At 8/1/2010 01:15, you wrote: > > But there are other solutions: if you want a brute-force > > window filter they're out there too. I have a few 5 MHz > > wide 1 dB loss UHF filters sitting on the shelf here that > > I acquired at the Dayton & Ft. Tuthill hamfests. ~$40 each. > >Seems like a good deal... but the 5MHz band-width is only >(for me) usable in some system applications. Some of those >applications where the 5MHz band-width would be excessive >but the expected Simrex band-width would not be...
Then we're back to the pass cavity solution. Just saying there are other inexpensive, well-engineered options out there. > > Again, the loss is low enough that in most cases a leading > > preamp simply isn't needed. > >Leading or Post Preamp? "Leading" meaning "pre-preamp". > > >Kind of makes it look like helicals are a sin doesn't it... > > > Ever wonder why the MVP/MastrII & Micors are so deaf > > compared to more modern RXs? > >Nope... > > > I haven't measured the loss of the UHF helical assembly, but > > the VHFHB front-end helicals have ~6 dB of loss. In all > > those radios, their own helicals effectively are all the > > post-preamp filtering you'll ever need. > >No it isn't... if you sweep the front end of Micor you'll >find it's actually fairly wide. I seem to remember sweeping >some GE Receivers and their front ends were relatively wide >on the order of at least a few MHz. From Memory I seem to >remember the Micor being at least 4MHz wide. The VHF HB MVP front-end helical assembly has a 3 dB BW of 1.8 MHz. At 40 dB down the BW is 5.7 MHz. Granted the selectivity curves of the Simrex preselector curves are narrower, but keep in mind that they are in fact misleading, since most of that selectivity is AFTER the preamp stage. Add to that the fact that the mixer in the GE radios has very high dynamic range (remember, the stock GEs don't have a gain stage ahead of the mixer), & you likely end up INCREASING the GE's susceptibility to IMD by using one. > > > > The GLB preselector preamp has 4 helical stages of unknown > > > > (unknown to me) coupling. > > > > > >Depends on the Pre-selector Model and age of the box ... > > >I have GLB units here with two pre-device stages and three > > >post-device stages. And I have versions with a more traditional > > >helical design and others with more of a lumped parts layout. > > > > > >A 2-Meter version I have lots of pictures of has 1 stage of > > >pre-selection and four trailing stages. The active device is > > >an MRF-901. > > > OK. MRF-901 NF @ 2M is ~1 dB, so maybe 2-3 dB NF for the unit. > >And that jives with my informal recorded notes for the 224 MHz >GLB Pre-selector with a dual gate Mosfet. > > > Not bad for VHFHB, but having only 1 little resonator in front > > of the active device doesn't offer it much OOB protection. > > Better put a (gasp) PASS CAVITY in > > front of it! ;) > >To quote someone who recently wrote: > >"As you so often like to state, it all depends on the >application - in many cases it simply isn't necessary." Agreed: in the above example the Simrex preselector isn't necessary: simply omit it & use just a pass cavity. > > >However, > > >In more than a few real world situations you might really > > >need the filter pre-selection a lot more than the most > > >optimum NF. A practical trade of pre-selection for a slightly > > >higher noise figure can and does sometimes make the difference > > >in a usable radio system. > > > OK fine. But again, we DON'T KNOW the noise figure for the > > device. > >It's not mandatory to know the NF for every situation, only >helpful for those specific situations where making a logical >assumption is not allowed. Kind of like saying you don't need to know how much output power your TX is running, so long as your users can hear it. > > Fun to play with? Yes. Can solve some IMD/overload > > problems? Certainly. But not a tool for any seriously > > engineered RF system. > > Bob NO6B > >Really depends a lot on whose money you're spending. I've >seen a lot of seriously engineered RF systems that don't >work very well out there in the real world. ...and in almost every case I've seen this, it's due to the engineering failing to take into account all of the real-world parameters. If your models are flawed, everything falls apart. Bob NO6B
