Improving Reciever selectivity-Tuned input or 2nd IF?

[KI4YAN]

Right now, I am working on a 6 meter receiver, mostly following the GE Ham News article for September-October 1951. Me and two friends are building the receiver in the article, and each of us will build a matching transmitter for a local net-trying to get more local guys into 6 meters again.

Here's the article: http://n4trb.com/AmateurRadio/GE_HamNews/issues/GE%20Ham%20News%20Vol%2006%20No%205.pdf

I've got all the bits to build it exactly as it's drawn in the article, including some actual made-for-5Mhz IF transformers. The article's author claims that the selectivity, while adequate, is not the best.

Here's where my parts differ from the article's parts:

My tuning capacitor is a 3-section unit, with two sections being 12pf, and one section being 18pf. My plan *was* to use the two 12pf sections, with the oscillator on the high side-as per the original.

My IF transformers are actually made for a 5Mhz IF-I did not have to modify them to get them on frequency.

That said, I have not punched the main chassis yet. I have built the RF amp and mixer/oscillator on a small separate chassis, and it does mix 50-54Mc down to 5Mc. AGC action is "OK" and selectivity is "OK" but it is a little wide. Sometimes centering up on one signal leaves you with one other just above the noise, but not every time. Sometimes that second signal is WAY above the noise.

If I was to change the LO to run under the band, using the 18pF capacitor, I would have two sections of 12pf, and perhaps the currently broadband input could be converted to a normal tuned input, with a little fiddling with the LO inductor.

Would that appreciably improve the selectivity, over say using a second IF of 250kc? I have three Collins F-250-A-67 mechanical filters, center frequency is 250kc, bandwidth is 6.7kc. I've considered using one of them as a second-IF, but I feel like it would be putting the selectivity in the wrong part of the receiver.

What would you choose to do?

[w8khk]

I looked at the article and schematic for the receiver you are building.  Looks like a fun and interesting project.

A 5.0 MHz IF will not provide good adjacent-channel selectivity.  When the article was written, there was probably a lot less activity on six meters, and transmissions were spaced further apart, so great selectivity was not as much of an issue then as it is now.

An additional tuned RF stage at the incoming frequency may help with image rejection, but it is likely not needed due to the high IF frequency choice.  It will not help with adjacent channel selectivity.

An additional lower frequency IF stage, 455 KHz or 260 KHz will greatly improve adjacent channel interference rejection, and with the lower second IF a filter might not be required to attain your desired performance.  Adding a filter will, however, require more gain in the IF section to overcome the insertion loss of a filter.

I would suggest experimenting with a second converter from the 5 MHz to a lower IF frequency of your choice, and use a signal generator, or signals received from your antenna, to determine just how much additional selectivity is required.  I would NOT suggest a lower IF frequency instead of the 5 MHz already planned, as this would result in less image rejection on six meters.

In the article, it is stated that a second conversion and lower IF was not included because the additional selectivity was not needed (at that time) and the added complexity and power consumption was not justified.  You have a clean slate, and can probably find some examples of double conversion six meter receivers and get some good ideas on how to implement your double conversion receiver.  In summary, adding an additional RF stage or tuned circuit at 50 MHz will not add any significant adjacent channel selectivity.

Hope this helps, good luck with your project!

[KI4YAN]

It does indeed. I only have two 5Mhz IF transformers, but I did think another IF stage would be prudent, as long as I had good AGC. Mostly because I've not seen many recievers with only one IF amp, but I have built a few and they all worked just fine.

If I was to use one of the mechanical filters, how much loss am I looking at? I can have an IF amp before and after.

[w8khk]

If you have two 5MHz IF transformers, you can easily implement one IF stage at this frequency.  Additional stages at 5MHz may not add significantly to the selectivity, but remember the sole purpose of the high IF is to afford image rejection.  Additional stages at this frequency may introduce stability problems.  Original TRF (Tuned Radio Frequency) receivers, prior to the introduction of the superheterodyne, were difficult to adjust, and several stages in cascade could oscillate and require neutralization.  So I would suggest you minimize the number of 5MHz stages.

Look at some handbook articles or search QST for some of the simpler double-conversion receiver designs.  Some just had a converter from a high IF to a low IF with no intervening IF amplifiers, and attained good image rejection, selectivity, and sensitivity.  Some research on these designs can be very revealing.

You might even get by with a single 455 or 260 low IF stage without a filter.  Adding the filter could incur an insertion loss of six to ten dB.  One other thing you might look at to gain additional knowledge in this area is the Kiwa filter site: http://www.kiwa.com/  specifically: http://www.kiwa.com/kiwa455.html

This will give you a much better idea of the impact of adding a filter, and they also have one with a built-in amplifier with zero insertion loss.  Study their other filters, insertion loss, bandpass, and of course price.

If you have access to the ARRL QST Members Only Web Services, search for articles in the 50s and 60s by Byron Goodman, he has written many articles on receiver design for good selectivity without getting overly complicated.   Many of his designs could be applied to your second conversion and low IF and filter needs.

One thing to keep in mind is the bandwidth of an IF amplifier, even without a filter, is proportional to the frequency of the IF amplifier.  That is to say, a 5 MHz IF amplifier has a much greater bandwidth than a 455 KHz IF, and a 50 KHz or 250 KHz is narrower than a 455 KHz amplifier.  On the lower bands, signals are crowded together, and filters are generally necessary; but on six meters, signals may be spaced far enough apart that just the bandwidth of the second IF after the converter may be sufficient to avoid adjacent channel interference. 

Be creative, but do not be afraid to study other implementations and integrate their ideas into your project.

[KI4YAN]

The 2 IF transformers are a good thing-I meant using a single IF amp at 5Mc, a fixed converter down to whatever IF can I can find or use one of the Collins 6.7kc wide filters I have.

[KI4YAN]

Here's as far as things got tonight. I have a National MCM dial for it-but my dial scale will roll backwards due to the odd tuning capacitor. The cabinet is an old National Radio one that I got from a mailing list 5 years ago, just waiting for a project I felt could do it proud.

Things are looking good so far...now it just needs to work as well as it looks!

[w3jn]

All of this depends greatly upon how much LO drift you'll have.  Doing a second IF with the mechanical filter will be great if you have a stable LO.  If not, you'll be constantly re-adjusting the receiver to keep the signal in the filter passband.

It's really hard to build a stable VFO, particularly at these frequencies.  Indeed, that might be one reason the original design doesn't have a lot more selectivity.  With the 5 MHz IF you're probably looking at a 50 KC IF bandpass.  Probably good enough for 6M these days, I don't think there's a ton of activity there. 

You could leave room for an additional mixer and IF stage.  Any crystal in the 5 MHz range would be fine since you can adjust that 5 MHz IF to anything from 4-6 MHz to convert down to 250 KHz.

You could also just use a 10.7 MHz IF and use some of those little three-legged crystal filters intended for FM - I got a pair to replace some in a service monitor on eBay for a couple bucks.  15KC would be just the ticket for this, I'd think.

[KI4YAN]

The 1st oscillator seems fairly stable so far. Benefits of using TV tuner parts-they had to be stable to stay on channel, so lots of those bits are already tempco'ed if you can lift them *all* and use'em. So far, the receiver receives and converts to the 5Mc IF fine-it's just a bit wide. I have some 540kc IF transformers for the second conversion, but I still need to check and see if they're good-they have the caps built into the base and may be junk or may need a cap-ectomy.

[KI4YAN]

Unfortunately, as-built the thing is broad as a barn door. Single-IF stage passband is 34kHz...Adding in the second 5Mhz IF stage gets it down to 15Khz but that's still about twice as wide as I'd like.  Haven't got the proposed 540kc stages going, the IF cans needed a cap-ectomy. The little silver mica caps inside had shorted, once those were removed the actual windings are ok so just need to tune 'em up and they'll work.

I went ahead and built the detector/BFO on a breadboard and it's "OK". AVC works decently, but the diode detector as layed out in the article struggles with strong signals and doesn't have the sensitivity to match the front end. (MDS when watching the scope using an outboard infinite impedance detector is 0.043uV, so -134dBm?) If I want to use the diode, either more IF gain or a more sensitive detector will be needed...maybe my 6AL5 is just flat.

To fiddle with the IF gain, I tried 6AK5's in the IF amp spots, and it did help some. I have some 6BJ6 here too, but they might be *too* hot and oscillate...In order to get *any* visible audio off the 6AL5 detector, input had to come up to 0.01uV. The infinite impedance detector doesn't do AVC though, I thought, so I started looking it up. I'm thinking of comboing a plate detector with an infinite impedance detector to get audio off the cathode and AVC voltage off the plate, we'll see how it works.

All in all, the front end from the article is definitely workable, and if you're in an area with low noise and a relatively quiet band, the rest of the reciever might work a charm for you. But for my location here, the noise floor is too high and the band too crowded to work exactly as the article describes. I haven't built the audio output section yet, though, because I don't have a 6AK6 pentode. I'm planning to use a 6BL8 or 6DX8, or maybe a 6AQ5. I'm not worried about power consumption, unlike the article's writers. My small bench power supply puts up 280V at 120mA and 6.3V at 3A, and may get the 5V winding rectified and regulated to 6.3V to run a small regulated supply for screens or something. I just haven't needed the regulated bit yet so I never added it in there.

So far, the RF amp, 1st mixer/oscillator, and 1st IF amp work pretty dang well. Like I said in the beginning, (and as is mentioned at the very end of the article) it's a 34kc passband, but good sensitivity and hard to overload the front end. I'll get the detector and audio stages worked out next, then work on the 2nd IF.

My local buddy who is also building the reciever has built it exactly to the article's design, and says he's satisfied with the performance. He's working on getting his license soon...but has been working on it for years. We'll see if building the receiver prods him into finishing his license.

[N8ETQ]

   Yo'

      Prolly better off installing an IF output jack
and sending the if to your R-390, or any decent RX
that tunes 5 mc.

   Just a thought, nice work on the RX..

GL es 73

/Dan

[KI4YAN]

Hoping to get the second IF running "proper" tonight. The he plan is to have the on-board audio and infinite impedance detectors hooked to the 540kc IF strip, and try to tune it down to 455kc and feed the IF of my little Sparton into it. (The Sparton is a very nice AM/FM unit from 1954, with excellent channel separation on AM.)

That should let me tune and adjust the 2nd IF and get the AGC voltage set up.

[Opcom]

If you have the 5MHz IF parts already and want to use them, maybe a second conversion to 500KC or something low for one or two stages would help narrow it up depending on what crystals are avail. for the converter oscillator.

[KI4YAN]

That's what I've done.

Current layout is 6AK5 RF amp, 12AT7 1st osc/mixer, 5Mc IF transformer, 6BJ6 IF amp, 5Mc IF transformer, 6BL8 2nd osc/mixer, 540kc IF transformer, 6BJ6 IF amp, 540kc IF transformer, 6AK5 infinite impedance detector, 6BL8 audio output.

I'll try to get my schematic posted up soon as it's mostly working properly. The infinite impedance detector has a diode detector tacked into the plate circuit to generate AGC voltage.

[KI4YAN]

Looks like the 2nd IF is going to be 567kc, driven by my drawer of crystals. Not sure if that'll be with in the tuning range of my IF transformers, but they did tune up at 540 to take the IF out of my other radio, so another 20kc should be there.

News at 11.

[K4RT]

I have enjoyed this thread. I would like to see the schematic, and some shots of the under-side of the chassis if you have any.

[Opcom]

I have enjoyed this thread. I would like to see the schematic, and some shots of the under-side of the chassis if you have any.

ditto esp. the schematic, love to learn more from others projects..