Alternator control problem

Greetings all,
I'm an engineer on a tugboat. Mechanically minded more than electronically. A long while ago I posted the following info on my problem on another NG
with no response. Wrong group or people just unable to assist, I'm not sure. Hopefully somebody here may be able to help.
Our system: A 50 KW alternator driven by an hydraulic motor in turn driven by an hydraulic pump directly driven by a Caterpillar 3512 diesel engine (used in tug propulsion) on a PTO. Caterpillar's speed is variable 650 to 1800 rpm. Hence the speed of the pump varies. Hydraulic motor is regulated to maintain alternator at 1500 rpm (50 hz).
Problem: The controller currently in use is more than twelve years old. The design of it was ahead of its time when comissioned. But is an electronic nightmare now. Two printed circuit boards. A driver board and an options board. Two different electronic technicians have looked at it and are unable to rectify the system. It will not maintain 1500 rpm. Each board has its own set of adjustables. Gains, P Gains, I Gains, Balance, Zeroes etc etc. To my mind just too many adjustments.
The way it is set up: Three inputs to the controller. 1) Speed sensor off alternator shaft. 2) Hydraulic pump swash plate position feed back. [swash plate governing the flow of oil to the motor] 3) Desired set point for speed. One output: Voltage to servo valve to position the swashplate of the hydraulic pump. Controller powered by 24 volt DC.
My electronics knowledge is not so crash hot. But I'm sure there must be in this day and age a simple process controller that can handle the task and is easy to set up.
Can anyone suggest anything please? I'll furnish more info if required. Thanks, PT.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Greetings all,
I'm an engineer on a tugboat. Mechanically minded more than electronically. A long while ago I posted the following info on my problem on another NG with no response. Wrong group or people just unable to assist, I'm not sure. Hopefully somebody here may be able to help.
Our system: A 50 KW alternator driven by an hydraulic motor in turn driven by an hydraulic pump directly driven by a Caterpillar 3512 diesel engine (used in tug propulsion) on a PTO. Caterpillar's speed is variable 650 to 1800 rpm. Hence the speed of the pump varies. Hydraulic motor is regulated to maintain alternator at 1500 rpm (50 hz).
Problem: The controller currently in use is more than twelve years old. The design of it was ahead of its time when comissioned. But is an electronic nightmare now. Two printed circuit boards. A driver board and an options board. Two different electronic technicians have looked at it and are unable to rectify the system. It will not maintain 1500 rpm. Each board has its own set of adjustables. Gains, P Gains, I Gains, Balance, Zeroes etc etc. To my mind just too many adjustments.
The way it is set up: Three inputs to the controller. 1) Speed sensor off alternator shaft. 2) Hydraulic pump swash plate position feed back. [swash plate governing the flow of oil to the motor] 3) Desired set point for speed. One output: Voltage to servo valve to position the swashplate of the hydraulic pump. Controller powered by 24 volt DC.
My electronics knowledge is not so crash hot. But I'm sure there must be in this day and age a simple process controller that can handle the task and is easy to set up.
Can anyone suggest anything please? I'll furnish more info if required. Thanks, PT.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Wow, sounds like a very cool electrohyrdraulic servosystem. I've worked on something similar in graduate school about 20 years ago.
Let me make sure I have this right. We're controlling the displacement of a hydraulic pump to control the speed of a hydraulic motor at 1500 RPM so that the generator keeps making 50 Hz AC power regardless of the speed of the tug's engine.
I assume there's a small rotory actuator being controlled by the servovalve to actually move the swashplate. If I remember correctly, the stubby little shaft that moves the swashplate is called a 'pintle'.
I suspect the reason you have two boards is that you have two nested control loops. An inner loop servos on swashplate postion (why else would you be measuring it?) and the outer loop is a speed control.
Of course, this doesn't help solve the problem. What are the symptoms? Does it hunt? Does it hold steady at the wrong speed? Does it just max out? IF you can isolate which controller (inner or outer) is screwing up, it'll help.
PT wrote:

Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

electronically.
sure.
old.
an
board
etc
in
is
Woodward make a range of alternators controllers for land & sea, both digital and analog, one of which I'm sure will solve your problem: www.woodward.com
Have a look at their analog controllers. The Woodward gear is not cheap, but in your situation I'm sure that reliability is a bigger issue...
Good luck!
Cameron:-)
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Cameron Dorrough wrote:

Thanks Cameron, Have sent them a copy of my posting.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
PT wrote:

The two sets of PID adjustments may be excessive, but it may be necessary excess -- read on.
Given that the thing is monitoring the swash plate position the second set of adjustments indicates that you have a two-loop system -- there'll be an inner loop that sets the swash plate with a PID controller, then an outer loop that sets the alternator speed by commanding the inner loop with another PID controller. This sort of thing is good when your system needs to react quickly and accurately to disturbances. If your hydraulic system uses a positive-displacement pump and motor then the ratio between the diesel speed and the alternator speed will be pretty firmly set by the swashplate system, so the overall controller will have to react promptly to diesel speed changes if you want to maintain a good 50Hz alternator output.
The way to tune such a system is to break the loop between the inner and outer loops -- in this case it means figuring out which controller is for the outer, disconnecting it's swash plate command and connecting your own setpoint in there temporarily. Then tune up the inner loop until its performance is satisfactory. Finally you'd reconnect the inner and outer loops and tune up the outer loop. Trying to do this in reverse, or trying to tune up the inner loop with the outer loop still active is a royal pain without the right instruments and a computer standing by.
You didn't give enough information to tell, but I can see four possibilities of what your problem might be: One, your controller is just plain broken. In this case you want to repair or replace the thing and find the original settings. Two, someone has diddled the settings in an attempt to improve performance. Given the two-loop nature of the thing this could be a bear to fix without separating the loops and tuning the inner one first. Three, something in your mechanical system has worn or broken. Four, you've changed a hydraulic motor, pump or the alternator and haven't told us.
If it's the third or fourth case then either your system is now just plain uncontrollable (particularly if there is a lot of backlash or stiction/friction in your swashplate) or it needs markedly different tuning parameters to work correctly. If it's just plain uncontrollable then going out and buying the World's Most Advanced Controller won't do you any good. If it is worn and needs markedly different tuning parameters to work you could conceivably tune the thing up, but if your mechanicals are at the end of life then you'll have to do it all over again when things wear some more (or just plain break). OTOH, if it has had a part replacement and is otherwise controllable tuning it up to the new part would be a good idea.
--

Tim Wescott
Wescott Design Services
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Tim Wescott wrote:

I believe you might have hit the nail on the head Tim. In this particular case, our driver card being the inner loop, And the options card being the outer loop.

This has been tried. The options card having been removed and the driver card being tuned, Then replacing the options card and tuning it in.

The cards have been inspected and tested. We believe all components operating satisfactorily. We even have spare cards which have been tuned in but does not rectify our problem.

Well yes. The settings have been altered. But only after the initial problem. Hydraulic system hunting erratically and the alternator tripping off the switchboard.

One technician that has looked at the system tends to follow this line. Myself, the other engineer that works on the tug, and the other technician working on the controller do not necessarily agree with him. Our reasons being that we have been able to tune the controller. And the system has and does operate satifactorily. But the fault comes in intermittently. Also we need to conclusively prove the fault is not electrical before we look at the hydraulics. In our instance we are looking at a five figure sum in $'s to remove the pump/motor and send to the maker for inspection/repair/overhaul. NB. We had a play with it yesterday and had it set up okay. I've just now returned from doing a tug job where I ran the hydraulic alternator. It tripped out when full speed rapidly put on main engine. On two occasions. Each time I was able to reset it and put the alternator back on the board. Maybe this time we're getting ahead and a little more fine tuning needed.

Negative on all three counts.

I can accept this

Scary. Seeing as the system has less than 6000 hours running time.

Many thanks for your reply Tim. As you can see, half way through my response, we had to do a shipping job. And I've just returned to finish off. We think that we're getting close to having it right. Still it would be nice to put a finger on what our original problem was. I'd like to add that we regularly send oil samples away for analysis. A long time ago we had a sample come back with high copper content. This may have indicated wear. But we should have been able to adjust for that. I've past since changed the oil and filters and after more hours running, I've just received another analysis back indicating all satisfactory.
Regards,
TR aka PT when using computer at work.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
TR wrote:

-snip-
Well, it sounds like you've done good basic fault isolation and ruled those out, then.

When you say the hydraulic system hunts erratically -- does it do this all the time, only after the engine has changed speed, only when the engine is changing speed, or completely unpredictably? If it's one of the first two then chances are your tuning isn't right. If it happens during speed changes it's because the control system can't keep up with the disturbance input, and if it's completely unpredictable then I'd vote for some intermittent fault, either mechanical, hydraulic or electrical.
If it's from the control system not keeping up it could either be because you don't have sufficient bandwidth in the controller to move the swash plate control fast enough, or that the servo motor that moves the swash plate control just doesn't have enough oomph to keep up with the engine as it revs.
If it's an issue of the motor not being fast enough you'll have to address that. If it's because the loop doesn't have enough bandwidth you may want to check to see if one or both of your controllers has a feedforward input, and if you can feed it with either the engine throttle position or the engine RPM. Essentially what you'll be doing is feeding the system a guess of how much it has to move, and only asking the feedback to clean up any little messes left behind. If the engine takes significantly longer to respond than the inner swashplate servo loop then you probably want to use the engine RPM, otherwise you'd want to use throttle position.

I assume you've looked at the servo motor that drives the swashplate, and made sure that it doesn't have any faults, either mechanical or electrical? If you can get to it checking both the motor and swashplate for mechanical bind would at least eliminate another possibility.

OK, I'm just not calibrated to tugboats.
> We had a play with it yesterday and had it set up okay. I've just now

"tripped out when full speed rapidly put on main engine" -- that's what's making me think of feedforward, and of seeing if the servo motor that drives the swashplate position isn't just too dang slow.

You said the problem is intermittent -- has it always been there, or has it developed over time? Do you have a new tugboat driver* who likes to work the throttles way more than the last guy? Have you installed a new control panel that may be tripping the breaker more readily than the old one? In other words, could this be a design problem with the system, and not a broken part?
* I have no idea of the correct term, unless it's "pilot".
--

Tim Wescott
Wescott Design Services
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
PT wrote:

Before you throw this system overboard, are you sure it cannot be made to work just fine if only you understood how to adjust it correctly? Do you have any documentation on this system? If approached systematically, it doesn't sound so hard to adjust (assuming all its parts still work).
--
John Popelish

Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
John Popelish wrote:

I'm beginning to believe John we are getting ahead and we don't have to toss the old system. The documention we have was very difficult to follow. Especially as the manuals had to be kept updated to all the revised versions of pcb's. Even then some of the technicians don't believe the procedures are exactly correct in all the manuals.
TR. aka PT on my work comp.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
TR wrote:

I would like very much to take a look at the documentation, but it sounds like it would be hard to scan and email a readable copy.
--
John Popelish

Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Polytechforum.com is a website by engineers for engineers. It is not affiliated with any of manufacturers or vendors discussed here. All logos and trade names are the property of their respective owners.