please sanity check my idea transmission line going through sliprings)

Thread Starter

strantor

Joined Oct 3, 2010
6,782
I have a machine that is a thorn in my side. it is a tubular stranding machine. basically a stationary machine with a giant spinning tube inside, and inside the giant spinning tube are stationary gimbles which contain reels of wire. there are 9 gimbles, and a VFD on each gimble. It has a profibus network, and the designers have run the profibus through sliprings. I am forever having issues with the slip rings. if a wayward piece of lint happens to light upon the slipring, the whole machine falls on its face and I have to dive into it (literally) and go clean the sliprings/brushes. This used to happen maybe once every month or so; now it has become almost once every shift. I am trying to come up with a solution. I think by doubling the # of brushes I can square the chances of success. please have a look at the drawing. I am hoping someone here is familiar with profibus DP, or any kind of input about transmission lines would probably be helpful.
 

Attachments

Thread Starter

strantor

Joined Oct 3, 2010
6,782
I've tried to sell the wireless idea up the chain for this issue in the past. It wasn't recieved well. That's why I'm now looking into fiber. I wish I could just open people's minds for them; it would make things a lot easier.
 

Thread Starter

strantor

Joined Oct 3, 2010
6,782
maybe the signal could squeeze through the sliprings a little better if I reduce from 1500 kBit/s (current setting) to 9.6 kBit/s (minimum)?
 

crutschow

Joined Mar 14, 2008
34,285
Sliprings can be a pain. Doubling up on the sliprings or brushes should help significantly. Then you would need to have a piece of dirt/lint on the same point (simultaneously) in both sliprings to generate an error.

Lowering the data rate should also help.

Another consideration is the load on the sliprings. If there is a pull-up or pull-down type load then any disconnect of the slipring may give a false one or zero. If you can make the load high impedance and capacitance only, then the slipring output signal will tend to stay at whatever level it is at when the slip-ring loses contact (sort of a sample and hold).
 

THE_RB

Joined Feb 11, 2008
5,438
What's the data baudrate?

If it is low enough you can run a severe RC filter (integrator) at the receiving end, the UARTs should have a schmitt trigger buffer anyway and you can up the RC values to where it will reject all noise <5% of the baud period. The final waveform should reach min or max within a single baud, but that still allows a lot of filtering.
 

thatoneguy

Joined Feb 19, 2009
6,359
New brushes and springs?

What's changed to make it a common error rather than a rare thing?

How would the fiber line up? Concentric ring receiver if I'm thinking of this the right way?

Perhaps a quick and dirty 3D isotropic sectional view of one unit?

Or fiber stationary going outside ends of giant tube, and the sliprings (little x in your drawing connectors or brushes?) supply the power?
 
Last edited:

Adjuster

Joined Dec 26, 2010
2,148
Where is the equipment supplier in this story - is it an in-house machine? Redesigning the failing system to fix it will make the equipment non-standard. This might lead to difficulty with future support if an external manufacturer or other third party may be involved.

There may also be issues with safety/liability/insurance following modification. You should only consider making such a change if your employer fully supports your doing it. My apologies if doing this sort of thing is part of your normal job description, but that's not known to me.

You might also want to be sure before modifying anything that the problem is really with the slip-rings, and not some other issue. For instance, external interference or other effect could be eroding the noise tolerance of the repeaters, worsening the effect of minor contact skips.
 

Thread Starter

strantor

Joined Oct 3, 2010
6,782
Another consideration is the load on the sliprings. If there is a pull-up or pull-down type load then any disconnect of the slipring may give a false one or zero. If you can make the load high impedance and capacitance only, then the slipring output signal will tend to stay at whatever level it is at when the slip-ring loses contact (sort of a sample and hold).
I don't think I can change the impedance, as it is part of the profibus DC spec. there is a specific value of termination resistor, and there are no t-junctions allowed, as this adds capacitance. so I assume I am not allowed to add any capacitance.

What's the data baudrate?
1500 kBit/s
If it is low enough you can run a severe RC filter (integrator) at the receiving end, the UARTs should have a schmitt trigger buffer anyway and you can up the RC values to where it will reject all noise <5% of the baud period. The final waveform should reach min or max within a single baud, but that still allows a lot of filtering.
that's over my head.

New brushes and springs?
All of the brushes have been replaced recently as a "shotgun solution" that didn't hit it's mark. springs may be, I hadn't really considered the springs because they all seem to have the same tension; I assumed that means they all have the correct tension. I will look into that.
What's changed to make it a common error rather than a rare thing?
I wish I knew. back when the problem occurred monthly, it was attributed to a particularly dusty (copper dust) product that is sometimes ran in the machine. Now that it's happening more than once per shift, it is obviously not caused by this dusty product.
There was an incident a couple of months ago (back when the problem was stilly a monthly occurance) where I had just cleaned the brushes and the guys on the next shift had a problem again. they suspected the brushes again (it wasn't the brushes) so they decided to clean them again, only "real good" this time. They ground down the surface of the brushes by hand with sandpaper and rounded them all off, removing the natural contour of the brush to the slipring (hence why I replaced the brushes), and they cleaned the sliprings with hard pumic motor commutator cleaners. the problems started escalating shortly afterward. I'm not sure if that was the cause or not.
How would the fiber line up? Concentric ring receiver if I'm thinking of this the right way?
The OLMs (siemens Optical Link Module) in the picture are these. they have one hardwire channel and 2 fiberoptic channels. They can be arranged just like in the picture I drew, in the ring configuration. I have set them up like this before. The only problem is that the siemens modules are only rated for 1G and I intend to mount them inside a tube that spins up to 400 rpm. I am looking for a higher g-rated device.
Perhaps a quick and dirty 3D isotropic sectional view of one unit?
I assume you are referring to slipring & brushrack assembly? I don't have one.
Or fiber stationary going outside ends of giant tube, and the sliprings (little x in your drawing connectors or brushes?) supply the power?
Yes the little Xs in my drawing are brush/sliprings through which the profibus signal is passed. The profibus signal does not power the modules. The modules are powered from a seperate 24V bus, which is already in place, and also goes through sliprings which are not shown in my drawing. The fiber I drew in blue will be permanently mounted inside the tube and will never move.
Where is the equipment supplier in this story - is it an in-house machine? Redesigning the failing system to fix it will make the equipment non-standard. This might lead to difficulty with future support if an external manufacturer or other third party may be involved.
The supplier is in austria. They are notoriuosly hard to contact. This is another added frustration. I must get my supervisor invloved any time there is an issue, because they will only answer an email if he sends it, or if he sends them an email demanding that they answer my multiple unanswered emails. Getting them to come here and actually lay hands on the machine is a huge ordeal and costs tens of thousands of dollars, if not more.
There may also be issues with safety/liability/insurance following modification. You should only consider making such a change if your employer fully supports your doing it. My apologies if doing this sort of thing is part of your normal job description, but that's not known to me.
Any safety issue that could potentially be introduced by this modification
is already present by original design. I always get the support before I make any major modifications. I am gathering information right now so that I can present my idea to my supervisor & management. I need to know that this will work before I do that, and I need to know the answers to any questions they may ask. My job description is written in grey ink. I was hired because of my experience with controls/PLCs, but I change gearbox oil more than I install/program PLCs here. My experience with controls/PLCs is not vast though; they were looking for someone young and excited to work, and that's what they got. Unfortunately I cannot be this, and also be a guy with 40 years of experience. They know exactly who I am and they take my gand upgrade ideas with a grain of salt (last year I tried to convince them to let me upgrade to wireless data transmission and get rid of slip rings and they said no). They also know that something has to be done about this machine. even failing once per month is not acceptable. so maybe they will listen to me, maybe they won't. maybe they'll call in a 3rd party to figure something out. But I'm still going to try to find a solution regardless. This machine should have never been deisgned with profibus signal going through a series of slip rings IMO.
You might also want to be sure before modifying anything that the problem is really with the slip-rings, and not some other issue. For instance, external interference or other effect could be eroding the noise tolerance of the repeaters, worsening the effect of minor contact skips.
Every time the problem crops up, I clean the slip rings, the problem goes away. That's what's got me convinced that its the slip rings. it may well be like you said. if that's the case, how can I tell?
 

GetDeviceInfo

Joined Jun 7, 2009
2,192
I've tried to sell the wireless idea up the chain for this issue in the past. It wasn't recieved well. That's why I'm now looking into fiber. I wish I could just open people's minds for them; it would make things a lot easier.
what I tend to do in these situations is to assemble a working prototype and present the concept based on that platform.
 

Thread Starter

strantor

Joined Oct 3, 2010
6,782
what I tend to do in these situations is to assemble a working prototype and present the concept based on that platform.
I have a machine I could possibly use as a test platform. it is an older machine that is rarely ever used. Actually I believe it's slated to be scrapped. only problem is that it is an open-air machine with an open frame. The machine I'm having problems with is a 1" thick tube inside a 3/8" thick sealed enclosure. the EMI effect may be different between demonstration and application.
 

THE_RB

Joined Feb 11, 2008
5,438
Are you sure it is 1500 kbit/sec or 1500 bit/sec?

It sounds very nasty running something as high as 1500 kbit/sec through sliprings. What were they thinking?

What are your cleaning the sliprings with, when you clean them? And post-clean lube?They are probably specced for a particular lubricant that reduces corrosion and electrical noise.
 

Thread Starter

strantor

Joined Oct 3, 2010
6,782
Are you sure it is 1500 kbit/sec or 1500 bit/sec?

It sounds very nasty running something as high as 1500 kbit/sec through sliprings. What were they thinking?

What are your cleaning the sliprings with, when you clean them? And post-clean lube?They are probably specced for a particular lubricant that reduces corrosion and electrical noise.
yes I'm sure, and that was what I thought "what were they thinking?". I assume they had a reason, so I'm not sure if it would be ok to change the setting, but I guess it's worth a try.
I'm using a lint free paper towel. Originally I was passing 300 grit sandpaper between the ring and brush (grit towards brush) a couple of times and blowing down with air, then rinsing with no-residue contact cleaner and cleaning the ring with a fine commutator cleaner. My supervisor contacted the manufacturer of the brushes and they said to never use solvents or abrasives, just to blow down with air, and to never clean the ring. They said that the brush is supposed to leave a patina on the ring ( discoloration) and that this patina helps conduction and it should never be cleaned off. The manufacturer said nothing of a lube.

Until you threw the fiber optics in I was going to suggest these - http://www.mercotac.com/
 

shortbus

Joined Sep 30, 2009
10,045
Could you show a link to the machine? Or a photo showing something similar? To give a better idea of what your up against.
 

GetDeviceInfo

Joined Jun 7, 2009
2,192
I wish I knew. back when the problem occurred monthly, it was attributed to a particularly dusty (copper dust) product that is sometimes ran in the machine. Now that it's happening more than once per shift, it is obviously not caused by this dusty product.
There was an incident a couple of months ago (back when the problem was stilly a monthly occurance) where I had just cleaned the brushes and the guys on the next shift had a problem again. they suspected the brushes again (it wasn't the brushes) so they decided to clean them again, only "real good" this time. They ground down the surface of the brushes by hand with sandpaper and rounded them all off, removing the natural contour of the brush to the slipring (hence why I replaced the brushes), and they cleaned the sliprings with hard pumic motor commutator cleaners. the problems started escalating shortly afterward. I'm not sure if that was the cause or not.
are your brushed knife edged running on a hardened slip ring? That configuration would be self cleaning, as opposed to a flatter surface that lifts away due to contanimates.
 

Thread Starter

strantor

Joined Oct 3, 2010
6,782
the brushes are exactly like this one (except newer looking):


and they ride on a set of rings like this:


I can post a pic of the actual setup on monday if that
 

shortbus

Joined Sep 30, 2009
10,045
Can't find a link other than this, but the brushes shown here that have the "dust grooves"(slots) in them may work better for you. The top part of the link shows the ones with the slots, the stuff that gets in them is moved out of the way by the slip ring motion.

Have you checked the spring tension on the holders? With different people working on them maybe some one stretched the springs?
 

Thread Starter

strantor

Joined Oct 3, 2010
6,782
Can't find a link other than this, but the brushes shown here that have the "dust grooves"(slots) in them may work better for you. The top part of the link shows the ones with the slots, the stuff that gets in them is moved out of the way by the slip ring motion.

Have you checked the spring tension on the holders? With different people working on them maybe some one stretched the springs?
there's no link in your post, but I think you are on to something there with the spring tension again. Every time the brushes are cleaned, they are pulled back to insert a cloth or whatever between the brush and the ring. this could explain why the problem happens exponentially more often. more failures leads to more cleaning which stretches the springs, which leads to more failures. I am going to see about getting new springs.
 
Top