EMRFD Message Archive 3422

Message Date From Subject
3422 2009-08-04 00:20:30 David Craggs Quadrature hybrid
I'm developing with a third-method receiver very much along the lines of G3XJP's offering in Radcom 1996.

It will have a tuneable IF of 6 to 6,5 mhz.

The front end consists of an RF stage feeding a quadrature hybrid (in the signal path rather than the usual LO path) which in turn feeds a pair of SBL1s (I'll probably replace these later with homebrew quad diode mixers with adjustable balance).

The hybrid is the usual twisted pair of inductances and two capacitors (the usual 50 ohm resistor has been replaced by a 100ohm pot).

With the theoretical values for 6.25 mhz - 250pf and 1.26uh I cannot get phase or amplitude balance. The best balance (but nowhere near useable) is achieved with the pot set to 0 ohms.

Increasing the capacitors to 560pf and with a deal less inductance I can get a reasonable phase balance but the amplitude is a long way off (of course I can compensate for this - but overall the gain is much reduced).

I've made no attempt to terminate the hybrid or the SBL1s in 50 ohms but I am feeding the input through my 50 ohm switchable attenuator.

I'll try simple diplexers at the SBL1 outputs (50 ohms + 100n to ground) and put a 3db attenuators between the hybrid and the SBL1s but otherwise I'm a bit short of ideas.

Please can anyone shed any light on this?

With thanks

David G3RYP



[Non-text portions of this message have been removed]
3426 2009-08-04 09:07:26 Rick Re: Quadrature hybrid
A few quick questions to start the discussion:

Have you looked at the quadrature outputs (using a dual trace scope) terminated in 50 ohm resistors instead of the SBL-1 inputs?

When using the SBL-1 mixers do you have LO drive?

Did you account for the capacitance between the twisted wires when building the hybrid?

Best Regards,

Rick KK7B

3428 2009-08-04 09:39:19 David Craggs Re: Quadrature hybrid
Thanks Rick

The answers to your questions are :

1 NO - First thing to try I guess

2 YES

3 NO -- second line of attack - perhaps I'll make the capacitance variable and see what happens.

I'll get back with the results

David

----- Original Message -----
3429 2009-08-04 09:56:47 w4zcb Re: Quadrature hybrid
> I'm developing with a third-method receiver very much along the
> lines of G3XJP's offering in Radcom 1996.
>
> It will have a tuneable IF of 6 to 6,5 mhz.
>
> The front end consists of an RF stage feeding a quadrature hybrid
> (in the signal path rather than the usual LO path) which in turn
> feeds a pair of SBL1s (I'll probably replace these later with
> homebrew quad diode mixers with adjustable balance).
>
> The hybrid is the usual twisted pair of inductances and two
> capacitors (the usual 50 ohm resistor has been replaced by a 100ohm
> pot).

David, there are 8 ways to hook up a pair of hybrids, only one of
which is interesting. They are very low Q circuits and I'd suggest
they might simply be hooked up wrong. You might just hook them back to
back and see what you get. Should be something like 0.1 dB IL at the
design frequency, and at that frequency, just work.

Regards and GL, hope you have your IF feedthru problems licked, some
very big signals at 6 MHz.

W4ZCB
3431 2009-08-05 04:18:07 David Craggs Quadrature hybrid
Thanks to all who helped with my problem. I'm getting somewhere.

I checked the hybrid values more carefully this time and discover that the
twisted pair has an inherent capacitance of 31pf (which surprised me a bit)
and also that the 18 turns on a T-50-6 was somewhat short of 1.27uh.
It's now 22 turns and the capacitances 220pf.

I also loaded the hybrid + and - 45 outputs with 50 ohms and now I'm
getting
closer and more sensible amplitude and phase balances over the 6 to 6.5
range.

Whether this tuneable IF is going to prove ok (given the number of strong
signals in that range) remains to be seen. I chose it because I've got
suitable crystals for most of the HF bands.

The tuneable IF system appears to have gone out of fashion with the advent
of DDS. (My 6 to 6.5 VFO is Huff and Puff stabilised by the way).

I had this notion that, if I made plug-in front-end modules for each band,
I
could use them with any transceiver/receiver I care to put together.

Now I'm going to see whether mixers with adjustable balance make much of a
difference.
.

David
3434 2009-08-05 06:00:44 w4zcb Re: Quadrature hybrid
The tuneable IF system appears to have gone out of fashion with the
advent
of DDS. (My 6 to 6.5 VFO is Huff and Puff stabilised by the way).

I'm interested in whether you are using Spargarrens original or the
"Fast Huff n Puff" of G7IXH's? I have used both in years past, likely
still have a couple of each. If you're using one of Klauss's, I'd be
tickled to send you an "upgrade". BIG improvement.

W4ZCB
3435 2009-08-05 16:47:21 Allison Parent Re: Quadrature hybrid
Hi,

The easiest and bes way to solve the IF leakage problem is shielding.
Put that IF in a box (one made of soldered copped clad works well) and filter every thing that goes in(dc and control) and everything that comes out. Use coax to connect to the shielded bandpass filer (6-6.5mhz I believe your using) which is also enclosed. Then make sure the converter stages are suitably constructed that only those
signals desired enter and exit. Makes for bigger and likely a bit
heavier but look at any of the radios (especially Collins boat anchors) and you will see good electronics and lots of metal
to keep signals where they should be. This generally makes the
difference between and OK homebrew and a stellar one more so
than any exotic electronics.

There are ways to do this with printed circuits but then layout is everything. Boxing up everything is time proven.

One hint even the VFO is boxed and the protruding shaft is broken with
an insulating coupling to the outside to keep unwanted radiati
3436 2009-08-05 16:47:47 Allison Parent Re: Quadrature hybrid
Hi,

The easiest and bes way to solve the IF leakage problem is shielding.
Put that IF in a box (one made of soldered copped clad works well) and filter every thing that goes in(dc and control) and everything that comes out. Use coax to connect to the shielded bandpass filer (6-6.5mhz I believe your using) which is also enclosed. Then make sure the converter stages are suitably constructed that only those
signals desired enter and exit. Makes for bigger and likely a bit
heavier but look at any of the radios (especially Collins boat anchors) and you will see good electronics and lots of metal
to keep signals where they should be. This generally makes the
difference between and OK homebrew and a stellar one more so
than any exotic electronics.

There are ways to do this with printed circuits but then layout is everything. Boxing up everything is time proven.

One hint even the VFO is boxed and the protruding shaft is broken with
an insulating coupling to the outside to keep unwanted radiati
3437 2009-08-06 22:50:45 David Craggs Re: Quadrature hybrid
Thanks for the tips Allison

Yes - copper-clad is useful stuff (if a bit pricey). Since I built my cutting jig it's quite easy to make really neat enclosures.

These days I make a lot of use of wood or MDF for the basic radio frame (perhaps lined with aluminium foil) and wire up the circuitry ugly-style on bits of copper-clad with shielding if necessary. I also make the enclosure a lot bigger than really needed.

I may build a more convential and better-looking? version later but more often than not it doesn't work as well as the prototype!

I'll leave miniaturisation to the professionals. It suits them because it's cheaper.

73

David


[Non-text portions of this message have been removed]
3448 2009-08-08 18:24:22 w4zcb77 Re: Quadrature hybrid
S'cuse the bandwidth, I did a bad thing yesterday and downloaded IE8. Bad decision!!! and the only way I could recover was to image my hard drive with the latest HD backup, which unfortunately was made 10 days ago, and I lost all my e-mails from then forward.

I have been communicating privately with G3RYP(I'm pretty sure) Whoever it was I was discussing Huff n Puff and the Bedford RX with, would you please drop me an e-mail? We'll pick up from where I lost you.

Regards
W4ZCB
3449 2009-08-08 23:14:27 Dave Brainerd Re: Quadrature hybrid
w4zcb77 wrote:
> S'cuse the bandwidth, I did a bad thing yesterday and downloaded IE8. Bad decision!!! and the only way I could recover was to image my hard drive with the latest HD backup, which unfortunately was made 10 days ago, and I lost all my e-mails from then forward.
>
> I have been communicating privately with G3RYP(I'm pretty sure) Whoever it was I was discussing Huff n Puff and the Bedford RX with, would you please drop me an e-mail? We'll pick up from where I lost you.
>
> Regards
> W4ZCB
>
>
>
>
That's a good reason to have a separate data partition to place all
your E-Mail, photos, schematics, etc. That way when(not if) you have to
wipe the program partition, your data is unaffected. Doesn't alleviate
the need for a backup of data.

Dave - WB6DHW
<http://wb6dhw.com>