Proportional Valve Hysteresis
The flow difference at one drive current depending on whether it was reached rising or falling; it comes from friction and remanence, and dither cuts it.
Hysteresis is getting a different flow at the same drive current; what settles the result is whether you came to that current by rising or by falling. When you open the lever slowly and close it back to the same point, the speed of the movement being different is the direct symptom of this. It has two basic sources: the static friction between the spool and the body, and the magnetic remanence in the iron circuit of the solenoid.
How it is measured
The measurement is made by raising the current slowly from zero to the peak value and then bringing it back down just as slowly. The flows obtained on the way up and on the way down draw two separate curves; the largest difference between the two curves is given as a percentage of full scale. Making the measurement slowly matters, otherwise the dynamic lag of the valve gets mixed into what you are measuring. The same curves are also read as the flow-signal curve.
Recording the conditions of the measurement matters as well: when the oil temperature, the supply pressure or the setting of the drive card changes, a different curve comes out of the same valve. That is why, when comparing two products, it is necessary to ask under what conditions the declared value was taken.
Where it comes from and how it is reduced
| Source | How it behaves | What reduces it |
|---|---|---|
| Static friction | The spool sticks at the start, then breaks free suddenly | Dither, clean oil, a suitable surface condition |
| Magnetic remanence | As the current falls, the solenoid force drops with a lag | Material and magnetic circuit design |
| Mechanical play | Part of the command is swallowed on a change of direction | Maintenance of the linkage and the joints |
| Absence of position feedback | The deviation is not measured, so it is not corrected | Closed-loop valve construction |
The solution that largely wipes out the friction-borne part is the dither current; its amplitude and the dither frequency are chosen to suit the valve. If the amplitude stays low the sticking goes on, and if it is opened up too far the load shakes visibly. The logic of setting the dither is explained under a separate heading. On closed-loop valves with spool position feedback the hysteresis typically comes down below the one per cent level, but the cost of the valve rises.
The practical result
On a system with high hysteresis you cannot settle the load exactly where you want it in one go; you make two or three corrections. The operator generally describes this as the lever not holding. On work where the load has to be placed to the millimetre, asking for this value when choosing a valve is more decisive than the rated flow.
Nor is its effect limited to positioning. If a different speed comes out at the same place on the lever every time, the operator stops giving small commands and starts working in coarse movements; that in turn makes the load swing more and lengthens the cycle time.
It does not matter on every job
On work where all that is wanted is for the movement to start and stop and where the load is placed by eye, hysteresis is not noticed; in an application like that the price paid for a closed-loop valve is not repaid. And not every failure to repeat is hysteresis: the system behaving differently when the oil is cold has to do with viscosity, while the movement not starting at all at the same command has to do with the deadband.
What to do in practice
When describing the complaint, note the change of direction separately: if the trouble comes out only when reversing, hysteresis and mechanical play come first, and if it comes out in every direction the deadband is in front. Do not skip the dither setting when commissioning; on many systems this setting on its own reduces the number of corrections needed to a visible degree.
The context in which this term is used: Why should it be commanded by an electronic signal?