Tweet
Does anyone have experience in nullifying a whites coil that drifted out of spec? I was able to successfully split the case to gain access to the inside of the coil. I can get it to stop overloading with a gain of 15 on my V3I but the 22.5khz frequency VDI is skewed. They have a small loop of copper wire towards the top center of the coils that helps with this skew but I'm going at it blindly. I'm not entirely sure what the V3I wants to see from the RX coil for a proper null. I did find a forum somewhere else talking about the x, r and V null but I'm not sure about how to properly measure these on my scope or setup the procedure. Could this all be done using the V3I as my signal generator and probing the RX coil inside the detector at the connector? Should I use a signal generator? I don't have much experience nulling a multifrequency coil so any help would be appreciated.
-
-
It's best to use a sig gen and an oscope. Connect the sig gen to the TX coil and the oscope to the TX coil (ch1) and RX coil (ch2). Set the sig gen to a 10kHz sine wave with the highest amplitude possible. Trigger on the TX and adjust the scope to see 2 or 3 cycles at max amplitude. The RX signal should be tiny, crank up the ch2 V/div to see it. If you have a digital scope, an averaging of 2 or 4 is often useful.
At this point it will be instructive to take a piece of ferrite and a small piece of household foil and wave them around the coil to see how the RX signal reacts. The ferrite should produce an RX that is exactly in-phase with the TX, or exactly 180° depending on how the coils are connected to the scope. No big deal if they are backwards. The foil should be close to a 90° shift, but more importantly as it is barely detected you should see the RX phase start to roll.
The large loop of wire does what the ferrite does: we call that the X-null (X==reactive). So as you move it around you should see big changes in the RX amplitude, but either at 0° or 180°. That is, until you get very very close to zero amplitude, and then you will likely see the phase roll around while the amplitude hits some minimum.
This is where you then do the R-null (R==resistive). The foil could do this, but White's usually did this with a small shorted loop of wire. It looks like I see one near the nose of the coil, though it doesn't appear to be glued down. Move this loop around until the RX signal phase is either 0 or 180. The readjust the X-null to a better minimum. Repeat if necessary. You want the null to be at least 60dB, or 1mV on the RX for every volt on the TX. That is, if the TX = 10Vpp then the RX should be 10mVpp. Preferably you want it better than that if you can manage.
The V-null was an extra step to ensure the null holds up at 22.5kHz. I think it was done by hand-selecting a coupling cap that's under all that glue. Hopefully you won't need to deal with that. Take the sig gen up to 22.5kHz and see if the null holds. If it does not, then try re-nulling X & R at 22.5kHz and then go down in frequency and see if it holds. If it does not then the V-null is bad and I don't remember what to do there. I'm sure it would require digging into that hot melt and figuring out what's under there and how it's connected.
As a final step, try dropping down to 3kHz and see what the null looks like there. With many coils, a 10kHz null will hold up at anything lower.
-
Thank you Carl for the very informative answer. This clears up a lot of questions I was after. Last night while trying, I was using the detector for my signal and wasn't comparing it to the TX side. The X and R null make more sense now as well. I was wondering what that tiny loop of wire was for. I have to get my hands on a signal generator. My TDS 540 oscilloscope should be up for the task. I will post back my findings once I get a generator. Hopefully sometime next week.Comment
Comment