Roller Concentricity

Hello everyone,

Can I bring up the subject of roller concentricity?

Last year we purchased a restored treadle press and the new rollers that came with it had, at their worst, 0.14mm/0.0055” of deviation in concentricity, measured at the runners or trucks.

As I am sure you don’t need me to tell you this resulted in a cycle of darker and lighter inking, as the roller slowly turned in use.

They remade them and two are vagely acceptable (I’m being picky here). However, I recently sent the rollers from our other treadle for assessment and recovering to another company and they have come back the same.

I do ink lightly, I’m not at all a fan of over inking, so of course the variation is more noticable.

I really want to crack this, as it feels like the next obsticle to refining our printing. I’m more than happy to have some new cores and runners/trucks made - I’m even wondering if they should be ground finished, rather than turned to ensure the quality I am looking for. But of course all that will be for nothing if the rubber is applied and turned poorly.

Any thoughts on this subject?



Log in to reply   8 replies so far

What tolerances does the roller people offer?

Where did you send them? I personally ink on the heavy side, but I’ve noticed zero roller issues with my rollers from Advanced Roller in California. A ground finish I believe is the standard for most letterpress rollers.

To me the most accurate way to make a roller would be to:

1. cut the roller core to slightly longer than finished length
2. chuck each end of the core in a lathe and adjust it with a dial indicator so there is little to no runout, but certainly less than .0005 inch. Then face off each end of the core and center drill it
3. lightly mount the core in the lathe between centers and check again for runout with a dial indicator. If runout is ok, proceed to step 4.
4. apply and cure the roller covering
5. mount the roller between centers in the lathe (don’t mount it any tighter than necessary), and grind it to finished diameter with a tool post grinder

Even if the roller is made this way, runout can creep in if the roller is mounted too tightly in the lathe for step 5, because that can bow the core while the roller is in the lathe. (This has happened to me). If the core is bowed in the lathe for step 5, the roller will be ground straight with the bowed core, but when the roller is taken out of the lathe, the core will spring back to being straight, and then the roller covering will be bowed.

For the rollers to run true in the press, the trucks and rails need to be in good shape as well, of course.

Thanks for your comments, all really helpful. I am UK based, so been using UK suppliers. Sadly none seem to specialise in letterpress or even printing rollers, they are more generalised (industrial rollers etc). If anyone in the UK has good experience with any company I would be pleased to hear it.

Both sets of rollers on both presses have centre marks from being turned between centres and my understanding is that they use these for grinding the covering.

The trucks or runners on the older press are actually loose fit onto the core spindles, the Arab press has a pin that registers them.

Has anyone ever thought of using bearings on a treadle press, like the Heidelberg platen does? I was looking at our Heidelberg the other day and wondered if that was a good way to go.

I wonder if there would be any advantage

I’ve been looking at - who specialise in letterpress rollers. They specify Durometer readings for their compounds. Is that something that anyone has played with or noticed differences with?

You can expect a roller to run out to less than .0005”? That is a half of 1/1000”. seems a bit tight to me.

James, Have used All Make Rollers for recovering my windmill rollers and no problems. All Make have now been taken over by Regal Rubber and they have recovered 2 more rollers for me, but not been used yet. Also in UK is Bottcher rollers. Both recover for commercial printers.

ericm, I meant the bare roller core should run out less than .0005”, in my opinion, not the finished roller.

A problem inherent in making rollers is, the roller maker references to the center holes drilled in the core, because that is how they mount the rollers during manufacture. The printer, on the other hand, references to the outside diameter surface of the core, because that is what the trucks go onto. So, if the center drilled holes are not drilled exactly in the middle of the core with reference to the outside diameter surfaces, runout will creep in. That is why I said that when a core is mounted in the lathe between centers and dial indicated to the core O.D., the runout should be =<.0005” in my opinion