Dither Amplitude and frequency for EDC valve

[Jpshockster]

Any recommendations welcome.

So I have read the API and every other document I could find with no answers to my question.

I'm working with a Linde Hydrostat HPV-02
My software drives the swash plate one direction for forward and the opposite for reverse, using C1p37 for Fwd and C1p38 for Rev.  Pinconfig 5 for both.  I have found the applying Dither 200mA @50hz to the Fwd (C1p37) gives me a very smooth transition from neutral to forward motion creep,  I apply the same values to the reverse (C1p38) and the transition is aggressive/jerky...... so I remove the Dither parameters from the Fwd (C1p37) and only use them on Rev (C1p38) and Reverse now has the very smooth transition from neutral to reverse motion creep. 
Now for the question
Is it ok to manipulate the Dither parameters or should they remain constant? 
I set it up now so that when the vehicle is in neutral there is no Dither applied to either output, once a direction is selected the Dither parameters are applied to the respective output.  I have noticed that a change to Dither parameters on C1p37 affects C1p37 - C1p40 so I have configured the application to only use 37 and 38 and moved my other output devices to C1p42 - C1p45
Has anyone ever dealt with this?
Could this be hard on the controller outputs?
TIA
Justin

[Marbek_Elektronik]

I can speak for Hawe valves:

ReqFreq: default 4000Hz
PinConfig: 5
Maximum current 1100mA
Dither frequenc 100Hz
CurChgLim: default 111
Dither amplitude 100mA

For all valves an all dirctions the same parameters.

[Jpshockster]

Marbek_Elektronik,
What did you do for your Minimum current?  Do you have an issue with the cracking current being higher than the drop out current? 
For my valve I have found that 440mA is the cracking current regardless of Dither parameters.  Once the valve opens I can lower my minimum to 390mA and maintain motion at a slower speed than that 440mA it took to initiate motion. 

The other question I would have for you is.......Do you set your parameters the same in every vehicle or do you have to tweak them per vehicle?

[Marbek_Elektronik]

It was a 24V system.
And it was only 1 vehicle and I adjust the parameters to the vehicle.
I don't know the starting point.

My understanding is:
- if you use no dither, there will be a big hysteresis,
- if the dither frequence is low and the dither current is high, there will be a small hysteresis,
but: maybe the system works with the dither frequence and the life time is not high.

[FluidPowerTom]

Well, I've found that the code will compile if you have logic to switch PWM frequency during operation, but the code will not compile if you have logic to switch the dither frequency.  What I've done with this is instead of running high frequency PWM with dither (pinconfig 5) I just do low frequency PWM without dither (pinconfig 4).  This more or less allows you to change dither frequency within your programming since it derives from the low frequency PWM.  I haven't tried any such logic with dither amplitude though...

[Marbek_Elektronik]

Hi FluidPowerTom,
nobody wants to change dither frequency during the running program.
But if you want to do it , it works: I have placed a NV-memory and changed it and it gives the value to the dither frequenc. It was a test.
You choose one time the correct dither frequency for your valve and then don't change it.

I think high frequency pwm is a good solution, because , it is like constant voltage.
Then add a am-modulation (dither) to change the current.

If you use low fequency, I think, it is very critical.
If the valve has another inductivity (tolerances!) the current will be not the same.
I don't think, it is a good way to do this.
You will have 100% dither, if you do it....

[FluidPowerTom]

Marbek, I'd done this on a control unit that is being used by our shop for testing valves.  We're using one of the old chipset controllers where the outputs are grouped, and the frequency needs to be 4,000 Hz for when they control PVE pilots and 100-200 Hz when they control typical proportional valves or an EDC.  It's a somewhat unique situation, but so far it's working.

[oiltronic]

What did you do for your Minimum current?  Do you have an issue with the cracking current being higher than the drop out current? 
For my valve I have found that 440mA is the cracking current regardless of Dither parameters.  Once the valve opens I can lower my minimum to 390mA and maintain motion at a slower speed than that 440mA it took to initiate motion.

Hey Jpshockster, did you get it all figured out?  I thought one of the points of dither was to get rid of that hysteresis, though I haven't actually had a use for dither anywhere yet.  If I'm reading it correctly, note #3 of the MC050 API docs states that if .PinConfig is 5 or 8 (dither enabled) then the PWM frequency for all pins in that group is forced to 4000Hz, regardless of the ReqFreq for that group of pins.  And I think the Lynde HPV-02 needs either 100Hz PWM or analogue with 35Hz dither.

Did you measure what was being sent to the pump solenoids, or maybe look at the checkpoints for the actual frequency & current?

[Krusty]

Marbek, I'd done this on a control unit that is being used by our shop for testing valves.  We're feeling secure with VyprVPN and using one of the old chipset controllers where the outputs are grouped, and the frequency needs to be 4,000 Hz for when they control PVE pilots and 100-200 Hz when they control typical proportional valves or an EDC.  It's a somewhat unique situation, but so far it's working.

How are you figuring out what the perfect dither frequency is?

[Jpshockster]

What did you do for your Minimum current?  Do you have an issue with the cracking current being higher than the drop out current? 
For my valve I have found that 440mA is the cracking current regardless of Dither parameters.  Once the valve opens I can lower my minimum to 390mA and maintain motion at a slower speed than that 440mA it took to initiate motion.

Hey Jpshockster, did you get it all figured out?  I thought one of the points of dither was to get rid of that hysteresis, though I haven't actually had a use for dither anywhere yet.  If I'm reading it correctly, note #3 of the MC050 API docs states that if .PinConfig is 5 or 8 (dither enabled) then the PWM frequency for all pins in that group is forced to 4000Hz, regardless of the ReqFreq for that group of pins.  And I think the Lynde HPV-02 needs either 100Hz PWM or analogue with 35Hz dither.

Did you measure what was being sent to the pump solenoids, or maybe look at the checkpoints for the actual frequency & current?

Its been a while since Ive been to the forum but after exhaustive testing and tinkering Ive found that you must always double check what a mechanical engineer is doing with the valving.  Im not sure how to best explain this but i will do my best.... The system had a "Small" built in leakage at the neutral position that would return to tank for cooling purposes. Because these are pilot operated the pilot pressures were slightly different and the built in leakage for cooling purposes was causing me all kinds of grief.  It was recommended to use the case drain on the wheel motor to return to tank for cooling.  this theory was implemented on a completely different platform and I have not had the issues with this method.  With this method I have since returned to pin config to 4 and everything is smooth in both directions once you figure out the optimum cracking current and use the Knee profile on the operator input ie: foot pedal to further fine tune the feel of creep mode.  hope this helps.
 keep in mind I am no hydraulics major so my descriptions may be somewhat simplified as it took two mechanical engineers and two sales reps speaking to their engineers to figure this out.  (if you ask a Mechanical engineer, its always the codes fault)  LOL  programmers can fix anything!   

[Jpshockster]

How are you figuring out what the perfect dither frequency is?

Lots of testing! I installed some memory so I could change the Dither freq and amplitude and just began playing and using the scope functionality of service tool the evaluate the results.  I highly recommend the API for your controller as there are only certain freq available and the max amplitude based on your application.   Good luck!