Circuit gets feedback when near something magnetic - video

[Stonent]

I've noticed on the CDV-700 Geiger counter that I'm working on, the general area of the transformer seems to generate feedback when anything magnetic is near it. I'm guessing the screwdriver is messing with the field and causing it to feedback into the transistor?

[SeanB]

Photo of the other side, likely the transformer is running at close to saturation, as only one magnetic pole will drive it into the region where you get noise. Possible causes are a missing air gap shim, a cracked core or side or the material has warped. If you cannot fix those use a small ferrite magnet to apply a bias to the core to keep it out of saturation, that is often done on pot cores to keep them in a linear or low loss portion of the BH curve.

[Alexei.Polkhanov]

I have two theories. First it could be same effect as in fluxgate magnetometers. Perhaps you have high permeability core somewhere on the circuit and that changes the hysteresis curve depending on strength and direction of magnetic field. To test the theory I would try to take a magnet and rotate while keeping at the same distance from circuit to see if feedback effect changes with angle.

Second - magnet should affect the Beta and Alpha particles in a tube, but not Gamma. As far as I remember beta is stopped by metals especially heavy - polyethylene is good for stopping alpha particles. So maybe you can try to add some selective particle blocking and see how magnet affects it then?

[Stonent]

Here's a general photo of the board, not my personal unit but the same model. The black box is one of the transformers and the one that seems to go crazy when anything magnetic is near it.

[Alexei.Polkhanov]

Ops, Sean had same theory about the hysteresis while I was typing my post :-)  Rotating magnet would be nice way to test it.

[Stonent]

Schematic: http://www.datasync.com/~rsf1/fun/cdv700sc.jpg

Q2 feeds T2 which then goes to V2 which is a Victoreen Corotron 900V vacuum tube regulator. I think T1 steps it back down again to provide the sound pulses and the meter pulses.

[Stonent]

Along with what you guys said, if I remove the piezo speaker, T2 makes its own noise when it feeds back. If I turn the voltage all the way down, it will stop until I turn it back up again.

[Alexei.Polkhanov]

Well here are some videos that show how to look at the effect on scope:



You should try to probe output from transformer to see how pulses are effected by magnetic field and how they change if you rotate it. Fluxgate compass sensors respond to position of Earth's magnetic field this way.

I would not discard all other theories yet :-)

[Stonent]

Ok the magnet test... I can with a magnet manipulate it to make the frequency go up or down or seem to stop the audible oscillating and work the way it is supposed to.

[SeanB]

Probably you should find an old monitor and look for the linearity coil and grab the magnet stuck on top of it to place on the transformer. Most likely the core in the potting has cracked at the top so is responding to stray fields as they drive the core in and out of the linear region as it receives the pulse. Either that or you need a new transformer, as the potted one is not likely to survive the depotting to either rewind it or epoxy the cracked cores back together.

[Stonent]

Probably you should find an old monitor and look for the linearity coil and grab the magnet stuck on top of it to place on the transformer. Most likely the core in the potting has cracked at the top so is responding to stray fields as they drive the core in and out of the linear region as it receives the pulse. Either that or you need a new transformer, as the potted one is not likely to survive the depotting to either rewind it or epoxy the cracked cores back together.

I'm not as versed on transformer specs as I'd like. According to the manual the primary winding between pins 1 and 2 is 4500 Ohms. The second winding between pins 3 and 4 is 11 Ohms and the third winding between pins 5 and 6 is 5 Ohms. With spread of +/- 20%

The 2 separate winding do not appear to be connected together in the schematic and seem to be separate.

I'm not exactly sure how I'd order a transformer to be like that. I checked mouser but there were lots of parameters that I had no idea about.

The ratios come out at about 410:1 and 900:1

[SeanB]

That will be a DIY one, using a RM10 or such pot core and some 40SWG wire and some kapton tape to provide insulation and hold the windings. Doable and easy enough to wind by hand. When you get to 10 000 turns of 48SWG wire a winder with a counter is not a luxury item.