On 2019-10-02, we had problems with the G4 (Wendy Butts) Hall Effect Probe while replacing the Sensor & Control Assy (see viewtopic.php?f=13&t=187&start=70#p2749). We temporarily replaced it with the probe from the Stage Butts and Penri took the suspect one home for testing. He was unable to identify any faults using his manual test rig, so I brought it home for testing against the old SAC that we removed on that day.
There is definitely something amiss with that probe, but it is intermittent. In order that there was a control test subject, I initially tested the 'blue' probe. This is the probe that I used to test the SAC after I had built it and has a blue top cap and magnet cover. Everything worked exactly as expected and I observed every level at 25 mm intervals from 0 mm to 900 mm.
I then connected the suspect probe (this is pink and has a concrete foot, unlike the blue one). The initial reading with the magnet at the lowest point was 50 mm, which seems to be the lowest that can be read because the foot stops the float going any lower. I then slowly moved the magnet up the pole and obtained anomalous readings almost straight away. My newest test software does not print out the conditioned voltage or the dip value. it simply prints the depth or "No Sensors Triggered" if the magnet is between sensors (surprisingly hard to do with the blue probe, but seems more prevalent with the suspect pink probe). As I moved the magnet up the pole, I got a lot of "No Sensors Triggered" outputs, then I got "975 mm" (?). This was when the magnet was around 200 mm. I then continued lifting the magnet and got some good readings and some more '975' ones.
To be on the safe side, I disconnected the pink probe and reconnected the blue one - no anomalies. I then substituted the pink probe and this time it seemed to work, albeit with a few more instances of "No Sensors Triggered".
I need a bit more time to think about this, but a possible cause is a fault on the '75' chain. If this was intermittently failing (eg going from some voltage to 0 V), then the software would definitely give us a '75' reading, but I can't think why it would be 975. Also, what we observed on Wednesday was no intermediate values not 975. I can't think of any failure modes that might cause this. If either the red or black wire was intermittent, then the probe wouldn't work at all. If one of the sensors was intermittent, then I'm pretty sure the result would be a dip at that level. If the '75' wire was intermittent, then I'm pretty sure the result would be a dip at every level, but not a 975 mm result.
Any ideas?
G4 Hall Effect Probe
Re: G4 Hall Effect Probe
Could it be that one or more of the wires in the cable coming off the probe is damaged?
If that happened, it could mean that there's conductivity only some of the time, and hence it behaves unreliably?
If that happened, it could mean that there's conductivity only some of the time, and hence it behaves unreliably?
Hamish
Re: G4 Hall Effect Probe
Anything is possible, but a failing wire would affect the reading at every level and not give a 975 reading (we would get 25, 125, 225, etc).
Terry
Re: G4 Hall Effect Probe
Fair point. What if there was some metal exposed so it shorted when the cable was bent a particular way? I can't really think of anything else to try at this point.
Hamish
Re: G4 Hall Effect Probe
I'm going to take the connector apart later, but I suspect that the circuit in the probe may be compromised in some way (leakage or something).
Terry
Re: G4 Hall Effect Probe
Terry
If you can take it back to WMT I’ll pick it up and go through it component by component.
Penri
If you can take it back to WMT I’ll pick it up and go through it component by component.
Penri
Re: G4 Hall Effect Probe
OK. All the wires seemed to be firm.
The probe is now in the Workshop.
The probe is now in the Workshop.
Terry
Re: G4 Hall Effect Probe
I just realised that there is a mechanism whereby the results I saw could result in that value (using my unit test code at least).
If the '75' line was faulty (either the wiring or the sensor stack), then the 'dip' could appear permanently in that line. Since the code checks for a dip at each level, starting at 0 mm, then it would find it at 75 mm and write '75' to the variable carrying the depth, then find it at 125 and overwrite the depth, then 225, etc until the last level is reached, which happens to be 975 mm in the limits.
So I guess we are looking for an intermittency in that stack.
That still doesn't explain why we only saw 100s on Wednesday.
Terry
Re: G4 Hall Effect Probe
Given the probe a bit of a going over, re-soldered joints which may have been dry, checked voltages at each node and checked the operation of each of the Hall-effect switches. I'm currently giving it a bit of a shake test to see if any problem surface.
It is passes I intent taking it down to WMT and re-installing in the stage, this afternoon, to see how it performs in situ.
It is passes I intent taking it down to WMT and re-installing in the stage, this afternoon, to see how it performs in situ.