Broadband amplifier sanity check and IF frequency.

[setq]

I'm building a small RF amplifier for the front end of a 20/40/80m receiver I'm working on, very slowly. Plan is receive SSB/DSB/CW. I can't be arsed with a license yet; I'm only interested in the technical challenge so far.

Plan is to have a signal relay switched set of tuned bandpass filters on the front end to reject other bands, then a broadband amplifier and then the first mix to IF. Is it better to have a broadband amplifier here or one tuned amp per band?

Also is there anything I need to know when selecting an IF? Currently looking a slightly non standard 10MHz for the sake of crystal availability as I already have a ton of these floating around.

Also trying to build a PLL based VCO with a 74hc4046 but that's another thread!

[nugglix]

Hi!

Sounds interesting, /me is doing something similar.
Also no license. 

I'm targeting the AM and 160m frequencies.
So a smaller frequency range for the input circuits.

WRT to amplifier:
Did you think about some attenuation?
Might come in handy on crowded bands w/ strong signals.

I'd go with a small amplifier for each band.
I think that a well behaving broadband amp is much more complicated
than a "simple" amp for each band.
In my view the pre-amp is nothing more than a 1 or 2 stage transistor amp.
Of course your idea might be completely different.

I decided to use 10.7 MHz as first IF.
Lots of filters available -- and cheap crystals
I'm heading for an NE592 as IF amp.

As LO I went with a "normal" LC Collpitts.
Which is was built 3 times, this version looks quite okay -- to my eyes at least.
Of course, drift is still an issue.

Hope to read more on that!

Cheers

[setq]

Thanks for your reply.

I did consider AM+160m as there's a lot of traffic on 160m here in the UK but a reasonable dipole antenna length is a little too long for my property unfortunately! Not sure I'd get far with a ferrite antenna!

Good idea with the attenuator. I will look at that. At the moment any signal would be nice

I've built a few colpitts oscillators and can't get a stable enough one with the parts I have on hand so decided to up the specs and go for a PLL based synthesizer.  I can get it frequency stable but I can't get one to be mechanically stable; if I put my hand near the VFO unit it drifts. This might be because I'm using crappy old recycled air variables and the shielding is crap. The shaft is insulated though. Very annoying. That's analogue for you though.

I've built a very simple PLL synth with a reference frequency of 10Hz divided down from a 3.2768MHz crystal just to get a feel for how they work and that works nicely. Just need to suss all the loop equations and scale it up now (the hard bit). Probably have to replace the VFO with a varactor tuned one yet.

I'll take a look at the NE592 - quick glance suggests it's a nice little device!

[nugglix]

Hi!

I've chosen the band because of the low frequency.
It's my first project, so I need something easy.

Idea was: just a radio w/ some SSB capabilities.
The current block diagram shows a double super...
That's how things go!

I got the drift of the oscillator down to a few Hz per minute.
The warm-up time is quite long though. But there is always something... 
A good shielding is essential, as I discovered.
My oscillator uses varactor tuning, which works quite well.
At least after you take the simulation only as a (very) rough guide.
The varactor model isn't the best. And I've no clue how to make it better.

Attached the oscillator schematic.
The 2 BF199 drivers are still missing, I was eager to see it working.
The "counter" one is for a frequency counter - and general measurement point.
The lower "mixer" output is meant to go into the LO pin of a NE602.

The PLL thingy sounds interesting. Would be nice to read more about that.

So, one short stop at the bench before bedtime.
Had some 10.7MHz crystal filters in the mail today.

Cheers

[CJay]

Grab a service manual for something like the Kenwood R1000 or other shortwave receiver that has switched bands and see how they do it.

Then grab a manual for a modern(ish) radio like a Yaesu FT817 and see how it's done in there.

Work out which version is better for you.

[retrolefty]

I would use a single 50 ohm matched broadband RF amp. Dynamic range rather then low noise is the best attribute to design to on these bands. Also consider allowing the RF amp to be switched out as strong overload is frequently found of the HF bands. The mixer stage dynamic range is a critical element also.

[setq]

Thanks all. Much appreciated. I'm going to use a canned mixer (mini circuits).

[dmills]

20/40/80? I would not bother with an amp, the sky noise from any kind of decent aerial is so huge that it will completely swamp any noise advantage gained from an amp in front of the first mixer.

For me it would be Aerial->switched attenuator->switched BPF (per band) -> mixer.

Note that mixers are termination sensitive, so following the mixer with a diplexer to provide a broadband termination is a good idea (And make the first IF filter somewhat wide, it helps the filters linearity).

Usually a modest 10dB of so of gain after the first IF filter is indicated for noise reasons (And to isolate the filters), with the lions share of the IF gain being after the main selectivity.

Regards, Dan.

[setq]

Yep 20/40/80 is target. Gotcha - that all makes sense. Thanks for the hints - much appreciated.

So consolidating ideas, the current RX block diagram is pretty much :

antenna -> L match -> attenuator -> BPF switched per band -> mixer from VFO -> bridged T diplexer -> IF filter -> IF amp -> 2nd mix from BFO -> LPF -> audio

BFO is switchable between upper/lower SB. Main VFO is a synthesized PLL based on 74hc4046 if I can get the thing stable else I'll probably cheat and use one of those nice Si570's. L-match is considering the TX half of it which may appear if I can be bothered to get a license or think I might annoy Ofcom

Trying to build the whole thing in modules like a lego kit.

Tasked myself with working out what the hell to do with an antenna today and it appears that step 1 is move house

[David Hess]

These days a broadband preamplifier is not a difficult task; offhand I do not know of any designs which use separate preamplifiers within the HF bands.

As dmills points out, HF noise levels are usually so high that any RF preamplification will not help with signal to noise level and will actually lower the dynamic range of the receiver.  What the RF preamplifier does however is block the first local oscillator and mixing products from radiating out the antenna so it is often included for just this reason alone.

[setq]

Interesting. So in fact the RF preamp will just provide some reverse isolation. I will consider that. If it provides impedance conversion/matching and no voltage gain that will have some value as well.

I've built a broadband amp already. Seems flat-ish (I have no decent equipment here ) between 1-30MHz with +12dB of gain. Really simple common emitter amplifier with some negative feedback between base and collector. No inductors other than an RFC in the supply line.

[PA0PBZ]

Plan is to have a signal relay switched set of tuned bandpass filters on the front end to reject other bands, then a broadband amplifier and then the first mix to IF. Is it better to have a broadband amplifier here or one tuned amp per band?

As soon as you put bandpass filters in front of the amplifier it's not broadband anymore, so the whole discussion about broadband/tuned amplifier can be skipped 

[setq]

This is true but the amp has to be able to amplify all signals at each of the pass bands equally and these might be disparate frequencies or extended later.

[C]

I might suggest a couple of things.

Plan for more "BPF" when you start with so you can add more easily at a later date.
 Think of a broad coverage communication receiver that also has added filters for ham band use. A filter switch lets you have best of both. This also allows easy addition for additional front end's to be added.

David Hess's reverse isolation point can become very important when adding a second radio. Do you want to pickup radio waves from air or your radio with second radio?

Think of using more than one AGC stage. That first attenuator is not to control desired signal amplitude as much as prevent RF amp and mixer overload. Having many outputs of your mixer allows adding a AGC module that covers just this part of radio. An additional output would also allow a PAN display. Many uses for outputs here but with some costs.

It easy to think of what desired signal you want, need also to think of what is present at each stage that you do not want. Think of that KW signal you do not want next to the QRP signal you do want.  You run the RF stage at max gain while preventing overload then start over and do same with the IF stage. Your  desired signal keeps increasing while not wanted gets reduced.

[David Hess]

You could also play games with varactor tuned bandpass filters before the RF amplifier which will relax some of the filter requirements for minimum passband and selectivity.

[C]

Need to be careful with varactor's in RF path as the RF signal could change the varactor's value causing filter to change based on the RF.

[David Hess]

Need to be careful with varactor's in RF path as the RF signal could change the varactor's value causing filter to change based on the RF.

And cross modulation could be a problem as well.  I suspect using a varactor tuned preselector is better than nothing if you have a wideband front end which would overload anyway without it but not quite as good as separate fixed preselectors.  There are some interesting tradeoffs there.

I think I have seen it done for the HF band but I know it has been done for VHF and higher.

[CJay]

Need to be careful with varactor's in RF path as the RF signal could change the varactor's value causing filter to change based on the RF.

And cross modulation could be a problem as well.  I suspect using a varactor tuned preselector is better than nothing if you have a wideband front end which would overload anyway without it but not quite as good as separate fixed preselectors.  There are some interesting tradeoffs there.

I think I have seen it done for the HF band but I know it has been done for VHF and higher.

It's been done and is still being done in cheap CB equipment and is apparently quite effective.

If you're curious grab a copy of the circuit diagrams for the Radioshack HTX10, President Lincoln, Anytone 5555 and a few of the Ranger/Galaxy CBs, they all IIRC use varactor tuned front ends.

I would imagine but don't know for certain that it's also used in higher quality equipment but as you say, there are tradeoffs which may preclude its use in better equipment.