Images in recievers?

[Ed/KB1HYS]

I should know this, but I can hear the same station in two places on a reciever.  I think I've misaligned it, would this be a cause - poor image rejection?  This happens on a 455kc IF reciever but only in the 1.6-5Mhz band, the BC and 6-18MHZ ranges have no images what so ever...  I can hear the oscillator runinng on another reciever. Is it possible that I've got the oscillator on the wrong side (producind a 455kc down instead of up).

[WQ9E]

Ed,

Most receivers use high side tracking, particularly for the lower bands, so the oscillator should be heard at dial+455 in the other receiver.

The most obvious symptom of wrong side tracking is inability to get the oscillator to closely match with dial readings across the band but this will decrease image rejection since it usually also results in the RF and Mixer stages not tracking with the oscillator.  For the same reason, misalignment of the RF and or mixer stage will result in reduced image rejection.  So first check that the oscillator is tracking on the right side (and is in reasonable agreement from end to end of the dial) and then check alignment of the other stages to make sure they are tracking with the oscillator.

A related question, PE did a two part review of shortwave receivers in Nov./Dec. 1972 and Hirsch Houck labs claimed an image rejection figure of around 45 db for the single conversion Radio Shack/Realistic DX-150A on 10 meters.  Anyone ever measured their DX-150A?  That seems like an incredibly high number for a low cost single conversion (455 Khz. IF) receiver.

[Rob K2CU]

Ed,

You might want to check the difference in the two frequencies you are receiving, at least to the accuracy of the dial you are reading. 910KC would be the difference for a 455 KC IF. Note that in many parts of the country, radio stations will transmit with multiple satellite transmitters spread out in a larger region than the principle transmitter will cover. These will display no known difference if dial readings. They rarely announce the additional id's of the satellite transmitters.  I think that WDRC AM in CT is one such station.
"WDRC is the flagship station of Buckley Broadcasting's four Connecticut stations running a talk format with the slogan "The Talk of Connecticut" (the other stations are WSNG, Torrington, WWCO Waterbury and WMMW in Meriden). "

You may be hearing the same program material for whatever station you are tuned to on multiple channels.

[KB2WIG]

Leaning on Robs' coments. A quick and dirty check is to grab a nuther ceiver and see if the 2 freqs. programing is in sync....  some of the local FM broadcashters repeat with a delay......  Hay, you never know.



If you have a GDO or a signal generator, see if you get 2 sigs out of the rcvr.

klc

[k4kyv]

I have never head an image on the 75A-4 or the R-390(A), although they do have a few minor birdies.  My Kenwood R-1000 is very poor in this respect unless some kind of pre-selection is used, since the front end runs wide open with bandpass filters instead of tuned RF circuitry.

The older classic receivers with 1 or 2 rf stages and 455 kc/s all have images to some extent. My old Nationals with one rf stage always seemed to do well through 40m, but images began to appear on 20, and on 10 the image was almost as strong as the real signal. Some of those old receivers used to be called a "railway radio" because it whistled at every station.

My HRO Sr, with two rf stages was image-free up through 20m, but the highest frequency coil performed miserably - hum, poor sensitivity and stability, and images all over 10m. 15 was marginal, with no bandspread (the set was built before there was a 15m ham band).  Some receivers go to double conversion at 7 mc/s and higher; that helps the image problem, but sometimes with double conversion comes spurious garbage and  poor dynamic range.

Hallicrafters came up with an interesting idea back just before WWII. One or two of their receivers with only one tuned rf stage included a passive "image rejector" circuit. I remember reading about it but don't recall exactly how it worked, maybe nothing more than a series tuned circuit to ground. It had a separate manual control on the front panel. The way it worked was that first you would tune in the station, then rotate the control to null out the image if there were any. They claimed it allowed image-free reception across the entire HF range, but the feature was soon discontinued.  Maybe it didn't work as claimed, or else users found it too complicated, inconvenient or difficult to use. This might be a worthwhile idea to research and maybe try to design an outboard unit to go between receiver and antenna. It should work better than an active pre-selector, which tends to degrade the dynamic range of the receiver.

I tried a web search for the Hallicrafters image rejector with no results. Maybe someone can come up with info on it.  I think it was advertised in QST, so maybe going back through some of the old issues would yield some results.  There may have even been a technical article on it in QST.

[KM1H]

Don, all that was is a tuneable image trap and dropped sensitivity of the wanted station a few S units. It was tried and died on the SX-18 of which I have and confirmed its a dog when the reject is used. Otherwise it is a basic SX-16 type 1 RF Halli.

I havent tried my 1935 HRO on 10M but a 1946 NC-183 works very well with about 20-25dB rejection when all the leaky mica caps are replaced and all the grime washed off coils, etc and all stages are well tracked. I have a HRO-50 with the AC coil and its seems very good but I havent measured the rejection.

[k4kyv]

One thing that will help with images is to use an antenna tuner* between the antenna and receiver.  If one uses a tuner on TX, installing the antenna T/R relay so that the signal comes into the receiver through the tuner just as it goes out from the transmitter through the tuner, will give several dB of image rejection. I would assume most stations would already be set up like that anyway.

*By "tuner" I mean a resonant tuner to match a balanced open wire tuned or hi-Z feed line, or single end-fed wire antenna to the transmitter, not a "transmatch" simply designed to tune out the stray reactance of a supposedly flat coax line.

[KM1H]

I wonder how much attenuation you will get at 910kc away on 10M with a single tuned circuit in the typical 300W MFJ tuner or the automatic one in many riceboxes?

Sometimes I'll use a modified DB-22A with only one tube which has been changed to a 6GM6 and set the gain pot to just enough to improve the SNR. With 3 tracked tuned circuits the rejection is at least 15-20dB which helps those 1 RF receivers a bunch.

[WQ9E]

There was an ER article about the SX-18 "Sky Challenger" and its "image rejector".  The Q of the tunable image trap was sufficiently poor that on the higher frequencies it attenuated the desired signal nearly as much as the image. 

It did make great advertising copy until units actually got into the field and owners discovered its decided lack of effectiveness.

[k4kyv]

There was an ER article about the SX-18 "Sky Challenger" and its "image rejector".  The Q of the tunable image trap was sufficiently poor that on the higher frequencies it attenuated the desired signal nearly as much as the image. 

It did make great advertising copy until units actually got into the field and owners discovered its decided lack of effectiveness.

I wonder if that was just poor design, or if it is inherent to the basic concept.  The Q would have to be higher on higher frequencies to discriminate between the image and the signal frequency, since they would be increasingly closer together as a percentage of the signal frequency as the latter increases.

I would try a series resonant circuit shunted across the signal path as the rejector, to be used in conjunction with the normal parallel-tuned grid circuit for the TRF stage.