ABB ACS800 drive - acceleration question

Total_newbie

Member
Join Date
Aug 2020
Location
Detroit
Posts
24
Members of this forum helped me out a great deal a couple of years ago with a question I had regarding acceleration during torque control of drives.

http://www.plctalk.net/qanda/showthread.php?t=126282&highlight=acs800

I am getting back to working on my software control of this drive (ABB ACS800), and I am seeing some behavior which I cannot understand. If drive is set to internal speed control (parameter 98.01) and I specify target speed with drive accel at max. rate - I see speed increase exactly at rate which I expect (based on specified motor torque and inertia which I have as load). E.g. it takes ~13 seconds to go from 0 to 740RPM, with constant speed ramp. (See attached pic - upper graph. Y -axis is 0.1X. Motor torque shown in brown is also 0.1X scale - so actually going to (-)100%.


However - if I switch to external (speed encoder) feedback (changing nothing else) - this same ramp takes twice as long. Also, accel rate is no longer constant - but increases with increasing speed. (bottom graph). Speed as sent from encoder is correct as measured with tach. (And encoder module was swapped for another one - no difference...). Motor torque still shows up as -100%.

Could anyone give me some idea as to why switching between internal and external (encoder) speed feedback should lead to such significant difference in accel rate? I cannot leave drive set to internal speed control, as this gives some error in speed at higher speeds and does not properly pick up zero speed.

(Note: I put this question up on another forum (Mr. PLC) - but didn't get any real ideas as to what might be the cause for this. Based on my experience in this forum - there are some experts who really know their drives...)

Thanks!

Capture.PNG
 
Just replying to bump because I'm kind of interested in this myself. This issue goes completely against the little bit I thought I knew about the ACS800 and DTC. It was my impression that with the ACS800 the encoder only provided actual feedback to the velocity loop instead of the internally calculated speed. The torque generation algorithm is always the same and independent of the feedback source. More correctly, transistor firing is not affected by feedback source.

This is as opposed to the ACS550 which operates more like most other drives on the market with more typical sensorless vector and FOC with feedback algorithms. It would appear I was mistaken.

Keith
 
Can you show torque in Nm rather than percent - if the motor is developing the same torque in both cases, the time to accelerate the motor/load cannot change.
 
if the motor is developing the same torque in both cases, the time to accelerate the motor/load cannot change.

Good point (and we don't need Nm or ft-lb or furlong-Whey;)).

I just noticed that the [external speed control] plotted torque data have a round edge that approaches the -10 (≡ 100%) torque limit smoothly, while the [internal speed control] plotted torque data have a hard edge at the limit. This suggests to me that that in the latter case the actual torque blew past that plot limit, and maybe there is something in the data collection that limited the plotted data.

@OldChemEng told me once that VFDs can go from zero to full torque in half a cycle, so my interpretation may be due to an artifact in the data and/or the plot, but as @LD notes the difference in slopes is a clear indication that the torque must be greater in the [internal speed control] case, assuming the load inertia is constant across both plots.
soft_vs_hard.png
 
Last edited:
Keep in mind these torque numbers are NOT coming from a torque sensor or some other calibrated device external to the drive. They are the drive's internal calculated torque. So the reported torque is very likely NOT the actual torque in either case. Supporting data would indicate the first, non-feedback, trend is closer to reality. However, both cases are calculated.

It would certainly be worth checking that changing the velocity feedback source DIDN'T also change the max torque value in the drive. That seems an odd tie but stranger things have happened. But all things being equal, the inference is that the drive is calculating torque differently based on the velocity feedback source. With most other drive manufacturers I would buy this. But this seems to fly in the face of ABB's calling card.

Keith
 
Thank you to everyone for your ideas/observations.

Insofar as the point "if the motor is developing the same torque in both cases, the time to accelerate the motor/load cannot change."... Yes - that seems to be make perfect sense. So... it must be that the scaling on drive for max torque % (which I am reading in as analog feedback -as % of full scale - from drive and plotting), is somehow being changed when switching between internal to external speed feedback. But (as was pointed out) - that doesn't appear to make sense (that max torque scale would be changed).

I don't have access to the drive right now to do additional testing, but when I do - I will check the current level during accel that the drive is putting out. Current level is in absolute units (Amps) - so at least we will know more clearly what the drive is doing, independent of any internal scaling.

I can also check whether max torque value on drive changes with switch from external to internal.

Let me know if there is anything else that anyone can suggest that I check when I get drive back for testing...
 
Just getting back to this (perplexing) issue of seeing different max torque from drive when switching between external (encoder) feedback and internal speed. I tried switching between external and internal on two other drives (ACS800) which we have here - and there was virtually no difference in accel rates between the two. Looking at current shown on problem drive during accel - I see max current (max as per motor rating) when operating with internal speed, and about 80% of this when using external. (Even though we are getting about 50% torque...).

The only thing that I can think of is that perhaps there is some ripple/noise in the external (encoder) signal, which when fed into speed control loop on drive - is alternately increasing/decreasing actual torque. Average torque (accel) may show up as smooth - but rapid up/down during control is impacting overall rate. (But... I did try changing some parameters of PID speed control loop on drive - and no effect at all on anything...)

We are trying to bring the experts in on this - but till now... they have not given us any clear ideas as to what this might be...
 
Problem resolved...

Just in case anyone else encounters this rather confounding issue (different drive accels depending whether internal vs. encoder is used for speed) - we think that we have found answer. Motor slip (which is parameter which is usually only set at setup - not sure if it was never set correctly for this drive) - was off. It should have been ~7RPM, and it was actually set to 15 RPM. When using encoder - drive is using speed feedback to get to max. torque (it thought it was going to max. torque...) - but since this motor slip setting was off - closed loop control was messed up.
 
That actually makes a lot of sense!

I remeber years ago, we had to run a separate program to autotune Yaskawa closed loop vector drives. It took several minutes to exercise the motor, measuring the actual slip was one of the functions.

I don't believe the modern drives bother to measure and adjust the slip parameter anymore.

I have had several Open loop drives act pretty stupid when the customer sets the wrong motor rated speed in the drive. The slip-comp will drive the speed loop nuts.
 

Similar Topics

I am getting an AD1_PLACE1 fault on my drive and am stumped, any pointers would be appreciated!
Replies
3
Views
1,843
Hi All, We have a ABB ACS 800 Drive, Have 2 AO for RPM and current output. As default setting, both are available on X9-12 terminal. I am...
Replies
1
Views
2,154
Dear team, i need to run ABB ACS800 two drives in master follower mode.Im using ABB RDC0-03C option kits with fiber optics lines. As they...
Replies
2
Views
2,283
Have any of you come across a easy to understand quick start manual for a ABB ACS800 Drive. ABB seems so complicated. Really just looking for the...
Replies
8
Views
7,158
Hi Guys, I know its not directly PLC related but it is connected to my PLC :) Just a quick one then, We have an ABB ACS800 Drive and when we...
Replies
6
Views
4,486
Back
Top Bottom