drbitboy
Lifetime Supporting Member
[I am turning into a dog with a bone on this one ...]
Another reason for buying a -VHSC card is that someone else should have figured out all of the pitfalls.
Analogous to the noise/wobble in the Freq.-to-Volt. circuit described earlier will be all sorts of issues if you try to count (de-rated) pulses off that pulser disk with a prox and try to calculate speed (i.e. pulse frequency) in PLC code; PLCs are not precision instruments. Unless the code can timestamp edges with interrupts, the pulses are controlled by the IO task. Even with a scheduled task, the code only knows the edge count since the last task to +/-1 edge, so it will have to make a tradeoff between measurement rate and accuracy. I don't know what the scan times are but with 5ms as a guess that is 100edges/s max and at least 1% = 35rpm+ noise at a 1Hz measurement rate. And time is another issue: of course there is the GSV of CST CurrentValue with μs resolution, but the code can only sample that in the task, not at the edge.
And on it goes. Unless the customer will be either satisfied with 1% noise or a longer measurement sample time, or is impressed by the use of duct tape over money, the cost of a -VHSC card will be far cheaper than the time it takes to do almost anything else.
Another reason for buying a -VHSC card is that someone else should have figured out all of the pitfalls.
Analogous to the noise/wobble in the Freq.-to-Volt. circuit described earlier will be all sorts of issues if you try to count (de-rated) pulses off that pulser disk with a prox and try to calculate speed (i.e. pulse frequency) in PLC code; PLCs are not precision instruments. Unless the code can timestamp edges with interrupts, the pulses are controlled by the IO task. Even with a scheduled task, the code only knows the edge count since the last task to +/-1 edge, so it will have to make a tradeoff between measurement rate and accuracy. I don't know what the scan times are but with 5ms as a guess that is 100edges/s max and at least 1% = 35rpm+ noise at a 1Hz measurement rate. And time is another issue: of course there is the GSV of CST CurrentValue with μs resolution, but the code can only sample that in the task, not at the edge.
And on it goes. Unless the customer will be either satisfied with 1% noise or a longer measurement sample time, or is impressed by the use of duct tape over money, the cost of a -VHSC card will be far cheaper than the time it takes to do almost anything else.