Other Parts Discussed in Thread: FDC2114, , FDC2112, FDC2212
Hello,
I'm working on a replacement solution for a mechanical liquid float / magnetic level probe.
The tank is Aluminum; U shaped about 7-8 high ~ 2.5M with a capacity of 2000 liters.
The tank will carry water or mucky liquid fertilizer. (REF L / ratio)
The for physical protection, the probe is to be immersed in the liquid instead of external wall attachment.
The FDC1004 was chosen for the immunity and noise however the FDC2114 maybe more suitable.
The target shape will most likely have a bend radius in it.
The size, length and environment is directing the "probe" to be composite.
Using the FDC1104EVM with two rapid build probe formats I wanted to get some data point for direction.
The data sheets were a great start as well as the reference TI flex probe.
First Probe Format
Base: HDPE 1' x 8' x 0.063
SHIELD Layer 60mm W copper foil
Sense Layer 40mm W copper foil
Kapton Tape is used for layer insulation.
Connections reflected the TI reference design.
First Results:
The level probe was near or at the 15+pf saturation
The ref were less, near the same but were not equal.
Compensation brought all open air values down.
Hand effect 3" and under, showed several PF response and direct contact shift to overload.
- the majority of the probe length was on the floor.
- streaming measurements were in constant change, connectivity routing was an issue picking up everything.
- fine pitch EVM points with flying leads to a proto board then to the foil. ( yuck)
Second Probe
The TI ref probe format was replicated on a;
2.5' x 5' rotomolded PE container
10mm wide copper foil
13mm wide Kapton tape
same flying lead format (yuck)
The first connection was still up above the 15pf compensation;
My ideal would be LOW maybe 1-2pf without compensation.
The Level Strip was separated until a small amount of compensation was applied.
The Level probe was 3' long.
Empty and full a 12+pf change was achieved.
Hand/body object movement from the outside (backside) had no effect.
The ref probes had some effect.
different distances, shield and insulation material were experimented with. incl30mm, 40mm,
streaming data; noisy .3-.5 pf variation so connectivity to probes an issue.
First Probe (return)
- The 12 wide Level and Reference were split into Level only and Ref only.
- FDC1004 EVM leads shortened to 1'
The Level only probe was only configured.
The 7' flex probe was hung vertical; not in ground contact.
Baseline 4pf
streaming numbers deviate less, more work here! on noise.
REF side is next on test; initial layout REF E was routed to close to the LEVEL.
- have not submerged.
- Fertilizer material will arrive soon.
AT some point what is gained will be transferred to a epoxy composite structure with a new set of dielectric properties.
Any recommendations? for noise? what magnitude of deviations should I see streaming?
Would the FDC2112/2114 offer a better noise to range resolution?
Cheers David
