Sauer 90 Series Hydrostatic Drive Control

[Hydguy100]

Good day everyone, I am working on a system that has 2 Sauer 90S hydrostatic pumps on board for track drive control. Both pumps are EDC controlled. I have pressure transducers on both ports of the pumps for feedback. I am controlling the pumps with Canbus Joysticks to a mobile PLC and from there providing a PWM signal with current feedback to drive the EDC's. My current feedback shows I am driving from 14ma to 85ma, which is within spec for the pump. I have drive ramps set up at 750ms from zero to full stroke from the PLC. Both pumps have the control orfices removed from the controllers because we need extremely fast actuation when the pumps are switched from drive function to drill function.

Problem I am having is that it is difficult to make small adjustments in drive control, lets say the operator is driving along and wants to climb a hill. Because we have ice and snow out here he has to make sure he doesnt spin the tracks in the snow or he may run the risk of sliding down the hill or flipping the unit. Yes the hills are that steep. So typically on the fully hydraulically driven units the operators watch the gauges and increment them up or down by moving the joystick slightly, this will increment around 100 psi to each motor and they can thereby control the output torque of the drive motors and prevent spinning.

On my system, the system is less easy to control, the pressure jumps by 500 psi with a small input from the joystick. I am thinking that if I am able to use torque (pressure) control on the unit versus using displacement (flow) control that the pumps will be easier to control.

Any thoughts out there on using torque control for driving a vehicle versus flow control from the pumps?

Attached is a link to a Youtube video showing me driving the unit.

Thanks for any help you can give me.

https://www.youtube.com/watch?v=-MtwusKufFY&feature=plcp

 

[Oldhydroman]

I recall a good few years ago working on a rail vehicle with a full hydrostatic transmission: variable displacement, over-center pumps mounted to the diesel engine (and running at a constant speed) coupled to fixed displacement motors driving the wheels via a set of three speed transaxles. There were about four pumps running in parallel with a common signal being sent to the driver cards on the individual closed loop displacement controls (proportional valves controlling pump displacement directly with LVDT feedback of swash angle). Anyway, the machine manufacturer had put some pressure gauges on the pumps and used micro-bore gauge hoses to connect to the gauges in the cab which were about 50 yards distant. The gauges were next to useless for showing anything meaningful because the movement of the needle was so sluggish. So we put some pressure transducers on the pump test points and some simple digital displays in the cab. The machine manufacturer went berserk – claiming that we had significantly altered the way the machine worked because as soon as they touched the joystick the pressures went up hugely. It was b*lls of course – what was happening was that, for the first time ever, the manufacturer could actually see how his own machine worked.

You have the benefit of seeing from the outset what your drive pressures are and the results shouldn’t be surprising. So let’s imagine that top speed on your machine is 10 mph and you have the ramp times set to 0.75 seconds from zero to max. Well that’s an acceleration rate equivalent to zero to 60 mph in 4.5 seconds – that’s quite some sports car you’ve got there. And what do those cars have? lots of soft rubber in contact with a dry road. Ask your sports car to accelerate like that on ice and all you get is wheel spin. So your problem is all to do with the track drive sprockets accelerating faster than the vehicle itself can accelerate. As soon as you increase stroke on the pump (even a fraction) then you will be pushing out more oil that the drive motors are absorbing and the result is a sudden and substantial increase in pressure. Using a pure displacement control on the pump means that as soon as you increase the stroke there will be a mismatch of flows.

You’re on the right lines thinking about a torque control (pressure control) but it’s not quite that simple. Consider what happens when you drive a car; there is no direct relationship between the position of the gas pedal and the speed of the vehicle. When you press harder on the pedal you let more gas into the engine, there’s a bigger explosion in each cylinder, each piston pushes down that much harder and the torque output of the engine increases. This additional torque is delivered to the wheels via the gearbox and drive shafts – and creates an acceleration of the vehicle. As long as you don’t get any wheel spin (or clutch slip) there remains a direct relationship between the engine speed and the vehicle speed. The acceleration tails off though, as the vehicle speed increases the engine speed increases with it and the amount of gas pouring into each pot PER STROKE goes down. The torque output goes down and the acceleration comes to an end.

If you apply a simple pressure control to the pump rather than displacement control then you will be controlling the acceleration directly and it won’t tail off as the speed increases. You will find that the vehicle control will be very jerky and it will be difficult to maintain a constant speed.

I think you might find some mileage in trying a POWER control algorithm. That’s actually what you get with the car – when you increase the rate of gas delivery to the engine (push further on the gas pedal) you are turning up the POWER. When you first press on the pedal the engine speed hasn’t yet changed but the torque output increases and the vehicle starts to speed up. As the engine speed increases (because the vehicle speed is increasing) the engine torque output reduces until you reach a balance point when the amount of gas you are supplying to the engine creates a torque sufficient to drive the vehicle at the speed that causes the engine to run at that particular rpm for which that amount fuel gives you the torque you need to run at that speed (I may have repeated myself there).

So if you have some control over the pressure cut-off or pressure compensator override on your pump then you can relate the “allowable” pressure both to the joystick position and also to the measured pump displacement. That way, when you push the joystick over you get a small increase in allowable pressure as well as a small increase in displacement. But as soon as you hit the pressure limit the displacement control will be overridden and the flow will be forced to match the vehicle speed. As the vehicle accelerates the pump will increase displacement to match the new speed. As the displacement increases the maximum allowable pressure will decrease and the acceleration will tail off. Note that when going downhill you need less drive torque so using a power control would allow the vehicle to speed up – but most drivers will be used to this characteristic because that’s what most automotive transmissions do already.

It’s up to you to determine the relationship between joystick position, displacement and allowable pressure (that pressure being a function of both joystick position and pump displacement). You could even have different relationships for different operating modes (hi range, low range, road travel, off road, inching etc.) and/or you could set different parameters for different traction conditions (ice, loose snow, mud, wet grass, firm ground etc.).

None of this actually solves the loss of traction problem – that’s all to do with how much track you have in contact with the ground, the profile of the tracks, the weight of the vehicle and the nature of the ground. If, however, you can make the vehicle a little more “drivable” then the operator stands a chance of intuitively making the adjustments needed to work the machine.

DOL

 

[Hydguy100]

Wow, that is a great response, thank you so much for putting so much thought into it Oldhydroman. When I have more time tonight, I will go over it in depth, programming another system right now.

Once again thank you, Oldhydroman.

 

[hydromech]

I'm guessing that oldhydroman is a very experienced hydraulic systems engineer who has retired from the game.

I'd love to have the time to write such detailed and lengthy responses...but there is never enough time. He puts us all to shame...

Having seen the video you posted, I assume that carrying out such changes are not quite straight forward...is it not easier to adjust the gain on the joystick. Again, the video shows that the vehicle has some decent control hardware, is that not an easier option?

 

[Oldhydroman]

Hi Hydromech

Not quite in my dotage yet Still hope to get in another 15 years before I have to give in. I was just typing away last night while the wife was watching some mindless pap on the TV.

DOL

 

[hydromech]

When I first started "helping" on this forum, I'd give chapter and verse. Always looking to help out, but too many people wade in with unhelpful and contradicting "advice" and there is never enough time. I've had to scale it back...

Anyways...keep up the good work!!!