Ultra 3000 following auxiliary encoder question

Hi,
I have a 2098-DSD-100X-HV indexing ultra3000 that I want to follow the speed of an auxiliary encoder. The encoder is a 2500ppr, single ended, from Automation Direct mounted to a rollform machine and is in contact with the material being formed. The encoder signal is sent to a FET Micrologix 1500 for length counting.
I want to use this encoder to also drive the ultra for a linear screw actuator (flying shear). I have set up feeders using Ultra's following pulses from ML1500's with the 24V signal converted through a Calex.
Is it possible to do this directly from this encoder? If so, could anyone provide a detailed idea?

Thank You,
Lou

Bump

Revisiting project. Please help.

Hi,
Quick question about setting up the load:motor ratio for the auxiliary encoder in Ultraware. I have finally installed everything and it's wired up, but I need to know what the ratio's should be for accurate speed/pulse tracking. Here is what I have:
1)2500ppr Automation Direct encoder #TRD-GK2500-RZD with 3.8" wheel.
2)Ultra3000 #2098-DSD-HV100X
3)Parker #ETA-100 linear screw actuator with Reliance #F-4075H-OO-AA servo motor.

I found I have 159850 pulses from the servo motor for every inch of linear travel of the actuator. The manual for the servo motor states the encoder has 2000 lines. My auxiliary encoder has 2500ppr and has a 3.8"D wheel that measures 12" per revolution.
I'm not sure how to do the math involved with figuring out the correct load:motor ratio to enter into Ultraware.
BTW- This is for a metal rollforming project where when the length measurement is reached, the actuator will move the shearing die to match the feed speed of the auxiliary encoder then shear and return to home.
Any help would be greatly appreciated.
Thank You, Lou

bump

Lots of information in the link

see page 10 for formulas for scaling encoders

http://www.dynapar.com/uploadedFiles/Products/Danaher_Encoder_Handbook.pdf

Thanks for the reply Mike,

Another question I have is, if using an auxiliary encoder, wouldn't one pulse from the aux encoder move the servo/actuator motor one pulse if load:motor ratio is 1:1? Thus, What I really need to do is figure out the actuators speed, not the servo motor driving it, but I have to use the actuator motors encoder pulses to do the math. Correct? Also, I would quad the aux encoder to 10000ppr and the motor would be 8000ppr. Is this right?

Thanks,
Lou

Think in terms of position scaling because that is ultimately what you are shooting for. You stated earlier that the actuator motor generates 159850 'pulses' per inch of actuator travel. I am going to assume you mean pulses here and not counts. That is to say the actuator motors turns just under 80 revolutions for one inch of travel. I am going to further assume you want one inch of actuator travel per inch of web travel.

The aux encoder generates 2500 pulses per 12" of travel, or 208.33333 pulses per inch. I don't recall the scaling set-up in the Ultra3K, but that would be 208.33333 master counts per 159850 slave counts.

2500 PPR is pretty coarse for what you are trying to do. It may ne a bit noisy when you try to move. Also, I see alot of these applications where someone says "All you have to do is..." or "It's as simple as...", or, my favorite, "We did something exactly like this except...". While the fundamental idea of what you are attempting is pretty easy to grasp there are alot of details that will make life difficult.

Keith

Ditto and......

kamenges said:

The aux encoder generates 2500 pulses per 12" of travel, or 208.33333 pulses per inch. I don't recall the scaling set-up in the Ultra3K, but that would be 208.33333 master counts per 159850 slave counts.

2500 PPR is pretty coarse for what you are trying to do.

Click to expand...

The slave must estimate the master's velocity and accelerations from the master's coarse encoder feedback and that will be difficult unless there is some fancy software in the drive. The slave needs to have accurate estimates of the master's velocity and acceleration to generate feed forwards and to use the derivative term.

Thanks Keith.
I think finally got it now. The actuator motor actually produces 15985 pulses per inch of actuator travel. I have the A&B channels from the aux encoder going into a Calex 8502 and the 5V TTL output going to the Ultra aux encoder A&B channels. Would noise still be a concern? The line speed max's at 100FPM.
Thanks,
Lou

I'm using the term noise in a different sense than you are. Noise is any information you don't want. In this case I am talking about false or incomplete information.

Peter's post describes the issue pretty well. The issue with low resolution is that the change in position per servo scan is too coarse to determine accurate velocity and acceleration. Without those you have no effective damping, which is where Peter was going this the derivative term point. In addition, the indexer can only move when it receives a master pulse. If your slave increment per master pulse is "large" then the slave will move in a series of short jerks, because that is all it knows to do. As Peter said there are some reasonably sophisticated tools that can be used to work with the coarse updates but the Ultra3K doesn't have them.

I recommend you get the highest resolutiuon encoder that will not put you over your input frequency and that you can afford. It is pretty easy to get 10,000 PPR these days at a very reasonable price. 60,000 PPR is not unheard of. Given the speed you are going, 10,000 PPR would only give you an input frequency of 16,667 Hertz, which is really low.

Having said all that, it sounds like you do this now by feeding command information out of a MicroLogix 1500 in the past. If you have gotten a system to work smoothly enough for you then if you match the resolution of the master encoder to the resolutoin you used out of the ML1500 in the past you at least have a reasonable chance at success.

Keith

Thanks again.
I've had success, with proper tuning, of getting the ML1500 to pulse out through the Calex to the drive with no issues. I've just eliminated the PTO in this case and split the encoder signal between the plc and Ultra letting the plc do the measuring and work the I/O's for control over everything.
I'll attempt a 10000ppr encoder when it comes in.

Thanks,
Lou

Do you recal what your ML1500 pulse increment was? Would you get close to that with the 10000 PPR encoder?

Keith

I'm not sure if I understand the question right, but the servo feeders I have, which I've used the ML1500 PTO, are geared 1:7(pinch roller:motor) and the pinch rollers are 2" diameter. These setups pulse out 73.5 for every inch of travel through our feeders, when the Ultra is properly tuned. Not sure of the motor ppi though.

Lou

You would be above that pulse rate by almost a factor of three with the encoder. So you should be OK, based solely on past performance.

Another possible issue is that you may get disturbances from the sheer contacting the web being picked up by the encoder and causing deviations on the shear. If your shear gets kind of twitchy during the cut you may be seeing this.

Keith

Thanks.
Fortunately for me, the encoder is located at the end of the roll former behind two forming dies and about 12 feet from the shear. I found deflection at the last two dies to be a big problem with accuracy/consistency. Also, the common problem of trying to explain the critical relevance of encoder mounting and material contact to the set-up operators who can only smile, blink, and nod. Lou