Hi all,

Well, I've been doing some further thinking on this idea of a homebrew
set. Gareth pointed me to the "minima" homebrew set:

http://www.phonestack.com/farhan/minima.html

I must say, that thing is truly inspiring. It got me thinking though.

I'd like to reach up into the VHF; 2m if possible. Now I bought some
Si570 ICs the other day, and they're quite capable of reaching 160MHz.
So 144MHz is no problem. Taking the 3rd harmonic should get me to 70cm.

Deciding on a suitable IF is the next challenge. The higher I make it,
the easier it becomes to make a filter that will reject images. I
considered a bog-standard 10.7MHz IF at first, but then considered what
would happen at 70cm where the band goes from 430-450MHz.

I can easily get a 30MHz oscillator brick, so I figured that'd be a good
starting point.

For HF, I whack the signals through a low-pass filter, tune the VFO
between 30-60MHz. It might get a bit hairy on 10m, but for 15m and below
we should be good. A notch filter right on 30MHz might make 10m workable.

For 6m, a low-pass filter on 54MHz should do it, tune between 80-84MHz.

2m and 70cm is where it gets interesting. I spent a few hours trying to
synthesize a 144-148MHz BPF in Qucs last Saturday. Not as easy as I
envisaged. I'd set it up to produce a high-order Chebychev filter, then
find out that the filter calls for 1pF capacitors (!) and very specific
coils. Cauer filters got closer, but not really the holy grail I'm after.

Then I did some thinking: the Icom IC-22b apparently used 5 helical
filters in its front-end to great effect. They're a smallish rig too,
and looking at a picture of the guts of one, I see 5 metal cans that
don't look terribly huge. They also look fairly straightforward
mechanically.

The question is, how good are these in terms of roll-off? I don't expect
them to put up much of a contest against a crystal filter but I'm not
sure what's realistic.

Regards,
Stuart Longland

"Stuart Longland VK4MSL" wrote in message
...

Well, I've been doing some further thinking on this idea of a homebrew
set. Gareth pointed me to the "minima" homebrew set:

http://www.phonestack.com/farhan/minima.html

I must say, that thing is truly inspiring. It got me thinking though.



Always my problem, an excess of thinking, a paucity of getting on with it
:-)




I'd like to reach up into the VHF; 2m if possible. Now I bought some
Si570 ICs the other day, and they're quite capable of reaching 160MHz.
So 144MHz is no problem. Taking the 3rd harmonic should get me to 70cm.

10.7 MHz is pretty standard for both*****, and there are helical cavity
filters for 145MHz,
as used in the "Simple Spectrum Analyser" design (G4PMK, Nov 1989 RadCom)
Nor sure about 70cms, but there might be SAW filters, as used for the 433.92
domestic products.

Failing that. a cavity filter or interdigital filter, is a possibility.The
most
professional design around is probably Scotty Sprowls's spectrum analyser
(http://scottyspectrumanalyzer.com) and that uses a cavity filter for a
1.3GHz if,
with the second IF being 10.7 MHZ.

BUT BUT BUT as emphasized above, these are thoughts and not experiences!


***** I've a scrap main board from an FT780 which is a 70com rig but it uses
IFs
of 56MHz, 10.7MHz and 455kHz

On Fri, 28 Feb 2014, Stuart Longland VK4MSL wrote:

Hi all,

Well, I've been doing some further thinking on this idea of a homebrew
set. Gareth pointed me to the "minima" homebrew set:

http://www.phonestack.com/farhan/minima.html

I must say, that thing is truly inspiring. It got me thinking though.

I'd like to reach up into the VHF; 2m if possible. Now I bought some
Si570 ICs the other day, and they're quite capable of reaching 160MHz.
So 144MHz is no problem. Taking the 3rd harmonic should get me to 70cm.

Deciding on a suitable IF is the next challenge. The higher I make it,
the easier it becomes to make a filter that will reject images. I
considered a bog-standard 10.7MHz IF at first, but then considered what
would happen at 70cm where the band goes from 430-450MHz.

At one point, there was a wave of FM communication receivers that used
21.4MHz (twice 10.7MHz) IFs, so those should be reasonably available. Of
course, I've seen some odd filters in the HF range used in such receivers
too.

When early clunky cellphones were obsolete, I was able to buy some at
garage sales. This was the mid-nineties. They all had crystal filters in
the frequencies above 30MHz, some around 45MHz but also some around 70MHz
if I remember. The good thing was they all converted a second time to 455
or 450KHz, so there was a conversion crystal there too.

If you have to convert to some other frequency for actual selectivity,
then the you'd need to get a crystal to do that conversion, which suddenly
makes those filters expensive.

The reality is there were lots of crystal filters out there in the HF
range. Probably more under 10MHz than above, but they were all over the
place, on odd frequencies. If you need selectivity there, then you have
to hope to find one that is for SSB, or go the ladder filter route. I've
seen SSB filters at really odd frequencies, so in commercial equipment
there isn't a standard.

WIth all the recent portable shortwave receivers on the commercial market,
there may be some now that are broken, offered up at a low cost, but
having an IF filter around 45MHz, and then a conversion to 455KHz or
450KHz, the latter filters not perfect for SSB but in some receivers there
is a narrower one than the AM filter.

Michael

On Fri, 28 Feb 2014 13:10:34 +0000, gareth wrote:

"Stuart Longland VK4MSL" wrote in message
...
I'd like to reach up into the VHF; 2m if possible. Now I bought some
Si570 ICs the other day, and they're quite capable of reaching 160MHz.
So 144MHz is no problem. Taking the 3rd harmonic should get me to
70cm.

10.7 MHz is pretty standard for both*****, and there are helical cavity
filters for 145MHz,
as used in the "Simple Spectrum Analyser" design (G4PMK, Nov 1989
RadCom) Nor sure about 70cms, but there might be SAW filters, as used
for the 433.92 domestic products.

Yeah, not sure what the bandwidth is like on those. Here in Australia,
2m extends up to 148MHz, so I was looking to plonk a filter at 146MHz in-
circuit with a 2MHz bandwidth.

I found a calculator for helical resonators he http://coil32.narod.ru/
calc/helical_resonator-en.html

It looks simple enough (perhaps deceptively so).

Failing that. a cavity filter or interdigital filter, is a
possibility.The most professional design around is probably Scotty
Sprowls's spectrum analyser (http://scottyspectrumanalyzer.com) and that
uses a cavity filter for a 1.3GHz if,
with the second IF being 10.7 MHZ.

BUT BUT BUT as emphasized above, these are thoughts and not experiences!

***** I've a scrap main board from an FT780 which is a 70com rig but it
uses IFs of 56MHz, 10.7MHz and 455kHz

That would make sense then, I couldn't see how a 10.7MHz IF would work
for 70cm, but put a 30MHz IF in front of it, then no problem.

On Fri, 28 Feb 2014 15:24:11 -0500, Michael Black wrote:

On Fri, 28 Feb 2014, Stuart Longland VK4MSL wrote:
Deciding on a suitable IF is the next challenge. The higher I make it,
the easier it becomes to make a filter that will reject images. I
considered a bog-standard 10.7MHz IF at first, but then considered what
would happen at 70cm where the band goes from 430-450MHz.

At one point, there was a wave of FM communication receivers that used
21.4MHz (twice 10.7MHz) IFs, so those should be reasonably available.
Of course, I've seen some odd filters in the HF range used in such
receivers too.

When early clunky cellphones were obsolete, I was able to buy some at
garage sales. This was the mid-nineties. They all had crystal filters
in the frequencies above 30MHz, some around 45MHz but also some around
70MHz if I remember. The good thing was they all converted a second
time to 455 or 450KHz, so there was a conversion crystal there too.
[…]
WIth all the recent portable shortwave receivers on the commercial
market,
there may be some now that are broken, offered up at a low cost, but
having an IF filter around 45MHz, and then a conversion to 455KHz or
450KHz, the latter filters not perfect for SSB but in some receivers
there is a narrower one than the AM filter.

I just had a look around, and found I can get an off-the-shelf crystal
filter for 45MHz, and a crystal for 44.545MHz. The filter has a 15kHz
bandwidth.

That sounds like a better combo. I was aware of IFs of 10.7MHz, wasn't
aware of 45MHz. 45MHz would work okay for what I'm doing.

The fun bit being now, working out how to best filter the image. Suppose
I wanted a signal on 146.400MHz… LO1 would need to be tuned to 101.400MHz
(it won't reach 191.4MHz). That'll give me an image at 56.4MHz.

I fiddled with doing filters to knock out such frequencies: the 45MHz IF
will be easier than 30MHz, but I still had a great deal of trouble
synthesizing a lumped impedance filter. Transmit is of particular
concern, since in addition, the filter has to handle high power. (If I
understand correctly, it goes *after* the PA to reject any harmonic
products from the amplifier?)

What rejection figures do people normally aim for when rejecting these
images?