Hacker News new | past | comments | ask | show | jobs | submit login
Why Paper Jams Persist (newyorker.com)
242 points by ohjeez on Feb 6, 2018 | hide | past | favorite | 77 comments



Exciting fact: Similar jams happen in hot steel rolling mills too. The YouTube videos are just prettier: https://www.youtube.com/watch?v=DVZ1DY0r4y4

In general the rolling mills call it a "cobble" when a whole or partial bit of steel escapes the rolling mill. Lots of exciting videos.


That's pretty insane. Wow, look at those coils of hot metal. It's interested how much of it is still hot after they clear the jam and run it through.

I remember years ago hearing about how the US mint misprinted over a million and had to destroy it (and on the Colbert Report, he joked about how them destroy it by burning it in front of poor people). We were talking about it at a party and someone asked, "surely they noticed? Why didn't they shut it down?" An engineer in our group said, "Have you ever been in a factory? It's so fast!" He worked in a toothpaste tube factory where you'd see hundreds of tubes come out every minute. If something went wrong, you'd have to wait for the whole pipeline to clear. Fab shops in China are even more insane.

I wonder if the mill would just sell those sheets at a reduced price as low grade/for non-visible construction or if they re-melt them down and run them again.


On the speed front, I concur. Back in 1998, when I was on summer break from university, I worked in a wines and spirits bottling plant. One day the labelling machine started applying wonkey labels to the wine bottles and a couple of hundred had gone through before the guy running the machine had noticed.

They weren't wildly misaligned, but far enough from true that we had to wash them off and run them through the line again. Took a couple of us most of an afternoon to sort it out.


At what point is it just more economical to sell them at a reduced price to restaurants (where the customer might not see the label)? I'm sure the guys (you) peeling the labels are probably the lowest paid in the factory but a few staff doing the work for the whole afternoon adds up.


At $10/hr for 2-3 people, assuming they could clear up 30 bottles an hour you're only talking $1/bottle. I don't think that small of a discount on a bottle of wine would encourage a restaurant to buy a veblen good that didn't look perfect.

[1]https://en.wikipedia.org/wiki/Veblen_good


This is totally an n of 1 type story, but perhaps relevant nonetheless.

I worked at a bar during university during a time where the popular beer sale would be "28 bottles for the price of 24". The bar owner would send me and a few guys in a truck to a beer store in another town to buy as many cases as the truck could handle. This was illegal as we would sell it at the licensed bar (all liquor was supposed to be purchased with the license, for which the sale wasn't applicable too). So the owner was willing to risk legal action for 4 free beers (x50 cases or whatever).

The point is... restaurant owners sometimes go to great lengths to save a few dollars.


Just speculation, but might there be regulations regarding the labeling of alcoholic beverages for sale that prevent accepting crooked labels or even price reductions?


It's called quality control.


Boom


I did some work with a drinks can manufacturer, who were looking to improve the speed of detection of printing errors. The lines are so fast that detecting a print error even a few seconds earlier saves thousands of wasted cans.


Or as a piece of artwork.


Any idea what the dimensions of that steel was? I'm imagining it was over 6ft across and seeing that jam being cleared was amazing.


It's even worse in our low-end inkjets, which are made of plastic parts and toy motors and are still expected to duplex-print on the cheapest crap paper imaginable. Our FW devs spend a lot of time perfecting the PID loops and mech sequences to make this happen. But do we get any thanks for that? Noooo...

Love that New Yorker article, though. I grew up reading this magazine in the 80s but drifted away. These great articles straddling tech and culture became too few and far between.


This is why I love the New Yorker. You don't see enough of this kind of long form journalism, shedding light on seemingly mundane issues that you'd never imagined could be so interesting.

The historical perspective, in particular, was great. But what took this to another level was detailing just how big the challenges are, and how enthusiastic the engineers are in trying to solve them. Even giving a sense of vicarious excitement from the descriptions of their attempts at problem-solving.


I dunno. I think this is one reason I stopped reading the Atlantic/Harpers/New Yorker. I mean, it is better than sensational TV where they take 45 minutes to get to the punch line where they could have explained it in 5 min. It's not even 40 min of interesting facts or history; just mostly commentary and talk to draw things out.

I do realize this article is different; there is a lot of historical fact and context which helps create a narrative. It has its advantages. Maybe I'm just getting old. I read a lot of novels and I'd rather spend the time on that type of narrative in a novel or even a non-fiction book than a news article.

People may criticize the YouTube generation, but it has helped in content creators who get a lot of information over in a very concise way (Kurzgesagt In a Nutshell, Smarter Every Day and Wendover Productions come to mind).

I think the opening of the historic paper jam made me want to fall asleep. The later office space references and Xerox design became more relevant. Even then, I feel like I don't read these articles the way I did back when I bought a physical Harpers from the book store. It's a lot of content, and in the scrollable world we just end up skimming more these days.


>People may criticize the YouTube generation, but it has helped in content creators who get a lot of information over in a very concise way (Kurzgesagt In a Nutshell, Smarter Every Day and Wendover Productions come to mind).

Reading the same info (in a non-narrated way) would be much faster. Also it's often concise over exactitude or neutrality. I can't sit through these videos, boring or caricatural. At least the New Yorker is sold as an enjoyable distraction, and I find great pleasure in reading nice writing.


It's the medium. If you read the same thing on paper or on a Kindle the temptation to skim would be much less.


> or even a non-fiction book than a news article.

This isn't a news article. Perhaps it would help if you thought of it as a freely available, very short non-ficton book.


I've been reading Delayed Gratification for a year now. It's all this sort of stuff. If you like long form journalism I highly recommend it.

https://www.slow-journalism.com/


Thank you for this. I've been deliberate in not reading news for a couple of years now. I subscribed to the Atlantic Monthly print edition which is slow enough to get some digestion rather than velocity. The most important outcome in looking for is reduction in anxiety and avoidance of getting sucked into pointless debates... It's very effective to say you never heard of Jordan Peterson (I heard of him last night) and you're thus physically unable to participate in the ensuing gender politics melee. What I notice about myself lately is that I go to Google news whenever I'm anxious or avoiding something. It's would be instructive to plot my browsing history to news sites with actual anxiety over time


Wanting to cut down on news consumption was one of my original motivations for subscribing. While I haven't cured myself entirely, I'm able to keep it a lot more under control. Reading about Trump and Brexit every single day was having a very negative effect on my mood and outlook. I remain aware of the stuff that matters, without having to deal with the blow by blow every single day.


Thank You!


Agreed. I read most New Yorker issues cover to cover. Their writers cover a wonderfully-broad range of topics in useful depth. It's the only print publication that I still subscribe to.


I like the London Review of Books.


Many, many years ago I had a student job with a small company doing industrial document finish, enveloping, and large-scale mailings. Apart from a still eaily induced paper eczema, the one thing I took with me was a rather comprehensive facility with the black magic of paper handling. Not an ounce of explicit science in it, but an awful lot of feel: When and how much to air and flex; where a grounded length of copper wire could work wonders; how much or how little moistening to apply, and where; how too touch the stream of paper and somehow know by the feel of the tip of your finger whether something would probably jam within a minute, and which tiny speed-adjustment up or down would probably forestall the disaster; when to to loosen or tighten this or that set of rollers because nightshift and the air getting cooler et cetera et cetera.

To this day I never just bang a fresh stack of paper in the printer tray. The full gymnastics of airing, flexing, feeling, airing and flexing again. I'm often laughed at. But my papers never jam!


In college I worked in an old style (precomputer) printing shop - learning how to work with paper like that is something that never seems to leave you (I always do it when I load a laser printer too)


I've seen people do this before, I always just assumed it was to make sure the pages weren't stuck together, and therefore reduce the chance of 2 sheets being pulled together.


Well, it is, sort of. But in a fast moving production line, it could make a huge difference in jam propensity whether the sheets were fed by somehow with a nice touch or not.


We take sheets of dried wood pulp, stack them so loosely you can always see a few that aren't 100% aligned with the rest, then feed them through a series of unlubricated rollers inside a machine that's usually indifferently maintained and clogged with dust, all while deliberately exposing them to heat and static electricity and spraying their surface with goopy ink or powdery toner.

The surprise isn't that they jam occasionally, it's that we ever get anything printed at all!


Coincidentally, just last night I was telling my wife how impressed I am with my work's copier/printer machine.

It's about the size of a 1980's copy machine, but my department easily prints a minimum of 5,000 color pages with it each day. Some days it's double or triple that. Sometimes with staples.

We go through five massive toner carts each week. They're about the size of a fire extinguisher.

In six months, I've seen exactly ONE paper jam.

I told her I thought maybe they'd figured out this whole printing thing finally. But I guess TFA says I'm wrong.


What's the make and model?


Don't know the model, but it's a Kyocera. Before this job, I'd only been exposed to that brand as one of the cheap printers for the Commodore 64. I guess it's come a long way!


I enjoyed the article. Just want to point out briefly that the explanation given in the article of how airplane wings create lift is incorrect.

http://www.explainthatstuff.com/howplaneswork.html


Also, slightly annoying was the description of the inside of a photocopier as being 'like a darkroom'. It seems like a small thing, but the whole point of the photocopier is that it is _not_ like a darkroom; it's a completely dry process. That was Carlson's innovation, copying without a wet chemical process.


For a more detailed explanation of how so many explanations we see for lift are wrong, see the page "Bernoulli and Newton" from NASA: https://www.grc.nasa.gov/www/k-12/airplane/bernnew.html

It's part of their excellent "Beginner's Guide to Aerodynamics" site: https://www.grc.nasa.gov/www/k-12/airplane/bga.html


I think they just messed up what side is flat, right?


No, symmetric airfoils generate lift just fine, and if you were going to spend half your time upside-down, you'd use one.

Making it asymmetrical merely makes it more efficient to generate lift in one particular direction (similar to how glasses lenses are asymmetrical to reduce distortion).

Furthermore, pretty much no airfoils have flat sides, in nature or in engineering[1]

The wing, when placed in airflow, disturbs the air in such a way that the pressure on top of the wing is lower than the pressure below the wing, generating net lift. This isn't a very satisfying description of lift, since it just begs the question of "how/why does it disturb the airflow in such a manner" and the short answer is "we put different shapes in a wind tunnel and measured" with the longer answer being "you don't know enough math to get a good intuition of that"

[edit]

After reading the article, the article did get lift right for the case of the paper though. Jets of air behave exactly as they say, and you can feel something similar by touching the back of a spoon to the stream of water from a sink; it will get pulled towards the stream of water (or just put a piece of paper on a table and blow over it). A few planes use blown flaps[2] to help get a similar effect (the introduction of a jet of air, which "hugs" the wing helps direct the airflow of the larger mass in the direction you want).

1: https://en.wikipedia.org/wiki/Airfoil#/media/File:Examples_o...

2: https://en.wikipedia.org/wiki/Blown_flap


The spoon in a stream is more of a surface tension thing and the classic blowing air over a piece of paper is the coanda effect.

The total effects of newton's laws and bernoullis law on pressure differences need to. Be combined but the gist is that you create a force difference using the shape and angle of attack of the airfoil.

The use of this is really quite neat especially in a printer!


>Because the top side of an airplane wing is flat, while the underside is curved, the air above moves faster than the air below, and the wing rises.

That's what the article says, but it's usually the underside that is flat and the top is curved.


In reality usually neither side is flat.


A flat bottom wing gives you nice stall and landing characteristics. Very small, 2-4 seat, trainer airplanes have them. (Cessna 150, J3 Cub)


I wouldn't call the Cessna 150 flat, as it has visible lower-camber, but you are right about the J3 Cub; it's super flat on the bottom.


That's part of it but it's also incorrect to say "the air above moves faster than the air below, and the wing rises."

The air speed is not necessarily different. Lift is generated because air is deflected from moving over the top of the wing and thus reduces the air pressure. You can create flat winged aircraft with careful orientation of the wings.

Air flow is not like say a spinning disk, where the solid state of the matter forces the outer edge of the disk to move faster in order to maintain relative position.


If you ever found yourself in a car, not driving, in a semi-desert place where it's sfe to put your hand out of a window, try that:

- First, put your hand parallel to the ground, and see what the wind does to it.

- Then, incline your hand so you raise your fingers, and see what the wind does to it.

Lifting has no relation to the wind being flat or curved. It's the overall shape (the one you change once you incline your had) that causes it.


I'd like to see another article on why in 2018 every meeting seemingly has to start with 5 minutes of fiddling with the screen projector.


Is that still a thing? For small meeting rooms (less than 10 people) our offices have large screen TVs (60 inch maybe?), and for larger meeting rooms it looks like there are NEC projectors (I just checked 1). I don't recall a project issue within the last 6 years.


Projector fiddling has been replaced with VTC fiddling. I hate Polycom/Skype with a far greater passion than any printer.


Preach. Seems like someone could make buckets of money undercutting Cisco with a better product


in my experience, 4 out of those 5 minutes is a result of apple changing the requisite number and configuration of display adapter dongles every 2 days, with the other 1 minute resulting from from wifi network access policies being innane in most building common areas


I still hate printers, and their jams, but I must reluctantly appreciate the engineering behind them after reading this.

This seems quite clever:

'Corrugating is when we put an intentional wave in the sheet, like in a piece of corrugated cardboard,” Vicki Warner explained. “It adds stiffness.” The plan was to corrugate the sheet lengthwise by running it over a line of rollers turning at variable speeds before “flying it” into the stacker. If a physical fix was necessary, a part might be 3-D-printed and installed, on-site, by one of the engineers.'


I have been a part of this process at Xerox.. it works quite well.. I've been involved in a number of different aspects of it as well. I've been the source of some of these fixes, I've been the source of some of the issues, and I've been involved in the investigations to solve them. Xerox has a few different labs for testing different things.

The "stub point" is a common thing, and can be very, very difficult to pinpoint. They had one specific finisher that kept jamming when you printed to the top tray. It didn't jam 100%, it was pretty intermittent. Well, this finisher had been in another lab for a few weeks or so and they couldn't find the source of the jams or a pattern to them. I had the finisher installed on one of my machines in the 3 days I'd solved it. There was a small stub of plastic on a flag for a sensor. The flag sits in the paper path, and when its moved, the sensor changes state. So this flag has an pretty smooth surface so as to not catch the paper, but somehow this one got a little nick in it. Enough to cause a jam when a sheet ran right into the nick. I filed it down and the machine never jammed again.

We had another issue with a printer at a customer site and the issue dealt with static electricty. a 30kv ESD gun and the inside of a printer.. we still never solved the issue.


The challenge helps explain why the user experience for desktop printers hasn't changed much in 25 years.


I'm surprised that Xerox pays engineers to do all this, when so many companies would pay their lawyers to make it someone else's problem when an odd type of paper stock jammed in tropical weather. Making custom parts and flying out to fit them in production machines? I thought technicians only got to do that during wars. I suppose you might do it when the machine is the one that prints the price labels for Walmart, though.


I don't know about Walmart but here at Target we just use some random small business Epsons for the signing and Zebras (previously Monarchs) for the price strips and stickers


Zebras are indestructible. It’s what Amazon uses in their warehouses for everything. I was looking for a label printer and my options were a new dymo label printer from 2015 or a zebra released 15 years prior. The zebra still won.


They're also a lot of fun to code against, if you're like me and have a thing for ancient serial cable text protocols designed by electrical engineers.


I haven’t coded against the Zebras but I have against their counterparts from Epson with ESC/POC and fully concur. It’s like a throwback to the 80s or early 90s.


I worked a couple years ago somewhere with iGen printers. Xerox guys were there all the time, absolutely top class service all the way when you're at this level of printing.


> (Before railroads transformed the transportation of lumber, logjams had to be addressed by “jam breakers”—experts who spotted and removed the “key logs” jamming up the river.)

River transportation of lumber kept going long after railroads came along. Jam would be broken using dynamite sticks.

Here's a movie on the driving of logs on a river in the 1950s: https://www.nfb.ca/film/drave/ Driving only stopped in the 1980s on that river.


Rivers are still in use for moving log booms today - you can see them on the Fraser river and other places in and around Vancouver BC - although they tend to be more orderly, with strict regulations around the logs being properly boomed and contained..


Thanks. You reminded me of "El río que nos lleva" (The river that carries us"), a beautiful spanish movie from 1989: http://www.imdb.com/title/tt0098239/?ref_=fn_al_tt_1


The "Secret Life Of Machines" episode about the photocopier is also illuminating: https://www.youtube.com/watch?v=S2NIAD5qn7E


Wow, that's really interesting. I didn't realize that's where the term "blueprint" came from.


Tim Hunkin's TV programs are excellent!


I work for a major ICT provider that sells about everything enterprise related, from servers to storage systems, laptops, scanners, etc. In terms of customer support the printers are the absolute worst: usually it's always a critically important since they're used to print the labels for factory outgoing and if they stop for one hour it's millions of euros lost. This means the ticket will be Priority 1 and everything needs to be solved within minutes or else people will freak out since there's a whole factory plant jammed with cars, tires, whatever, that can't leave the factory without the printed stamp.


I wonder if we could go jam-free with Continuous Stationary ( https://en.wikipedia.org/wiki/Continuous_stationery ). That paper with the holes on both sides from the 80s, and some legacy systems today.

It seems really helpful, but I'm not sure which problems discussed it would actually solve.


Had a dot matrix printer that used continuous paper. Notorious for jamming. Usually something getting stuck on the little holes, or the perforated sides tearing mid-print.


The worst was when it lost traction, shredded the holes, usually in a tangle of paper.

There'd also be cases where it kept pulling on the holes but those separated and got gummed up really badly.

When you try and tug the pages out they're all perforated so you get, at best, a single page. It can be a lot of work to clear a jam.

Most people think of dot matrix printers as slow, lumbering things that would take measurable time to print a single line. The industrial ones, enterprise grade, made by companies like IBM were dizzyingly fast, pages per minute, so if a jam happened it was usually severe. Most had multiple print heads to speed up printing.

These also used much wider paper than normal for things like accounting reports. When they jammed, they jammed.


One of my first "tech" jobs was refurbing dot matrix printers after we had them in the field for a while. The amount of little bits of that perforated paper that would get stuck in every moving part in one of those was kind of staggering.

People were also really rough on those things. Yanking at the paper to tear it as the printer was just finishing it up.


I had several dot matrix printers and can't remember any jams. And I did manage to accidentally dump the whole database to the one at work, the stack was at least 5 cm thick by the time we managed to stop it. It had it's own room though because of the sound so not your normal home printer...


Something similar could probably work if your print jobs required all the same width paper. Printers today need to handle many sizes (legal, letter, tabloid, envelopes etc etc).


There must be a market for "I only ever need A4 or letter" (depending on your area).

I honestly don't remember printing anything but A4 for the last 20 years. We printed some labels in the 90s, that was the only diversion...


I print so rarely, that about half of what I do print is A3.

If I don't want to read the thing on a screen, then doubling its size is often useful.

But, for metric sizes, two consecutive sizes could be supported. A4 would print landscape, and A3 portrait. The large photocopier-like printer at work does this.



Worked with plenty of line printers that used this type of paper in the 90s, and they still jammed. And it was often very bad, as a lot of paper could get jammed together before the printer and/or operator noticed there was a problem - it's not just one sheet getting stuck.

One of the jobs our operator had was to constantly clean, dry, and dust off the rollers that went in to those little holes on our high-speed line printer - the amount of paper dust produced by a few boxes of paper was astounding.


Inst this easy to solve by not feeding sheets, but from a role and cut after printing? Cheaper production and transportation.


Page wide, Precision core.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: