We can improve the circuit by using four OP471 [high-speed quad opamp]. IC1 for precision diodes with voltage follower for impedance matching. IC2 for two 2nd order low pass filters. IC3 for the RC oscillator based on the distance between two differential coils and the feedback And IC4 for adding the two signal and diode capacitor envelopdetecton followed by difference amplifier to get constant output. It will be a closed loop system. But in modern time, we cam condense all of that in 1 small thing like PSoC5 or SLN47004. Cheers!!
This seems like the nerdeist way of having a position indicater for a vacuum variable capacitor tuned mag loop then you could have so.e code running on a pie to correlat voltage with frequency and use that as a way to punch in numbers to almost instantly tune to a given frequency obviously you would have to do some set up tests say every 10kc on each band note the voltages assign them a value or even just straight frequency and have it so you can just type in a frequency when in use and it will go straight there
Yes, I did something very similar with an open loop stepper control system a while back. STM32 controlled the pos and stored a calibration lookup table in memory. The only issue with vacuum variable caps is that most "modern" ones' shafts don't extend when unscrewed. So you're left to track motor rotations instead of measuring linear displacement. I have a trick up my sleeve to do it optically. Maybe it'll make a future episode. -Derek
If you’re interested in very small displacements, I guess you’d want the two sense coils to be very narrow. I wonder what kind of positional resolution is possible with a DIY LVDT? I’ve also wondered if machining the moving slug to have thinner disks on the ends rather than it being a constant-diameter cylinder would increase resolution, as the flux would be more concentrated? You’d have a much shorter linear region, but if you were just concerned with small displacements, that would the limiting issue.
Decades ago a couple of guys from Tektronix set up shop in Newport OR and set the standard for precision measurement of paper mill belt wear. The company was called Emvico and at the heart of the machine was a precision LVDT in a heavy cast stainless shell with a microstylus capable of micro displacement measurement. The belts for those machines cost $100k and so the mills want to know when to replace them at the last minute. To do this they measure the roughness of the paper surface using these machines. To this day, 50 years or so on, those machines still run and are still the standard for taking that measurement mostly in toilet paper manufacture. So the next time you wipe up, thank Emvico and LVDT's!
We can improve the circuit by using four OP471 [high-speed quad opamp]. IC1 for precision diodes with voltage follower for impedance matching. IC2 for two 2nd order low pass filters. IC3 for the RC oscillator based on the distance between two differential coils and the feedback And IC4 for adding the two signal and diode capacitor envelopdetecton followed by difference amplifier to get constant output.
It will be a closed loop system.
But in modern time, we cam condense all of that in 1 small thing like PSoC5 or SLN47004.
Cheers!!
Indeed a lot can be done to improve the circuit.. thank you for the suggestions; maybe on a revisit episode. -Derek
I've used LVDT's in manufacturing and I've always been curious about how they worked. Thanks!
Thank you! Glad I could help someone -Derek
This seems like the nerdeist way of having a position indicater for a vacuum variable capacitor tuned mag loop then you could have so.e code running on a pie to correlat voltage with frequency and use that as a way to punch in numbers to almost instantly tune to a given frequency obviously you would have to do some set up tests say every 10kc on each band note the voltages assign them a value or even just straight frequency and have it so you can just type in a frequency when in use and it will go straight there
Yes, I did something very similar with an open loop stepper control system a while back. STM32 controlled the pos and stored a calibration lookup table in memory. The only issue with vacuum variable caps is that most "modern" ones' shafts don't extend when unscrewed. So you're left to track motor rotations instead of measuring linear displacement. I have a trick up my sleeve to do it optically. Maybe it'll make a future episode. -Derek
Ok!! (Puts my hard hat on)
How did you decide on the number of turns for the coils?
Cool project! In my opinion, winding the wire is the most enjoyable part! You can also try making a fluxgate magnetometer, which is similar in design.
Ooh interesting. I'll have to give that circuit a look. Thanks! -Derek
wearing that shirt is a good way to get clicks 👍
Let's see if it works! -Derek
What if you sum these 2 signals?
If you’re interested in very small displacements, I guess you’d want the two sense coils to be very narrow. I wonder what kind of positional resolution is possible with a DIY LVDT?
I’ve also wondered if machining the moving slug to have thinner disks on the ends rather than it being a constant-diameter cylinder would increase resolution, as the flux would be more concentrated? You’d have a much shorter linear region, but if you were just concerned with small displacements, that would the limiting issue.
Decades ago a couple of guys from Tektronix set up shop in Newport OR and set the standard for precision measurement of paper mill belt wear. The company was called Emvico and at the heart of the machine was a precision LVDT in a heavy cast stainless shell with a microstylus capable of micro displacement measurement. The belts for those machines cost $100k and so the mills want to know when to replace them at the last minute. To do this they measure the roughness of the paper surface using these machines. To this day, 50 years or so on, those machines still run and are still the standard for taking that measurement mostly in toilet paper manufacture. So the next time you wipe up, thank Emvico and LVDT's!
hi, excellent video. could you check if you use 1k or 10k in the actual circuit?
Low pas filter after precision rectifier ?
where can i download ltspice ?
Excellent!
Thank you sir!!!