Screening local oscillator coils

[dr pepper]

I've thrown togther a double conversion am receiver for lf to hf frequencies, the front end and 1st osc is a Ne602 mixer/Vfo.
It works but if I turn up the volume on weak signals I get mechnical type feedback noises, I traced this back to the 1st local osc (ne602) coil, using my sig gen as the local osc removes the issue altogether.
I realise that older receivers using a coil tended to have them in a can made from tin, some ally, the former isnt antimagnetic and the latter is, maybe the tin shielded just enough.
To get around this I'm going to put the whole mixer/osc front end in a shielding can, if this was a minimal mass produced item what would be the work around, stick the coil in a can, use a different design coil one more like a transformer with a closed magnetic path, or some other method.

 

[ge]

I'm a big fan of grounded coplanar waveguide around any RF traces and putting oscillators in RF shield boxes. Companies such as Laird & Leadertech sell stock at distributors for RF shield boxes.
The mixer should be less sensitive. The LO leakage could be an issue but shielding and coplanar will not solve this.

 

[vfone]

I think would be enough to put the oscillator coil in a shielded can to solve the microphonics issue.
Use thicker wire for the coil and wind tightly on the coil support.

 

[dr pepper]

Thicker wire less tight is interesting, I'm using a ready made ferrite which is dipped.
Are you thinking this for mechanical microphonics, and/or lower resistnace & higher Q.

 

[BigBoss]

But shielding the coil will change its inductance value. Instead, screening whole oscillator will be more beneficial to prevent from interferences and obtaining the stability. Professional oscillators have always screen shields all over the oscillator circuit.

 

[dr pepper]

I have done some more experiment testing, I'm no longer sure what the issue is, except its in the oscillator.
One test I did was run the receiver with its own audio amp and speaker, as well as its own 1st local osc (Vfo), I get mechanical feedback that changes playing with the speaker wires.
Moving the speaker 6' away on long wires doesnt change anything, connecting my rf gen to pin 6 of the Ne602 makes the radio work fine, even with the volume all the way up.
Using a seperate battery powered audio amp, and using the Ne602's built in Vfo also works fine.
The Ne602 has supply 1k decoupling resistor and o.1/10u cap.
Looks like ground currents are causing the issue, I'm going to try winding a torroid for the osc coil.

 

[RCinFLA]

The core is likely powered iron not ferrite but the issue is mechanical vibration on the high permeablity core and windings around it.

The coil must be made more mechanically stable and isolated. You can try to pot it in epoxy or fabricate with lower permeable core. High permeable core usually also have a lot of drift with temperature.

 

[dr pepper]

I dont understand why mechanical vibration would affect the core when the speaker is 6 feet away on the floor.
I made an aircore inductor with some 0.6mm 'telephone ' wire, it works, now theres no feedback if I tune to a Lw broadcast station even with the volume up, but if I tune slightly away and the Agc winds up the gain I get the squeel again, inbetween stations theres no feeback.
With the old inductor I got feedback all the time with the volume anything over fairly quiet, it does seem inductor related.
Should i wind another coil on an old If can and dip it in some transformer dope?

--- Updated Jan 2, 2021 ---

I made the coil into a solid piece using some hot melt.
Also restricted the freq response of the audio amp to 10kc, but I dont think this made much difference.
I have to think junk from the speaker ground connection is getting into the oscillator tank, the cap/coil are connected to ground, and this is being mixed into the IF and being fed back.
The receiver board and amp board are star grounded, however theres a 1m lead from the ground point back to my bench supply, I guess a few 100uV caused by speaker current may be getting its way into the oscillator circuit.

 

[dr pepper]

I added some 2.5mm2 solid copper wire to the power rails, and a piece of tin plate as a shield over the bottom of the board near the mixer.
Works well now, 'cept for some drift over time so I'm looking into either Afc or a Vfo stabiliser.