RG9-1485

Day of the clones

Let's face it, we're never going to make a ten page essay on a separation pad the raciest read on the Web.   But these clones did cost a day, because they resulted in an extra site visit and then more effort to solve the problem.

The RG9-1485 is the paper separator pad for the HP Laserjet 5000 and Laserjet 5100 printers. Canon and HP have both withdrawn the part.   We thought only compatibles were left, however after having the problem we got a few dozen more “ originals ”.   After that, we are all in the “aftermarket ” where compatibles rule.

Contents, this page:

Part 2

We're writing the problem up to record the specific issue. Mainly, however, this is a case-study of what goes wrong with compatible parts.

Basically it's what you might expect.   Compatibles can be little better than look-alikes that don't work well or stand up to much examination. That doesn't prove that it's impossible to make a compatible part at a good price,  just that not every manufacturer does.

Some spare parts sites are silent on the issue of compatibles and counterfeits. There are a couple of operations with low prices who can only be selling low cost compatibles because their prices are lower than UK distributors discounted bulk prices.   We normally sell originals, it's less trouble for us and for our customers who are generally service engineers and site technicians for whom the price of a part is a small part of the cost of doing a job.

Selling originals simply won't do in this case because the supplier isn't making any more.   What we do here is take a compatible part to bits (literally and conceptually) to see what can go wrong - and how things might be got right.

Compatible parts create more problems than supposed originals but not massively more. And it can be very difficult to tell when a part is original or compatible because manufacturers often don't label them and the packaging does change over time. There are hints on what to look for below. Usually originals have cleaner looking plastic mouldings because compatible moulds see to be made from the original (details below).

Compatible parts have at least two roles in the market: they provide manufacturers with competition and ultimately their existence should be an environmental good.

Competition

People often buy printers just assuming that the price of cartridges and spares will be in some way “reasonable”. Manufacturers spares and repairs policies aren't always reasonable. For instance, there just aren't any spares for most consumer inkjets. Lexmark part prices are notably high. One of the UKs distributors made a fortune in the 1990s substituting cheaper Canon parts for HP's identical item. Once a manufacturer has sold you a printer they tend to think they have the upper hand. Compatibles are a way of reminding them that people can find alternatives.

HP are the worlds largest printer brand with just under half the market. It may be a surprise to know that they came into the market in a significant way only in 1984 with two new designs, the Thinkjet and Laserjet. Still more surprising is that nearly all the Laserjets use Canon engines. It even surprised HP management, well into the 1990s, that Canon could not or did not take this market away from them (see A Quarter Century at HP by David S. Cochran). Canon themselves have something like 20% of the market .

HP have always been pretty good at making spares available for their laser printers, perhaps it's their scientific instrument making heritage. Other manufacturers have been less easy to deal with. In the case of the LJ 5000 family HP are withdrawing spares afte a decade; its not unreasonable.

Environment

If you are a printer (or car, or washing machine) manufacturer you probably have some lifetime in mind for your goods. After that you'd like the customer to happily buy a new one. But printers don't rust. Like other IT equipment their processor and memory speeds do fall behind market trends but then a lot of people aren't doing anything that complicated. We have customers still getting good service from a LaserJet 4. The rush to replace old with new is good for manufacturers but bad for the environment.

So if manufacturers won't make parts available at reasonable prices then it's good that compatible makers will. But compatibles aren't always good. The compatible maker has no access to the print-engine makers dimensions and specifications so they reverse-engineer from what they see. Their primary interest will tend to be injection moulding rather than printers. What seems to have gone wrong with our batch of RG9-1485's is that the maker hadn't quite thought through how this superficially simple part must work. To get these parts right you'd need to take some printers to bits and look closely at how they work. However what actually happens with cloned parts is that come from a great number of rather anonymous producers through a manufacturing and distribution system that scarcely ever makes its workings clear.

Separation pads look about as simple as a part can get - a spring to hold the thing up, a plastic holder and the working surface is a rubber pad. It's about as complicated as a bicycle brake pad and less safety critical. What could go wrong?

LJ 5000 Family

The RG9-1485 is one of about 20 critical parts used in the LaserJet 5000 series printer. Without the pad the printer is essentially just so much waste plastic. (It is largely plastic, there's no metal chassis).

The HP LJ 5000 Series is an A3 mono printer widely used in engineering. A3 laser printers are not commonplace, although most manufacturers have one in their product line-up. The advantage of A3 from the user point of view is that it presents twice as much information in one go as A4. For instance engineering diagrams and charts have more of the contextual information that is often needed to understand a problem. The 1200 dpi maximum resolution is sufficient for pictures and diagrams to be detailed. Another use for A3 is newsletters; A3 folded in half creates a professional looking leaflet. Colour leaflets might be preferable, but they are more expensive.

The LJ 5000 printers were introduced in 1998 at around the same time as the LJ 4000 - they are stylistically similar although the control panel is swapped to the opposite side. Machines in the family are the LaserJet 5000, 5000n, 5000dn, 5000gn, 5100, 5100tn and 5100dtn - the suffix letters indicate networking, duplex and an extra tray.

It's fair to say that HP have fulfilled their obligations to customers by supplying parts for ten years. The machine engine was actually made by Canon, so they were an alternative source of parts, however they recently stopped supply of parts as well.

With an older colour printer everyone might agree to call it a day. New colour printers are much better than those made in 1998 (clunky old things based on a shared drum).

Possible Replacements

However these mono printers are still perfectly good. Nothing has very convincingly replaced them - the HP LJ 5200 is faster, has more memory and is somewhat re-engineered so uses different parts. (The separation pads are a modern style, nothing like those in the 5100). Unfortunately the LJ 5200 isn't cheap; well made A3 printers aren't. An LJ 5200 is priced at about £1600 in 2011. There are competing products from Brother (HL 3260) at a slightly lower price and from OKI (B840) at about half that price. However a lot of corporates have experience of HP products and stick with them. In terms of long term value it's a decision we wouldn't argue with - after all the printers have lasted well and HP have gone on providing parts from 1998 until a couple of years ago - that's pretty good.

So a replacement printer will be faster and have more memory. It will also have a slightly bigger 12,000 page cartridge but it's got a killer chip. So basically if you have a LaserJet 5100 then you'd probably like to hang onto it. There may be other problems than just getting parts, in the future driver support may be dropped, but 32 bit drivers should still work.

Can it Live?

HP might stop production of the C4129X cartridge and at one point suggested they had, but when we check it's still available and there is stock - as well as Silver Reed, Xerox, Data Products and Longbow equivalents - so the cartridges don't seem to be obsolete.

The first part to run out in distribution has been the RG9-1485 separation pad for tray 2 (and optional tray 3). We've got a supply of original parts for a few dozen more printers.

There is nothing unusual in manufacturers withdrawing parts whilst users hope to go on using the equipment. When a manufacturer won't provide support any more its an after market Manufacturers dislike the after market, they had hoped to sell new products for old. On the other hand they don't like to complain too directly and often take some pride when equipment outlives its expected life. In the future micro manufacturing and 3D printers may make the after market more effective. The really big after market is for automobile parts, reckoned to be worth $257 billion an 4.5 million jobs in the US.

In this case the RG9-1485 pad is a little part, it would be insignificant except that it's failure brings a good old printer to a halt. So it might be worth an extended look.

What the Pad Does


Paper feed is one of those things that seem simple until you think about it. With a typewriter you loaded the paper manually - that was true with some old ink jets like the Canon BJ-10 as well. Dot-matrix sprinters use sprocket feed paper. Some printers had the tray stick out of the machine at an angle so paper was part gravity fed. Laser printers pick up and feed paper automatically from a cassette that slides neatly under the machine.

When a printer kicks paper out of a tray it is possible for just one sheet to move. It is quite likely that the downward pressure of the roller and a combination of friction and adhesion between sheets of paper will move several. A separation pad tries to ensure only one sheet makes its way out of the tray.

When the pad is working correctly paper is pushed towards it by the roller and possibly more than one sheet moves at the same time. If there is more than one sheet the lower and any subsequent sheets will run fractionally onto the pad but will not then move much further because friction against the rubber of the pad resists their movement. Only when the paper is directly driven by the roller above is friction of the roller sufficient for the paper to move.

When the pad is worn it no longer grips the lower sheet of paper properly. Either the texturing of the surface caused by chipping by the leading edge of the paper, or the development of an indentation by friction with moving paper reduces the grip of the separation pad. Lack of grip by the pad makes it likely that two or more sheets might feed at the same time. Too much grip caused by a ridge worn in the pad will stop paper feeding at all. Either way, the printer will probably detect this because sensor timings will be incorrect, giving a 13.1 error.

Paper feed faults like this tend to be erratic. A fairly frequent scenario with pads is that the first couple of print jobs of the day fail, but then several succeed. Eventually misfeeds get really troublesome and the pads have to be replaced.

From HP and Canon's point of view theses printers have been out there for more than ten years and they don't feel obliged to go on supplying parts indefinitely, they would rather sell new printers. Their business mission is to sell you ink and the newer printers have killer-chip cartridges that aren't quite so easy for their competitors to refill.

How Pads Age

Pads age mainly because the constant movement of paper onto its leading edge erodes into the rubber surface roughening it.The roller turning and pushing the paper through also wears a slight recess, although this wear seems less than we might expect. The difference can be felt with a finger perhaps a bit more easily than it shows in a photograph. If you stick a worn pad in the jaw of a caliper the recess worn by the paper is more visible.

Some rubber like materials also decay becoming harder or softer with time.

In principle it is only the rubber that actually needs replacing. Digging the rubber strip out of the recess in the plastic holder means removing the metal strip and overcoming the adhesive layer on the rear of the pad, which might distort the plastic holder. Reliable repairs need the whole pad assembly, not just a new rubber insert.

Problem Pads

There seem to be several sources of compatible components for these printers. Unfortunately the principle supplier in the UK was shipping parts that look OK at first glance but proved defective in practice.

The problem found with a specific batch is that the metal lip which the paper slides up should slightly overlap the rubber of the pad. In the defective pads the metal ends at the pad, leaving a fraction of a millimeter step onto the rubber. Paper will intermittently and perhaps continually fail to feed at all. The metal should fractionally overlap the pad, leaving a slight drop onto the rubber that is overcome by the feed roller.

One reason for writing this problem up is that it illustrates some general problems about compatible parts - or indeed any part when it doesn't work properly.

We found out about the defects with the compatible pads by having them prove bad on site. Having repaired a couple of printers from an entirely non-working state we returned them to site and found they were misbehaving. A little bit of experiment showed the paper hit the front of the pad but got no further and closer examination showed why. The fault hadn't happened in the workshop, but then the printers were tested on a solid surface, it is very typical of feed problems that slight differences make faults evident.

As it happens, the site these printers work on is a subcontractor for a major European energy supply company and the one that malfunctioned was given to one of their on-site engineers - which resulted in a bit of a spat between managements. So we had to get things sorted pronto. One of the problems with compatible parts is that you can't save face by saying "manufacturers originals used throughout gov" or words to that effect.

Counterfeiting

Compatible parts are sometimes criticised for inferior materials. For instance US parts site Printer techs has an article about counterfeit rollers

here.

Which notes that a roller failed by slipping excessively

That’s when we looked very closely and noticed the rubber had a slightly different texture and feel. The rubber compound was too cheap and caused the roller to slip on the paper.

...

What really annoys me about this is the fact that there is a factory or factories out there pumping out tons of shoddy counterfeit parts with the intent to deceive people into thinking they are buying genuine HP parts

Printer techs seem to have hit an issue where the packaging was amazingly like the bar coded bag these parts normally ship in.

Our experience has been a bit different. There certainly are compatible parts but the packaging is bland intended neither to inform or deceive. (In the past we have sometimes been deceived by it. We investigate more closely now).

Counterfeiting rollers looks unlikely - why would anyone counterfeit bits of plastic that retail for £10 or less. There again, apparently Lego has a battle against counterfeits. In truth the manufacturing cost of a separation pad is probably about the same as that of a Lego brick and the price is higher, so if copies can be made cheaply and easily and sold as original they will be. We look at the cost of component manufacturing elsewhere

(below).

In this specific case the parts weren't counterfeit, they were intended to be compatible. However compatible parts often aren't sold with a brand on them and it's up to the distributor whether they want to mention the manufacturer. In the UK they don't at present , most compatibles are unbranded. The problems are therefore much the same.

Manufacturing Quality

In this case the materials did turn out to be different but that wasn't the immediate issue. The maker hadn't made the thing properly or fully looked at how the part would work in practice.


A close look at the way this pad is made raises several questions.

The parts come in different packaging. There may be parts out there that are in packaging very similar to HP/Canon packs - and that would be counterfeiting. These aren't counterfeit. We bought them as compatible when we found originals had run out.

Comparing an original part with the compatible they have a different finish to the plastic moulding. The original part has a smooth finish with very clean edges. A part number RB2-2038 is just discernible on the rear as is a legend saying ABS-FR

The compatible looks as though it was cast from an impression made in fine powder. It could conceivably have been made in an epoxy resin mould. Some metal moulds must have been made because the hooks at the back of the moulding aren't ortho-normal to any plate surface, there were probably a couple of slides withdrawn to release the part from the mould. The places where the ejector pins engaged the rear of the part have signs of irregular flow on the compatible part. There is a slight parting line on the original, but nothing is visible on the compatible, given the texture of it's finish that isn't a surprise.

The working part of this little assembly is the rubber pad. The rubber of the compatible is a quite different consistency, very much softer than the original. It is also about 0.5 mm narrower and that could also be the source of the problem here. In practice the rubber does work, the items that were found faulty on site were replaced temporarily with more from the same batch that were more correctly made and they worked. However given the carelessness about the metalwork we'd have to doubt whether a different rubber was used for a good reason.


It was defective metalwork on the leading edge of the pad that triggered this investigation. The metal has probably been punch-pressed from 0.1mm stainless steel and then put through an air-press or press-brake forming stage. On an original pad the outer tabs that cover the edge of the moulding are not quite formed the same way as the middle bit that engages the pad - between the bend-relief slots. Presumably the designer or perhaps the machinist didn't realise this bit was to be left straight and doing that might have required a special tool. Whether because of this or because of an error in dimensions the total length nose to tail is 4.48mm whereas the proper part is 5.0mm. On the original part the tabs that lock the metal to the plastic have been bent into a hook shape, on the compatible this wasn't done, they are just bent downwards. If the tabs had been made into a hook the metalwork might have engaged against the pad better.

Finally the original part had super glue holding the plate to the plastic. This would have been inconvenient to do. Perhaps there had been an issue for HP/Canon with the part slipping and they did this to counter it. That measure was absent on the compatible.


The top metal on both is more securely held, it forms a hook that inserts into a narrow recess in the plastic body. Both original and compatible seem identical, although closer examination shows the originals have a brighter finish.

At a casual glance these pads look OK; its only when things go wrong that you start to look more closely. The plastic moulding is rubbish, but that is just a clue to careless manufacture, its not the cause of the fault. The cause seems to be the metalwork, its not the same shape. We think it was designed wrongly, it could conceivably just be operator carelessness on the pressing machine. The rubber pad is slightly wrong as well, not quite wide enough, so that a gap between it and the metal is still more likely. Just for a clincher they didn't form the hooks or super glue the metal to the plastic.

Interestingly they could have made this part better than the original. The top metal has a tab that locks it into the rear of the plastic. The tray-1 pad has a tab holding the metal in place. A couple of similar tabs on the lower metal would have held it in place better. Compatibles could sometimes be better than originals.

Compatibles Generally

This problem raises some general issues about compatibles.

Close examination of the plastic moulding shows that it isn't as cleanly made as the original. The plastic surfaces aren't shiny and have a lot more texture. The injection points visible in the plastic are uneven. We might suspect that the mould was made by taking a casting of the original as that would be a quick way to do the job.

There is no branding on the packs these parts come in. We're not sure who made them. The UK distributor who supplied them is a subsidiary of a European group, who own a compatible parts maker. However they say these parts were from a different supplier. Recently they have offered to take back defective parts, but luckily we didn't have all that many in stock (and we've cut some of them up writing this).

Packets for compatibles are not branded because distributors seem to prefer this. It allows the compatibles to be substituted for HP / Canon originals when a distributor finds it convenient. We have to watch out for distributors accidentally/deliberately shipping compatibles when we specifically ordered originals. Shipping compatibles when asked for originals amounts to counterfeiting but when found out the response tends to be oops -picking error. We suspect that the big printer manufacturers are vitally interested in cartridge counterfeiting, but have been less concerned about parts.

Branded Kits

That seems to be changing. HP are increasingly producing branded parts in boxes. In some ways this is good news for the customer because these parts are intended to be end-user fitable and come with instructions. Reservations are that part prices have gone up; and on one of the current kits we can make no sense of the instructions and one of the parts. HP make mistakes - but that's a different story.

Who Makes Parts

Of course, there is a question as to what is original when the print engine was not actually made by the brand who put a label on the case. Most of the parts come in packs that "look as though they were made by Canon" and that is about all we can say. Canon have their own mould making plants in Japan and Virginia US so they clearly know a lot about injection moulding.

We don't know whether Canon actually make all their own parts, but suspect not. A great many Chinese manufacturers are claiming to make parts for HP and Canon. Printers seem to be made a bit like like cars - but not even the assembly plants always belong to the badge-name manufacturer. Its as though Nissan cars were made by Toyota and no one noticed.

It is quite possible that some "compatibles" are made in the same factory and from the same moulds as originals. Manufacturers may not like this, but it may be a sacrifice they make for better pricing. However the part in question here clearly wasn't made that way; its not the same at all.

Responsibility and Control

One thing we do have with original manufacturer parts is a chain of responsibility and control.

So original is actually a chain of control that generally avoids these problems by quality control inspection. They also sometimes take responsibility for problems when they occur, we have known HP to replace parts that didn't perform as expected. The problem with compatibles is that the actual manufacturer is anonymous and the distributor may have a tendency to brush problems off, so it's buyer beware.

Compatible part makers have access to examples of the parts that need replacing, but not to the CAD production files, dimensions, formulations and trade secrets used in parts. They do have access to the text of patents but would have to pay a license fee to use the techniques - and we can assume manufacturers aren't going to grant any such license.

Laser printers have been around for 25 years or more so the feed rollers and pads are mainly and established art. There are patents: Matsuo Toshiro and Okuyama Hideyuki of Sumitomo Rubber have US Patent 6481707 for a feed roller particularly suited to inkjet paper feed. The patent is useful for making clear the enormous number of compounds a ‘rubber’ roller can be made from. So perhaps we should be grateful that the compatible manufacturers ever get it right.

We look at a few of the problems the compatible maker faces in the next section .

Lessons

One of our friendly distributors has a slogan "Imitations have their limitations"; our investigation shows a truth in this. And although supplies of this part had officially run out we were able together a limited end-of-line supply of originals from them. That part of the problem is solved for now. But in the end there will be nothing but compatible parts left, so if you want to go on running this printer (and we want to sell parts) the compatible parts have to be got right.

We also have a new batch of compatibles coming in and are assured they will be from a different supplier. Wanting a different supplier would be a natural reaction. The problem is that, at the moment, we don't have any great reason to believe that any specific compatible supplier will consistently get things right. They are actually in a difficult game, so we look at that in the next section.

We do have customers intermittently saying things along the line of this isn't original (subtext, I want a refund) ”. We relay that back to the distributors. They tend to say We all hear this from time to time, it was original . And of course it is possible for the original manufacturer to get things wrong (- hence, perhaps, the expedient of super-gluing the metal to the pad).

Vigilance is part of the answer. We watch the packaging on parts and note changes. We're discussing the practicality of checking bar codes, batch and item traceability, detailed photographs and accurate weights.

A US firm in a similar line of trade is making test rigs for part quality. If the part performs right it's good. We are considering the same thing. It might be better if the compatible makers had the laboratory rather than making parts that go into distribution and only get trashed after an argument.

The problem with all this is that it adds to costs, We don't currently work at the end of the parts process where we could avoid passing those costs on. We have to compete on price on the Web and can't charge more just because we're taking care.

It might be better if, rather than leave everyone guessing, the printer manufacturers clearly labelled their own parts or said what packing should be expected.  We guess they don't because packaging and stock cost money and the distribution part of the spares logistic chain doesn't like it. The logistics of spares distribution is expensive. When a manufacturer (or its manufacturing partner) orders 10,000 rollers from a subcontractor they want those through the global distribution chain and out of their hands in no time ( as they would be for printers or cartridges). They want the ambiguity of being able to ship a part as theirs when it was actually made by a competitor.   Printing, packing and labelling costs money- frankly they'd rather put effort into shipping another inkjet cartridge.

So parts distribution is a funny game - the next section takes a look at some of the issues .



Copyright G Huskinson & MindMachine Associates Ltd 2012