Hi everybody,

could you please help me in understanging this text regarding how leakage is compensated :
In valve controlled system, both chambers of cylinder keep a static pressure balance due to the leakage and negative overlap of servo valve in its neutral position.
To the ideal and symmetric servo valve, when in neutral position, it will make the pressure sum of both chambers equal to the setting pressure of relief valve. It is this characteristic of servo valve, that can compensate the flow deviation caused by leakage and compression of
oil. When the hydraulic cylinder is accelerated or acted by outer load, in order to keep this constant pressure sum, the both pressures will change toward opposite direction with the same amplitude. This pre-load of cylinder also increases the system's natural frequency
and stiffness and makes the system possible to be operated in four quadrants. So when the position of cylinder is controlled closed loop with speed variable pump, the system also should be preloaded like valve controlled system

Your statements are quite confusing, and broad... Honestly I have no actual experience with modern servo and will just response a little bit ...

IMO servo valves has almost zero overlap and if the clearance fits between the spool and the housing/sleeve is almost impassable by any viscous fluid, then internal leakage to either A/B ports is not possible. The only problem that may occur is the possible expansion of trapped fluid on A/B ports due to temperature rise(at stall) which can be easily rectified by adding components.

If a leakage can occur on the A/B, then you have to analyze the effect on what actuator you have( differential cylinder, equal area, or motor?). If there is really a substantial leakage, then probably your servo is defective. But any internal leakage(very small amout= 1 drop/min ), can be compensated by imposing more inlet pressure(maintaining the pressure drop).

A close loop servo system can do a lot of job... think about it...


I just want to respond to this quite complicated thread((Bud T said, we will try to answer all topics , but this is hard witout knowledge or book... you or some guys can do the rest.

BTW, any effect on the natural frequency of the system due to internal leakage are probably pre-designed(if there is).

This message has been edited. Last edited by: maglub,

Thanks for the reply,
It is more related to the internal leakage compensation cross the differential cylinder. It says that in servohydraulic systems, where servovalve is used to move a differential cylinder, the sum of pressure upstream and downstream of the cylinder is constant and equal to the supply pressure. AND because of this property, the internal leakage cross the cylinder can be compensated. (my question is that how this property can compensate the leakage by itself?).

Hi,
At first you should understand the available spool
configurations in On/Off, Proportional and Servo Valves.
Have a look at the attached file.

OVER-ZERO-UNDER-LAP.pdf (175 KB, 79 downloads)

quote:

Originally posted by Nahum Goldenberg:
Hi,
At first you should understand the available spool
configurations in On/Off, Proportional and Servo Valves.
Have a look at the attached file.

Isn't there also an another aspect/difference in spool configuration/design depending on if you have a Constant Flow, CF, system versus an Constant Pressure Unloaded (or Load Sensing,LS), CPU...

I know this is more applicable on mobile hydraulic applications where you use multiple, simultaneously and precicely hand operated functions, like an excavator boom/stick, knuckle boom cranes etc, but still......

In a CF system you have an open center valve, where the A and B ports are closed....operating pressure is built up by closing the center and at the same time opening the A and B ports....full pressure and flow available when center is fully closed....so here P-T, P-A, B-T can be more or less open/connected at the same time

In a CPU system you have an slightly open center and closed A and B ports...the center is a restricted (by a 0.035" orifice) compensator-signal line...Before connecting P to A, pump need to be pressure loaded by blocking the signal line thru the center, that means the center have to be fully closed before opening P to A and B to T...So here P-T and P-A can not be open at the same time

BTW... both these system can be setup with basically the same directional valve body....just some minor valve body modifications and different spool configurations/design/overlap what ever...
Check it out at Parker....

I hear terms like overlap, underlap, zero lap, positive overlap, negative overlap, no overlap etc...Is there an official "lap"-nomenclature used by professional "spool valve" people like spoolvalve design engineers etc????

I like to stick with Bosch terminology. There is Positive Overlap, Zero Overlap and Negative overlap.
Positive overlap results in a deadband in the neutral position and is found in open loop (control aspect) proportional valves.
Closed loop (control aspect - meaning with position feedback) have a slightly negative overlap but is often refered to as zero overlap for simplicity and marketing. The quality of the nuetral overlap effects the gain of the valve. Ideally a 2% input signal will result in a pressure gain at the A or B port of 80%. Too much overlap creates a deadband decreasing gain, too little overlap wastes energy through leakage and provides too much gain.
The fouth position on a directly actuated proportional solonoid valve (which the spring naturally wants to offset to with no soloid force) also functions as a fale safe, for if the valve lost control power with a negative overlap in neutral, a differential cylinder will become a desplacement cylinder and start extending with no feedback compensation. On a valve that has 2 stages, the main stage has only 3 positions, with no fail safe. The neutral position is spring centered if the feedback loop is interupted, so the cylinder will start extending, usually requiring the addition of an isolation valve to block flow in the event that control power is lost (Bosch, and I would assume others slightly offsets the mechanical null of the main stage on two stage valves so the cylinder can either slowly extend or retract on power loss).

All of this applies to Bosch proportional servos, or MOOG Direct Drive (DDV) or whatever manufacturer uses a similar technology; ATOS, Vickers etc..., if you get into older types of valves that have a flapper/nossle or jet pipe pilot valve, you can offset the null of the pilot valve(mechanical) so that in the event of control power loss, the cylinder will either slowly retract or slowly extend, whichever way you choose to offset the null.

This is a huge subject with lots of variables. I'm sure it must have been addressed at some point in time in the Forum (I'm pretty new) especially with Peter around. (Of course I guess you won't find those past discussions now unfortunately).

quote:

This is a huge subject with lots of variables. I'm sure it must have been addressed at some point in time in the Forum (I'm pretty new) especially with Peter around. (Of course I guess you won't find those past discussions now unfortunately).

Sad to say Doug, Servo and Proportional valves have had little traffic on the Forum's over the years I've frequented them. Until Peter started posting 2 or 3 years back there were almost no questions or answers with anything about Motion Control. He and Jack Johnson are the main ones that reply to any post's concerniing Motion Control.

My thoughts are, Motion Control is applied to less than 2% of Fluid Power circuits, at least in my small territory in Southwestern Indiana, Western Kentucky an Southeastern Illinois. Im referring to Industrial applications since that has been 99% of my sales/work. Some local Smelting amd Rolling facilities have more Motion Control circuits than all the rest of that territory by far.

All the types of Motion Control valves you mention are in these facilities since one has been operating since the early 60's.

My question is about that part of the sentence that says in an ideal and symmetric servo valve ( means no laekage in servovalve), the pressure sum of both chamber is equal to the setting pressure of relief valve.When there is no load on cylinder, pressure at each chamber of symmetric cylinder would be equal to the half of the supply pressure. And also when there is load on cylinder,sum of pressure at each chamber would be equall to supply pressure, but pressure at each chamber would be half of the supply pressre plus ( at uppstream) or minus (downstream) half of the load pressure.The statement says that if we consider that there is an internal leakage in cylinder, this sum pressure property can compensate the flow deviation caused by leakage. My question is that, how this property ( sum of pressures) can do that?
thanks

In order for a cylinder to hold a position, there needs to be a force/pressure balance. In a symetrical cylinder with no load in which both retract and extend surface areas are equal (double ended rod) there needs to be a pressure/force balance. With no external force, the only balance is internal force or pressure. So the ideal "NO-LOAD EQUILIBRIUM POINT" in this case would be the exact null position of the valve, which results in a pressure differential of the two work ports to be 0. 0 PSI diff and 0 force imbalance.

In a differntial cylinder, the NO-LOAD EQUILIBRIUM POINT is off it's NULL position because a net pressure differential or ratio is required to match the piston surface area ratio. Now as a load is applied in either direction, this Equilibrium Point keeps changing.

As the postion of the valve spool changes, the pressure differential changes according to the gain of the valve. In an ideal symetrical valve, the two ports pressure gains are inversely proportional to each other (in theory).

As a load is applied, the cylinder enters an imbalance of force and starts to move, this results in an increase of pressure on one side of the cyilnder and a decrease on the other side which moves the equilibrium point. The servo mechanism, be it electrical soloinoid with cylinder position feed back (ie. LDT), or a traditional servo with mechnical feed back, positions the spool to the new EP.

So I think maybe what there getting at is if there were leakge in a cylinder, the change in pressure differntial will result in cylinder movement, in which the resulting error signal from the feed back would compensate for the leakage by readjusting the equilibrium point of the valve and restoring the required pressure differnential across the cylinder.

I might not have all of this right, but without Jacks book, I wouldn't have any of it right. LOL.

Amir, where is the text that you are refering to? Where can I find it?