Low Band CDM RX Tune?

[maxkelley_kc2spy]

Hi all,

I have a sensitivity of about .41-.45uV on my lowband 42-50MHz CDM1250, and I'm looking to tweak up the front-end filters a bit. Looking in the service manual however, under the RX tune procedure, it says that RX front-end filter tuning is not needed for the low band models... Why, and is it not really needed? Are there manually adjustable helical filters on the board, or something?

Thanks,
Max

[fineshot1]

Receiver sensitivity is sometimes a matter of perception on the users part.

If the squelch settings are too tight then that can change your perception of receiver sensitivity.

I would try doing a squelch alignment first via the professional radio cps global tuner.

[Bill_G]

Nope. Nuthin to adjust. Fixed tuned filters before and after the rf amp.

[maxkelley_kc2spy]

I will say that my squelch settings were a bit tight, but I don't have a nicely quieting signal until about .41uV... We don't use a sinadder at our shop, but most of the time do it by ear. I know most other radios I see .23-.27uV for what we consider 12db SINAD, but I don't hear that on this one until .4 or so. Bill, I took a look at the service manual last night and was disconcerted to see that, about the fixed filters. I was testing the top end (50.125, or whatever the highest test frequency is) and getting this sensitivity... I plan on moving this radio to 6m, however, so maybe I can tweak some capacitor values to get better performance at the expense of my lower end RX...

Here's a question: All of the RX front-end filter schematics for the low band boards show a R1010, which is unplaced, which effectively shorts (I think) L1003... I wonder what the purpose of this is? I don't know enough about L & C filtering networks yet to be able to analyze such a thing...

It's all a big learning process! Thanks!

[mike m]

Is the receiver front end schematic available somewhere ?

I looked around for it but I couldn't find it, if I had a copy or just a section from the front end pre-selector I or others on this groups could suggest changes, if possible.

There are a lot of on this group who convert various models of Motorola and other MFg radios to other frequency bands out of sheer boredom (were addicted to radio designs and such), many of us on this group work in the commercial radio design field and are just plain crazy when it comes to radios and such.

If you can make a copy of the receiver front end section and then PM me then I could look it over for you and make suggestions.


Mike
WB8VLC/7

[maxkelley_kc2spy]

Here it is... R1010 is supposed to be a 0 ohm resistor, but it's unplaced... Kind of interesting. I wonder what kind of effect that would have? There's a couple other unplaced caps and whatnot, too.

[mike m]

Thanks that works for me, now I just need to keep this POS dell from crashing when i pull up my modelling program.

Come to think of it this POS dell crashes even if it's opening a PDF file, it may be a few days between Dell POS crashes before I get some filter bandpass numbers.

[mike m]

Wow my PC didn't crash simply amazing, maybe I'll buy a lotto ticket or go Blonde surfing tonite, I feel lucky.

Ok this was faster than I thought it would be what with my POS Dell PC and the blue screen of death that didn't rear its ugly head this time.

My Genesys computer modelling shows a bandpass from 38.5 to 54.5 MHz at the -3db points and 40 to 53 MHz at ~ -1 dB points and the S11, S22 (in/out) match is better than -15 dB RL from 45-53 MHz degrading to around -12 dB RL below 44 MHz.

So the filtering ahead of the pre-amp looks fine.

Putting in the zero ohm resistor across L1003 screws up the overall noise figure by degrading the input output match to only -1 dB return loss and the filters loss goes up by 7.5 dB so this doesn't help you at all.

What does the back end filtering, after the receiver pre-amp Q1001 look like ?

There could be some narrower filtering after Q1001 that isn't shown in your picture.

It could also be the transmitter low pass filter cutting into the overall bandpass, the receiver and transmitter circuitry should both go thru the final Transmitter low pass filter so if you can also supply a picture of the Final Low pass filter from the RX/TX Pin diode switches up to the antenna connector this will also help.


Some receiver designers will add a slightly wider filter at the pre-amps input for lower loss figure only to add some much narrower filtering after the pre-amp for the final bandpass response.

Also can you program in some other frequencies say 47, 48, 49 and 50 MHz and tell me what the sensitivity is at these frequencies and send me a picture of the filtering after the receiver pre-amp and any additional filtering up to the mixer input ?

Also maybe I'm making it too hard, going overboard with PC modelling and all, maybe I should have asked first if there are other group members who have converted this model to 6 meters with better sensitivity than you are measuring ?

Mike

[Bill_G]

Nice work Mike.

The spec on the rcvr is .3uV for 12db sinad. So, depending on the accuracy of the service monitor gen, test cables, etc, a .4uV measurement does not surprise me. When the difference is this small, parts tolerances come into play even. And, depending on the environment the rcvr operates in, if there are just a few db of desense - which is not unusual at all in low band - the measured sensitivity probably exceeds the effective sensitivity.

[mike m]

Bill your correct on everything,

I was thinking that he probably doesn't have a directional coupler to measure his effective sensitivity other wise he would see that once the antenna is connected to his radio and the sensitivity measured with an inline directional coupler he would probably measure somewhere around 1 uVolt or worse and any improvements to the front end filter would be unnoticeable.

[maxkelley_kc2spy]

OK! Yeah, I was just kind of curious... I've not done too much with low band before, just being used to tuning up VHF and UHF stuff, so I wasn't really sure what to expect, I guess. It seems to be working OK, actually, but that's cool that you were able to model all of that! I'll have to post more schematics (from my other computer) of the other filtering...

I love learning about this kind of in-depth technical stuff... These are the kind of conversations I like on BatBoard!

[Bill_G]

Oh yeah Max - the noise floor in low band can be horrible. It's one of those bands where you can talk 50 miles truck to truck on a good day, and hardly 2 miles on another. If you're out in the country, back in the mountains, cruising the tundra, there's nothing like it. Pack it with some horsepower with a good antenna and it talks forever. Crawls around hills, down into hollers, punches through trees. But, if you get into an urban area, you can forget it. Just about anything will noise up the receivers. Loose fenders, seat springs, wheel bearings, florescent displays on the dash, bad tail light wiring, ignition noise, fuel pump noise, fan noise, neon signs, traffic signals, detector loops in the road. I could go all Forrest Gump on you with the many kinds of noise that turn your radio into stone. That's one of the reasons guys who are into it build their own antennas - so they can pick out the signal from the noise. Even a .4uV lowband rcvr can hear a gnat pick it's teeth ten counties away when the noise is low. .4 is a good number when you consider everything else in the world will push it up past 1uV effective sensitivity easily. It's fun stuff.

[mike m]

I don't mind doing it,

I have 3 weeks left on a Agilent (Genesys) modelling program evaluation and I need to use it before the time runs out.

[mike m]

I have some more information,

I finished some lunch time modelling today taking into account a plus 5% tolerance shift of filter all of the pre-selector caps and inductors and came up with 4 capacitor changes to compensate if the parts are at the high end of the tolerance range and this also measures the most sensitive components in the filter.

If the filter L's and C's are at a minus 5% shift then the top end of the filter is pushed further out which isn't a problem.

But if the parts tolerance is on the plus side then the filters cutoff frequency could shift down and there could be a -4 dB or greater loss above 51.9 MHz.

If you have better sensitivity for instance .3 uv or better around 50 MHz but much worse above this frequency then the following cap changes may help out, its worth a try but since i don't have real radio to play with then someone else needs to do the following changes.

I don't know what size chip caps the radio uses but I checked Digi-Key and they have 0805 and 0603 size ceramic chip caps in the following tolerances ranges which should cover the vintage of these radios.

Make the following changes

Change C1003 to a 82 pf (+- 2%) tolerance capacitor 0805 or 0603 whatever the proper size is
Change C1007 and C1009 to a 160 pf (+-5% or less) capacitor, 0805 or 0603 size
Change C1012 to a 200pf (+-5% or less) capacitor, 0805 or 0603 size

These 4 cap changes should account for a +-5% tolerance change in filter components to where the -3db cutoff will always be above 55.25 MHz with less than -2 dB insertion loss below 54 MHz if all parts had drifted up in value.

Again if the parts tolerance shifts down in value then there is no adverse effect to the 40 to 54 MHz filters pass band so if your sensitivity is within a db between 49 to 54 MHz then these changes will not help and you don't need to do them but if you do measure a 4 or 5 db difference then it may help out.


Mike

[maxkelley_kc2spy]

Neat-o! I should actually have some of this components sitting at our shop, perhaps... I will have to give it a try when I get a free minute, and get back to you.
Before doing it, however, I'll take readings for sensitivity across the band from 35-55 MHz, just for fun, and do the same afterwards... It'll be interesting to see what it does.

Thanks for your help and interest in this... And Bill, I've gotten a little share of the noise already I know that HF, of course, is affected by all sorts of different stuff. HF mobile is always fun to try to dig down and find noise sources... I guess VHF Lo isn't too far away!