A question on sensitivity

**mitaux8030** · 01-10-08, 07:42 AM

OK, so I have two different models of (handheld) radio, same manufacturer, just one is a newer model.
Gave both a good look over in the workshop to confirm they're within spec... and we're not talking about some backyard CBer here, we're talking calibrated Hewlett Packard sig gens, frequency meters, spec-an etc etc.
Both radios have the same IF bandwidth filters.
The older model has a sensitivity of about -10dBuV (too lazy to convert to raw uV) on the sig gen for 12dB SINAD.
The newer model has a senstivity of about -19dBuV, considerably more sensitive, and the mute opens a good 15dBuV earlier than the old model.

Yet, 'on air' the observations are completely reversed. The older model has the better ears and the mute opens earlier, where as the new model that on the test bench was so much more sensitive now isn't all that enthusiastic to do anything - its still acceptable but instead of being 10dB better than the other radio, its more like 3-6 dB behind 'on air'

The test location is well away from any other sources of RF that would cause in or out of band receiver blocking etc. Both using the exact same antenna, and confirmed their connections were good to each radio. Both running off their own handheld battery, fully charged, for both tests.

Can any one offer an explanation as to why this might be happening?

**beer4life** · 01-10-08, 08:26 AM

G'Day Cobber,
Using a single RF source to establish S/N ratio does not tell the whole story.
The input bandwidth is the final determining factor in real life.
The first RF stage is the determining factor, not the IF filter.
A good illustration of this is that Universal LNBs, despite lower S/N claims, do not give as good a quality signal as a single narrower B/W one.
Kindest Regards, " The Druid ".....

**bulbous** · 01-10-08, 11:02 AM

A radio with excellent sensitivity on a test set often performs poorly on an antenna due to mixing and intermod in the front-end and mixer. There will be a heap of mixing products you may or may not hear causing blocking/mixing/desensing..

A classic case of this is the ICOM 208 compared to an old IC22A/S - the 208 is miles ahead in sensitivity on the bench but just doesn't hear a weak signal as well as the 22A/S because its front-end just falls over when hit with all the shit coming down from an antenna. In this case the 208's problem is its 'barn door' wideband capability compared to the 22A/S using a 4-section helical bandpass filter for 144-148MHz.

Your newer radios probably need a tougher front-end device or even the same device running more current. You may lose some ultimate sensitivity but have a stronger front-end in the real world.

The 208 is fine out in the country but hopeless in the city.

**trash** · 01-10-08, 08:17 PM

My brain isn't much in gear tonight, so the answer isn't obvious to me, but it should be.
You have signal, noise, bandwidth, squelch and audio sections.
Bandwidth is the same. How about the gain and front end noise figure of each model ?

With the squelch open on both radios how do they compare ?

I don't normally work with 2 way radios and squelching circuits, but what I do know a little about the various types. Spectral power involved with the various types etc.
How do the squelch circuits vary between the two models ?

I'm guessing, but you might have more noise power entering the inband spectrum that comes with more gain of the newer radio.

Have you tried throwing this question out on vklogger ?

**beer4life** · 01-10-08, 09:20 PM

The more tuned RF stages, including the antenna, the lower the induced noise.
That is why they use tracking tuning of RF stages in super-heterodyne receivers.
Part of the tuning process involves making the RF stages' bandwidth track with the mixer(s) across the band of interest.
The wider the front end bandwidth, the worse the signal to noise.
The narrower the front end, and more linear the RF amps, the less noise mixer products generated before the first mixer.
You cannot get max gain across an amateur band with resonant antennae, they are usually made resonant in the segment of interest. ie: CW, AM or SSB.
Once generated in the front end, a narrow band IF filter does very little to help. Cost is usually the final determining factor for mass production.
Been there, done that.
Kindest Regards, " The Druid ". aka ' VK2-AVF '.

**mitaux8030** · 02-10-08, 07:02 AM

Both radios are single band (UHF) commercial two way units, both covering the same designed band coverage. That's design, not actual/practical bandwidth. I know they use different front end technology, one uses stripline with varactors the other ceramic resonators with varactors for front end tracking.

The 'on air' test environment I've discounted as the reason for the difference as the location chosen was well away (>20km) from any strong source of RF, either in-band or out-band. Both radios unmuted for the tests.

Just has me buggered that in a low-noise on-air test environment the better bench-performing radio works worse than the other when everything else appears to be a level playing field. I'm guessing actual/measured front end bandwidth is going to figure in the answer - but if thats the case why isn't it apparent with the sig-gen test too?

Hmm, this could be an interesting little discussion to get the grey matter thinking...
de VK7ZJA (oh love my 208H - live & work in a remote-rural environment 99% of the time so suits me well)

01-10-08, 07:42 AM	#1 (permalink)
mitaux8030 Member iTrader: (0) Join Date: Jan 2008 Location: A rock in the ocean Posts: 41 Spent time on board: 12:24:27	A question on sensitivity OK, so I have two different models of (handheld) radio, same manufacturer, just one is a newer model. Gave both a good look over in the workshop to confirm they're within spec... and we're not talking about some backyard CBer here, we're talking calibrated Hewlett Packard sig gens, frequency meters, spec-an etc etc. Both radios have the same IF bandwidth filters. The older model has a sensitivity of about -10dBuV (too lazy to convert to raw uV) on the sig gen for 12dB SINAD. The newer model has a senstivity of about -19dBuV, considerably more sensitive, and the mute opens a good 15dBuV earlier than the old model. Yet, 'on air' the observations are completely reversed. The older model has the better ears and the mute opens earlier, where as the new model that on the test bench was so much more sensitive now isn't all that enthusiastic to do anything - its still acceptable but instead of being 10dB better than the other radio, its more like 3-6 dB behind 'on air' The test location is well away from any other sources of RF that would cause in or out of band receiver blocking etc. Both using the exact same antenna, and confirmed their connections were good to each radio. Both running off their own handheld battery, fully charged, for both tests. Can any one offer an explanation as to why this might be happening?

01-10-08, 08:26 AM	#2 (permalink)
beer4life The Druid. iTrader: (2) Join Date: Feb 2008 Location: The Nether World. Posts: 632 Spent time on board: 1 Month, 3 Weeks, 2 Days and 18:30:59	Signal to noise ratio. G'Day Cobber, Using a single RF source to establish S/N ratio does not tell the whole story. The input bandwidth is the final determining factor in real life. The first RF stage is the determining factor, not the IF filter. A good illustration of this is that Universal LNBs, despite lower S/N claims, do not give as good a quality signal as a single narrower B/W one. Kindest Regards, " The Druid "..... __________________

01-10-08, 09:20 PM	#5 (permalink)
beer4life The Druid. iTrader: (2) Join Date: Feb 2008 Location: The Nether World. Posts: 632 Spent time on board: 1 Month, 3 Weeks, 2 Days and 18:30:59	Fundamental. The more tuned RF stages, including the antenna, the lower the induced noise. That is why they use tracking tuning of RF stages in super-heterodyne receivers. Part of the tuning process involves making the RF stages' bandwidth track with the mixer(s) across the band of interest. The wider the front end bandwidth, the worse the signal to noise. The narrower the front end, and more linear the RF amps, the less noise mixer products generated before the first mixer. You cannot get max gain across an amateur band with resonant antennae, they are usually made resonant in the segment of interest. ie: CW, AM or SSB. Once generated in the front end, a narrow band IF filter does very little to help. Cost is usually the final determining factor for mass production. Been there, done that. Kindest Regards, " The Druid ". aka ' VK2-AVF '. __________________

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode

01-10-08, 11:02 AM	#3 (permalink)
bulbous Thaumaturgist iTrader: (0) Join Date: Jan 2008 Posts: 308 Spent time on board: 1 Week, 1 Day and 22:54:45	A radio with excellent sensitivity on a test set often performs poorly on an antenna due to mixing and intermod in the front-end and mixer. There will be a heap of mixing products you may or may not hear causing blocking/mixing/desensing.. A classic case of this is the ICOM 208 compared to an old IC22A/S - the 208 is miles ahead in sensitivity on the bench but just doesn't hear a weak signal as well as the 22A/S because its front-end just falls over when hit with all the shit coming down from an antenna. In this case the 208's problem is its 'barn door' wideband capability compared to the 22A/S using a 4-section helical bandpass filter for 144-148MHz. Your newer radios probably need a tougher front-end device or even the same device running more current. You may lose some ultimate sensitivity but have a stronger front-end in the real world. The 208 is fine out in the country but hopeless in the city.

01-10-08, 08:17 PM	#4 (permalink)
trash Senior Member iTrader: (0) Join Date: Jan 2008 Posts: 336 Spent time on board: 2 Days and 16:32:17	My brain isn't much in gear tonight, so the answer isn't obvious to me, but it should be. You have signal, noise, bandwidth, squelch and audio sections. Bandwidth is the same. How about the gain and front end noise figure of each model ? With the squelch open on both radios how do they compare ? I don't normally work with 2 way radios and squelching circuits, but what I do know a little about the various types. Spectral power involved with the various types etc. How do the squelch circuits vary between the two models ? I'm guessing, but you might have more noise power entering the inband spectrum that comes with more gain of the newer radio. Have you tried throwing this question out on vklogger ?

02-10-08, 07:02 AM	#6 (permalink)
mitaux8030 Member iTrader: (0) Join Date: Jan 2008 Location: A rock in the ocean Posts: 41 Spent time on board: 12:24:27	Both radios are single band (UHF) commercial two way units, both covering the same designed band coverage. That's design, not actual/practical bandwidth. I know they use different front end technology, one uses stripline with varactors the other ceramic resonators with varactors for front end tracking. The 'on air' test environment I've discounted as the reason for the difference as the location chosen was well away (>20km) from any strong source of RF, either in-band or out-band. Both radios unmuted for the tests. Just has me buggered that in a low-noise on-air test environment the better bench-performing radio works worse than the other when everything else appears to be a level playing field. I'm guessing actual/measured front end bandwidth is going to figure in the answer - but if thats the case why isn't it apparent with the sig-gen test too? Hmm, this could be an interesting little discussion to get the grey matter thinking... de VK7ZJA (oh love my 208H - live & work in a remote-rural environment 99% of the time so suits me well)

Sponsored Links

Sponsored Links