Eurotherm 590+DRV Questions

Has anyone out there ever set up a 590+DRV as a helper (load sharing) drive?
I have an old (early 80's) Marathon drive that I'm trying to have the 590 load share with. And before anyone asks, this wasn't my idea. I suspect that I will have to set up the 2 drives to look at a common speed signal and then feed the speed error signal (only one I could find on the Marathon) back in to the 590's current loop. Anyone done something like this before?

This is actually 2 separate machines, both pot controlled with a(just added) pneumatic clutch linking the drive shafts together.

I think you want to look at the torque as well. Otherwise one of the drives could take over and the other one more or less just coast along.
I would contact Parker SSD (formerly Eurotherm).

Simple load sharing (Not ratioed, motors the same HP and voltage, motors physically connected together on a shaft)?
If that is the case, take the speed regulator output signal from the master drive, and feed it to the slave drive as a torque command.

If the motors are not physically connected, you need to be sure to only enable torque mode on the slave when coupled by the process web.

A slightly more complicated (and less responsive) method would be to feed a speed command of something on the order of 10% higher then the master drive to the slave, and then feed the current limit input of the slave drive with the speed regulator output command from the master.

I'd probably go with the simple load sharing when the clutch is engaged (as long as it doesn't slip). Route Tag 356 to an analog output on the master drive, feed that to analog input #2 on the slave (Default for current limit/torque control). Enable "I Demand Isolate" on the slave when the clutch is engaged. I think "I Demand Isolate" is defaulted to digital input C8, but you may have to assign it). When the clutch is disengaged (independent mode), disable "I Demand Isolate", and feed 10VDC to AI2, and run as a normal speed drive.

NOTE WHEN USING THE ABOVE:
Keep the slave drive in RUN mode until the master has reached zero speed. If you give it a stop command together with the master, it will go directly to a speed mode stop, overriding "I Dem Iso", and bad things will happen.

Thank you for the reply's.
I would prefer to keep this as simple as possible, as the shafts are mechanically clutched.
The drawback is the speed error signal appears to be inverted. When the older drive is at minimum speed the feedback signal is at 10 volts, as the drive speeds up it settles down to around 6.5-7.3 volts.
The older drive is NOT a Eurotherm, but a 1982 Marathon.
Is there a way to put logic in DSE lite that will allow me to invert that signal or am I looking at purchasing one of the more sophisticated software packages?
I can't find a zero speed signal on the old drive, what bad things will happen if I stop both drives?

PS: This is my first attempt at programming a Eurotherm DC drive, up until this I have mostly done simple VFD's

This is the way the SSD guys do it. The document shows a 690+ but the same applies for the 590+. I personally would use the method that rdrast described initially. In fact I did on a driven rewind once. Then the SSD guys came in and changed it to the method shown in the PDF, likely because they didn't understand what I was doing and/or the method in the PDF "is the way they've always done it". That isn't a valid engineering justification, by the way.

Moral of the story is if you want SSD support going forward you should probably use the example in the PDF.

Keith

Soggy Canuck said:

Is there a way to put logic in DSE lite that will allow me to invert that signal or am I looking at purchasing one of the more sophisticated software packages?
I can't find a zero speed signal on the old drive, what bad things will happen if I stop both drives?

Click to expand...

You can scale the analog output from the SSD easily by changing the scaling value from 1.0000 to -1.0000.

On stop, issue the stop command to the master (and I'm hoping it is the SSD), and then when it actually drops out of run, you stop the slave drive. This allows for a controlled ramp down. I'd probably say change out the Marathon drive right off the bat.

kamenges said:

This is the way the SSD guys do it.
...
Then the SSD guys came in and changed it to the method shown in the PDF, likely because they didn't understand what I was doing and/or the method in the PDF "is the way they've always done it". That isn't a valid engineering justification, by the way.

Moral of the story is if you want SSD support going forward you should probably use the example in the PDF.

Keith

Click to expand...

SSD Apps folk don't usually do a lot of thinking, unfortunately. I've rolled many of their load share applications from versions like that PDF back to direct torque sharing, and gained huge performance improvements.

One problem with the 'speed trim', is that you will almost always have the drives fighting each other (often stressing the connecting element significantly).

I was sitting by once, starting up another line, while the SSD factory folk were trying to make their speed-trim load share work on two 200HP motors absolutely rigidly connected together... It wasn't pretty. After a few days, they decided the best bet would be to install resistors to trim the fields of the motors to better match. /sigh. It never worked, and the answer was to do it the simple way. The company eventually sent the entire machine back since it couldn't be made to run.

Many drive manufacturers LOVE speed mode for some reason, SSD is about the worst for that. I can't even count how many "SSD Factory Designed and Set Up" dancer applications I've fixed by spending 2 minutes changing from the 'factory approved' speed mode PID to torque mode and injecting the output directly to the current regulator.