixer modules - they are both pre-built, not kit ones. There did appear to be some improvement in received signal strength, but that was over a relatively short path, some 8km. The 32km result was that SSB offered some improvement in copy over FM, but that is to be expected. SSB copy was 59 on the weakest end, (ie with the Elcom), but with no sig showing on the meter, Vs S6 on the other end, with the Crystal Osc LO (Kuhne). Interesting that noise would result in an increased signal level (on the 817's meter). I will consider another Kuhne LO. (what's 200 Euros anyway?)
We'll see when we attempt 60 km, hopefully before the humidity imcreases too much.
Stay tuned!
cheers,
Dan
Re: Phase Noise
A signal source for 47 GHz - for the DB6NT basic TRV’s with no image rejection!
Because that TRV has no image rejection filter, will RX (and TX) on both 47088 and 46800 MHz.
Oscillator mixer injection at 46944 MHz (assuming the sub harmonic mixer output) a 144 MHz IF has a +144 at 47088 and a –144 at 46800 MHz.
So 1300 MHz, from say a 23 cm FM standard C6000 TX rig pumped in hard to a miniature harmonic diode will output 46800 MHz. That is 1300 MHz x 36 = 46800.
Now with one of these TRV units, a real twist here. When receiving your QSO partner, you may hear him twice. The offset, if say 20 KHz apart will be the offset of both parties oscillator injection.
And both will be the same side band.
Steve, ZL1TPH
Because that TRV has no image rejection filter, will RX (and TX) on both 47088 and 46800 MHz.
Oscillator mixer injection at 46944 MHz (assuming the sub harmonic mixer output) a 144 MHz IF has a +144 at 47088 and a –144 at 46800 MHz.
So 1300 MHz, from say a 23 cm FM standard C6000 TX rig pumped in hard to a miniature harmonic diode will output 46800 MHz. That is 1300 MHz x 36 = 46800.
Now with one of these TRV units, a real twist here. When receiving your QSO partner, you may hear him twice. The offset, if say 20 KHz apart will be the offset of both parties oscillator injection.
And both will be the same side band.
Steve, ZL1TPH
-
- Frequent Poster
- Posts: 130
- Joined: Tue Nov 25, 2008 8:45 pm
- Location: Mid North Coast NSW
- Contact:
Re: Phase Noise
Dan
Do both of your LO's put out the same RF level? Is the RF level out of the Elcom unit higher than the Kuhne LO?
I have an Elcom unit here that puts out +11dBm on 11.xxxGHz, and this is probably too high for a lot of mixing circuits (+7dBm) and multipliers.
Try measuring the RF out of the LO's and you may find that your Elcom LO is putting out too much level and overdriving your mixer. (Just a thought).
73... Bruce... VK2EM
Do both of your LO's put out the same RF level? Is the RF level out of the Elcom unit higher than the Kuhne LO?
I have an Elcom unit here that puts out +11dBm on 11.xxxGHz, and this is probably too high for a lot of mixing circuits (+7dBm) and multipliers.
Try measuring the RF out of the LO's and you may find that your Elcom LO is putting out too much level and overdriving your mixer. (Just a thought).
73... Bruce... VK2EM
73
Bruce VK2EM
Bruce VK2EM
Re: Phase Noise
Guys,
A new Kuhne LO has solved the problem! It injects 40mW - far more than the Elcom, and obviously less phase noise, or so it would seem. I now have two equally good receivers - In fact we set a new Australian Record of 58km for 47GHz!!! Woopee. It was a very weak signal, with fade and some noricable drift, probably attributable to drift in the 817, as the Txvrs were locked on Rubidium both ends. Humidity was probably about 51% - not bad for such a warm day as yesterday. Once dishes were locked at each end, we both had reasonable SSB copy, with terrible drift. A great day out. (The path was LOS - Gan Gan to Mt Sugarloaf). Many thanks to Peter, VK2YGM, Jack VK2TRF, Les VK2APE, and Heather & Irene for the catering!
Dan
A new Kuhne LO has solved the problem! It injects 40mW - far more than the Elcom, and obviously less phase noise, or so it would seem. I now have two equally good receivers - In fact we set a new Australian Record of 58km for 47GHz!!! Woopee. It was a very weak signal, with fade and some noricable drift, probably attributable to drift in the 817, as the Txvrs were locked on Rubidium both ends. Humidity was probably about 51% - not bad for such a warm day as yesterday. Once dishes were locked at each end, we both had reasonable SSB copy, with terrible drift. A great day out. (The path was LOS - Gan Gan to Mt Sugarloaf). Many thanks to Peter, VK2YGM, Jack VK2TRF, Les VK2APE, and Heather & Irene for the catering!
Dan
Re: Phase Noise
Well done Dan - pioneering stuff indeed! Dave told me the good news on air yesterday during my drive home.
73 - Rob VK2GOM / G0MOH
73 - Rob VK2GOM / G0MOH
Re: Phase Noise
Dan,
congrats on the 47 GHz contacts.
I don't understand the drift you were getting. An FT817 would not normally drift noticeably, whether on 144 or 432.
Is the frequency locking of the transverter LO working? From your description it isn't.
Was it drifting by several hundred Hz per minute, or several kHz per minute? Or more?
Andrew
congrats on the 47 GHz contacts.
I don't understand the drift you were getting. An FT817 would not normally drift noticeably, whether on 144 or 432.
Is the frequency locking of the transverter LO working? From your description it isn't.
Was it drifting by several hundred Hz per minute, or several kHz per minute? Or more?
Andrew
Re: Phase Noise
Andrew,
upon checking my 10MHz source, I found a broken cable!
That would account for the drift. Better results next time!
cheers,
Dan
upon checking my 10MHz source, I found a broken cable!
That would account for the drift. Better results next time!
cheers,
Dan