Question:
What has to be considered for the use of stretch factor and gearbox factor in function block DFSET?
Answer:
If in function block DFSET as well the stretch factor (VP-DIV / C533) as the gearbox factor (RAT-DIV / C33) is used for signal matching, the following is to be considered:
- The temporary result according to stretch factor (1st factor) is internally limited to ± 32767, before the gearbox factor (2nd factor) is included.
- After consideration of the gearbox factor an internal limitation to ± 32767 is also executed, which corresponds to the 16 Bit format of the analogue output value.
- In case of stretch factor > 1 (big numerator and small denominator) the internal limitation can be reached, which influences the output value correspondingly. The higher the stretch factor the more probable it the reaching of the format limitation.
Calculation in DFSET:
1. Stretch factor result = (DFSET-IN in inc/ms) x VP-DIV / C0533
2. Limitation of stretch factor result to ± 32767
3. Final result = Stretch factor result x RAT-DIV / C0033
4. Limitation of final result to ± 32767
For the practical use:
- The required numerator and denominator values are to be distributed to stretch and gearbox factor in a way, that neither the final result nor the temporary result is limited.
- With regard to an optimal drive behaviour the stretch factor (1st factor) should be selected in a way that the value range of ± 32767 for the stretch factor result is completely exhausted, however, without reaching the limitation in the 'worst case'.
- In case of stretch factors of > 1 it is to be examined, if the limitation can be reached, see below.
- The final result is processed as output value in the 16 Bit format, therefore the following is applied: The final result of stretch factor and gearbox factor is not supposed to be ± 32767. Otherwise a limitation of the output value must be accepted.
Examination, if in case of stretch factor (1st factor) the internal limitation can be reached:
- The maximum value at the phase difference input 'DFSET-IN' in the internal format (inc/ms) should be known.
- The code digit C0539 shows the input value for DFSET-IN, however in rpm. A conversion is necessary, see 'Master frequency, standardisation and conversion rpm in inc' under article 20030003.
Example:
Max. speed: 4000 rpm = 4369 inc/ms
Stretch factor: 245 / 4
Gearbox factor: 240 / 3120
Unfavourable arrangement of the factors:
1. Stretch factor result = 4369 x 245 / 4 = 267601 (Value too high!)
2. Limitation to 32676
3. Final result = 32767 x 240 / 3120 =
2520 (Wrong result)
Correct arrangement (denominators swapped):
1. Stretch factor result = 4369 x 245 / 3120 = 343
2. (No limitation)
3. Final result = 343 x 240 / 4 =
20580 (Value ok)
For comparison the result in 'Floating-Point' format (pocket calculator):
20584.71 inc/ms
Tip for 'Residual processing' in the DFSET block:
The final result deviates a little bit from the exact value given by the pocket calculator, which does not mean, however, that the master frequency processing in the DFSET is inexact! The calculations for stretch and gearbox factor are executed with a so-called 'Residual value processing '. This ensures that - after several calculation cycles (ms) - the exact result given by the pocket calculator will be reached.
The optimal selection of the stretch factor (see above) lowers the temporary deviations to the exact result and thus leads to a further optimisation of the drive behaviour.
Remember: If no limitation is reached, even in case of non-optimal selection of the stretch factor, the exact value will be obtained via several calculation cycles (no incremental loss).
