VFO tracking LPF for superhet. rx image rejection?

[ChristofferB]

Hi!

I've run into the (apparently) age old problem of always listening on 2 frequencies on my home-built HF receiver due to me having no input filter, and that's fairly annoying.

The old school solution is to have a tracking BPF or LPF at the input, with the variable cap mechanically (by belt, axle, etc.) connected to the VFO tuner, but that'll take up a lot of space (and i suspect alligning the VFO with the filter would be hell, if i'm just using scrap air variable caps) so i'm looking for alternative solutions.

Here's a beautiful ASCII rendition of my receiver:

                                                       +30db                                               AD831              MuRata                  +30db      AD831     opamp   
 LONG WIRE ANT.--->[1:9 balun]---->[LNA]--->[LC ant. tuner]---->[LNA]--->(X)---->[455KHz xtal filter]---->[IF LNA]--->(X)--->[AF AMP]--->speakers
                                                                                                                   ^                                                               ^
                                                                                                          VFO(AD9833) 4-12.5MHz                                   BFO(455KHz)       

So i want to be able to tune about 4-12.5 MHz and my IF is 455 KHz. I think it's easiest to tune to the lower signal and call the higher one 'image' thus only needing a low-pass filter and not bandpass. The image is then always 455*2 above,

I was messing around with the idea of a 3 component chebyshev filter, and then switching in parallel capacitors and shorting/unshorting inductor taps with relays/transistors/4066's, to give me a filter for each whole MHz of tuning, with an Fco just low enough to cut out the image signal, but that's not very elegant, having 4-5 relays per filter component...

Is there an elegant solution to this? maybe a varactor tuned filter?

thanks in advance!



   

[radiogeek381]

First, ROCK ON! for building an HF receiver from scratch.   

Second: short answer, a passive tunable LPF with a knob can be effective, and is a time-tested approach. (Look at lots of older receivers for a knob labeled "preselector" or "resonator.")

More Detailed Answer:

Why is there an LNA between the antenna and the first filter?

For an HF receiver, where there is such an abundance of man-made noise, the LNA at the front end is unlikely to be of much use.  But more importantly, the gain element functions as a mixer.  This mixer creates intermodulation products from unwanted stations beating against each other. This can put more than images into the IF passband.  That bare LNA is likely to cause more grief than pleasure.

For the same reason, you may find varactor tuning in the front end problematic.  There probably are HF receivers with varactor tuned front-ends, but passives are a good first-line-of-defense.  That junction in the varactor just begs to become a mixer.

For an example, I looked back at a venerable (and perhaps revered)  TRX design from the 70's -- the TenTec Triton IV.  It had a preselector before the gate of the first RF amp.  The inductor was slug-tuned (!) from the "resonator" knob on the front panel.  I notice that the Signal/One CX7 also had a passive preselector before the first RF amp. (The latter design is really something -- I owned one back in the 80's and have regretted selling it ever since.)

At a minimum, you'd benefit from a high-pass filter with a cutoff around 3MHz if you are in an area with broadcast stations in the 540-1600 kHz region... And with any kind of broadband gain near the antenna, a low pass with a cutoff between 13 and 50MHz would be a good idea too.  Intermod from FM broadcast stations can be particularly noxious as the signals start out very wide.

The scheme you have looks like you can move some things around and experiment with the gain distribution/budget.  This is a really big deal, as not all gain is good gain.  You might also try alternative mixers for the first mixer.  The AD831 looks really cool, and has pretty good IP3 specs.  I wonder how it would compare to a passive mixer like the TUF-1 or SBL-1 (perhaps there's a museum near you that has one  ) The latter are passive, and are more "bullet resistant."  (This is nothing but a suspicion, but I wonder how the 17dB difference in LO drive level affects spurious responses.  The AD831 requires -10dBm of LO drive, while the SBL-1 requires 7dBm.  Imagine a spurious tone (say from the BFO)  that couples to the 831's LO input at -50dBm.  That's only 40dB down from the real LO.  But for the SBL-1, that same spur is 57dB down.  Like I said, this is only a feeling, not the result of careful study or experiment. )

Again, good-on-ya for building an HF receiver. 

[ChristofferB]

Hey Thanks for the reply!

The first LNA is a (don't judge) TV antenna amp (i know, 75 ) i added it because the entire receiver was low on gain (no real measurements, but all signals were very quiet).
I'll replace it when i get another MMIC LNA board, but right now it lets me hear a huge amount of stations that were too weak without.

I made a mistake in the drawing, the two first LNA's come right after the ant tuner:


                                                                                +30db      +30db     AD831             
 LONG WIRE ANT.--->[1:9 balun]--->[LC ant. tuner]--->[LNA]---->[LNA]--->(X)--->

So any semiconductor in the signal path will receive radiated signals and mix them in, gotcha. maybe putting a HF can on each module would be good.

You're absolutely right, every block diagram block is its own little pcb with SMA interconnects, so the entire RX is basically lego!

I'll try making a passive front-end! actually i think i might just re-purpose the antenna tuner, since it doesnt really seem to help much..


Having a HPF and a LPF to box in the region i'm interested in is probably a good idea too.

I did consider the SBL-1, but then i came across these: eBay auction: #https://www.ebay.com/itm/AD831-High-Frequency-RF-Mixer-Inverter-Module-0-1-500MHz-9-11V-single-ended/192325536668?epid=18006529230&hash=item2cc77ecb9c:g:ULIAAOSwFYxZ1e8G

Most of my LNA's
and they looked so handy i decided to base the entire design on them!

Thanks for the ideas! I'll have a look at the manuals of the CX7 you mentioned and see if i can't learn something from the preselector!

[vk4ffab]

The main trouble you have is the IF at 455khz and the low frequencies of interest. The easiest solution would be to make the receiver double conversion, with the first IF somewhere high to avoid all the image troubles then the 2nd IF at 455khz where you do the AM, SSB and CW filtering.

[bd139]

This. A double conversion is not much more effort these days.

TBH I’ve built a couple of DC receivers and you get used to the double image problem and learn to tune around it after a week or two. There’s something nice about not solving problems sometimes

[ChristofferB]

You're right, you can absolutely tune around it, if you wanna listen to a freq. that has a bad image ontop of it, chances are if you switch sides (high/low), you can find the same freq. with a less terrible image, but not always, and when you're just 'surfing' for interesting stuff, it's tiring for the ears to always have RTTY in your SSB 

Maybe i'll make it double conversion at some point. I have played around with the thought of making a generic superhet. fixed freq. receiver at 21.4 MHz with 455KHz IF, and then use that for everything, make a front-end for each band i'm interested in. i think that would be pretty elegant.

Thanks again! it sounds like the immediate solution is to throw filters at it, and long-term is add an IF stage. I'll get right on that!

--Chris

[bd139]

Also worth looking at single conversion quadrature detector based receivers. See: http://www.norcalqrp.org/files/AustinNC2030Presentation.pdf

[vk4ffab]

Thanks again! it sounds like the immediate solution is to throw filters at it, and long-term is add an IF stage. I'll get right on that!

--Chris

If your frequency's of interest are much below say 5mhz you will still struggle to remove the image with band pass filters as you just will not get a sharp enough skirts on the filter to totally keep them out, obviously that problem becomes greater the lower in frequency you go. Even if you were interested in just the 40m amateur band and designed a bandpass filter centered on 7.150 that is 400khz wide, when you look at where the image lies with a 455khz IF it is not all that far out of the bandpass.

Solution, use a higher IF, go direct conversion and use phasing as @bd139 pointed out, or go dual conversion. From the sounds of it, you are making a general coverage receiver rather than having an interest in specific bands, if that is the case, then dual conversion is really the way to go. Make the first if about 45mhz, use an off the shelf 45mhz IF filter, and make all the image issues go away. You can then get away from having lots of preselector filters in the front end also, just have a low pass at the highest frequency of interest and a high pass at the lowest frequency of interest.

You would need another mixer module, an LO that is higher in frequency than the 12mhz max ATM, and a couple of xtal oscillators for the 2 BFO's and with careful selection of LO frequencies, you can mix and match where you have sideband inversion and where you do not, for receiving LSB and USB on the same IF filter.

[HB9EVI]

No doubt, that the IF=455kHz is pretty low for a reasonable image rejection - normally for a single Super, you would rather choose a high IF around 9 to 12MHz. Additionally to mention it: the AD9833 has quite a poor performance signal wise; clocked with 25MHz I absolutely don't see a usable signal at 12MHz.

so in short words:

- choose a higher IF, based on crystal ladder or lattice filter
- choose a different DDS, at least a AD9834 with 75MHz clock or better a AD9850 with 125MHz

[Z80]

455kHz is old school from the days when crystal filters were expensive so a high first IF was used to split the image, then a low IF for selectivity with cheap ceramic filters.  Nowadays you can build your own crystal filer easily and cheaply and make a single conversion superhet that will have similar performance to a commercial HAM set without too much difficulty.  I recently made an SSB transciever with a 10MHz IF from junkbox parts which is pretty much on par with my FT817 on recieve.  You don't really need a preamp on HF, in fact an attenuator is more useful especially on the lower bands.  You do however need a good input bandpass filter, the narrower the better if you want performance.  The less crud that hits your first mixer the better the performance will be.  Keep most of the gain in the IF after your crystal filter.

[bd139]

Yep they're not hard. I have made a couple.

Reference article I used was in ARRL QRP Classics (out of print): http://www.skzw.pl/A/ARRL%20QRP%20Classics%20(1990).pdf (page 199).

I made a G3UUR crystal analyser described here: https://www.mikrocontroller.net/attachment/195160/Crystal_Motional_Parameters.pdf and a counter and matched a box of 100 crystals from aliexpress with it.

[David Hess]

Varactor tuned preselection is common in wide range VHF and UHF receivers but it would still require switching to cover the whole HF range.  I would worry about the varactors creating more intermodulation products especially with so much RF gain.

Some HF receivers can use the transmitter's switchable harmonic filter for preselection but it is not narrow enough to handle the 455kHz image.

Current driven diode bridges work well as switches for capacitors in a low pass filter but as you point out, this gets complicated fast.  Sometimes transistors are used instead.

I think you would be better off adding gain after the IF or adding a 2nd IF so the RF amplifiers can be removed.