Soba rotary table

[thaiguzziParticipant @thaiguzzi]

I bought my HV6 Vertex new in the early 90's. It was worked hard in a professional shop until 2003 when i retired it to hobby use along with myself.

So that's well over 25 y/o.

I still have all the paperwork, instruction manual, and factory inspection data.

I also have period early 90's Vertex product brochures.

It was most definetely made in Taiwan.

[Neil WyattModerator @neilwyatt]

Posted by Oxymoron on 13/03/2019 14:23:26:

I have recently been looking at rotary tables for use on my mill (SX2.7). This discussion has been very useful to me.

However a question: Some tables seem to be sold with dividing plates and I am wondering if these are actually needed. In my imagination rotary tables can be positioned to within a few seconds of angle. So why are dividing plates needed as surely a spreadsheet could quickly give the number of degrees, minutes etc for any number of divisions.

What have I missed please, are dividing plates worth having with a Rotary table?

Regards

Dave

Shh… it gives people a great deal of comfort to know they can work to a level of precision far higher than the accuracy of their equipment

Seriously though, they can make dividing quicker or less error prone, but things like vernier adjustments and worrying about rounding errors is pointless for 99.9% of work. I've only ever made gears working to the nearest 0.1 of a degree (ten seconds) which is within a thou at the edge of a 2" diameter gear.

Neil

Edit: Looking at the comments above, it really does matter to match the precision you work to to the job in hand. Working to a precision significantly greater than the process you are using or the capability of the tooling might give you a warm glow, but in reality it's just a waste of effort better put into some other aspect of the work.

My astronomy hobby has taught me an awful lot about working with different levels of angular accuracy. You need to be within about a degree to find objects. You might work to a few arc-seconds to accurately align them in a camera frame. You work to seconds of a degree when tracking, which needs constant error correction. With the clock ticking it's pointless to work to more accuracy than you need.

Edited By Neil Wyatt on 13/03/2019 15:19:29

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Neil Wyatt on 13/03/2019 15:04:22:

… to the nearest 0.1 of a degree (ten seconds)

.

[OxymoronParticipant @oxymoron]

Thanks Guys. Pretty much as I assumed, you can manage without the plates but they make it easier. And less error prone. I need all the help I can get!

[geoff walker 1Participant @geoffwalker1]

Posted by geoff walker 1 on 13/03/2019 09:23:31:

On accuracy this table is 7.5 degrees out over a full 360 degree rotation of the table. The hand wheel dial is indexed with 7.5 degrees and the hand wheel rotates exactly 47 times for one rotation of the table, should of course be 48.

LOL. Cost saving measure. One tooth saved on each worm gear means that every 47th gear they get one for free.

I bet you had some headaches until you worked that one out though. Who would think to check that?

Hi Pete,

Yes I did have some headaches!!!

When I first got this table I set the 0 on the circumference dial to the cursor and 0 on the hand wheel dial to the cursor, then turned the table through 360 degrees and much to my astonishment the dials lined up again. I thought WOW!!! that's accurate. Of course at the time I didn't realise that I was one turn short, 47 x 7.5 degrees and not 48.

The table has a cumulative error of 1.25 minutes per degree of revolution.

As I said in the original post I can still use table by making a coarse setting with the circum dial, which is accurately engraved and a fine setting with hand wheel dial. So for say 132 degrees turn 130 on the circum dial and 2 on the hand wheel dial. This gives me "theoretically" 132 degrees with an error of 2.5 minutes which is fine for what I do

Geoff

[Anonymous]

Posted by Michael Gilligan on 13/03/2019 15:13:04:

Posted by Neil Wyatt on 13/03/2019 15:04:22:

… to the nearest 0.1 of a degree (ten seconds)

I think we should minute that for future reference.

Andrew

[JasonBModerator @jasonb]

Posted by Michael Gilligan on 13/03/2019 15:13:04:

Posted by Neil Wyatt on 13/03/2019 15:04:22:

… to the nearest 0.1 of a degree (ten seconds)

.

Don't point that out to him Michael, he is now going to realise that he has set his telescope wrongly too and been looking at the wrong planets for the last couple of years and what he thought was Uranus was something completely different

[Michael GilliganParticipant @michaelgilligan61133]

Posted by JasonB on 13/03/2019 15:54:21:

… was something completely different

.

in the elbow constellation ?

[not done it yetParticipant @notdoneityet]

Posted by Michael Gilligan on 13/03/2019 15:13:04:

Posted by Neil Wyatt on 13/03/2019 15:04:22:

… to the nearest 0.1 of a degree (ten seconds)

.

Nearer 2 thous if 1/10 th, and almost 3 thous if 10 seconds – to the outer edge of a 2" circle?

One might only need a mirror, JB?

[Chris TriceParticipant @christrice43267]

Posted by geoff walker 1 on 13/03/2019 09:23:31:

On accuracy this table is 7.5 degrees out over a full 360 degree rotation of the table. The hand wheel dial is indexed with 7.5 degrees and the hand wheel rotates exactly 47 times for one rotation of the table, should of course be 48.

I'd argue this qualifies it as worthless junk since it renders many of the tasks a rotary table is reasonably expected to do as unreasonably difficult and therefore not fit for purpose. I'm sure if your life depended on it, you could achieve all sorts of miracles using it if you had to but you shouldn't have to, regardless of how cheap it was.

[JasonBModerator @jasonb]

Posted by not done it yet on 13/03/2019 16:11:38:

Posted by Michael Gilligan on 13/03/2019 15:13:04:

Posted by Neil Wyatt on 13/03/2019 15:04:22:

… to the nearest 0.1 of a degree (ten seconds)

.

Nearer 2 thous if 1/10 th, and almost 3 thous if 10 seconds – to the outer edge of a 2" circle?

One might only need a mirror, JB?

10 seconds would be quite a bit less, 10 mins would be about 3 thou.

[not done it yetParticipant @notdoneityet]

Oops, not concentrating!

[SillyOldDufferModerator @sillyoldduffer]

Posted by Oxymoron on 13/03/2019 15:13:16:

Thanks Guys. Pretty much as I assumed, you can manage without the plates but they make it easier. And less error prone. I need all the help I can get!

…

Dividing for all but the simplest divisions soon becomes extremely tedious and it's very easy to lose track.

For example, say you want to cut a gear with 41 teeth in it. Without dividing plates you have to set the scale by eye in 8.78° increments as accurately as you can, 41 times. 82 or 123 times if the teeth are nibbled out or generated.

It's more accurate to count turns and part turns of the handle as dictated by the table's worm ratio. This method is hard to get right without pre-calculation and mechanical help.

Dividing plates come with a printed index table where someone else has done the sums. (Not always correctly –  check for mistakes before cutting metal! ) Looking up my HV6 table with a 90:1 worm ratio, to cut 41 teeth, I need to fit division plate C, then to take two complete turns of the handle plus an index of 8 holes into the 41 hole ring around the C plate. This automates the integer arithmetic and a pair of clock-hands keep track of the next increment and all the others. As mistakes result in the wrong number of teeth or damaged teeth, it's important to minimise them.

Even with help from division plates the process is error prone. The best way to drive a rotary table is with a stepper controller doing the maths and keeping count for you.

I recommend getting a set of division plates if you intend doing any complicated divisons.

Dave

Edited By SillyOldDuffer on 13/03/2019 17:05:22

[OxymoronParticipant @oxymoron]

Dave, (SOD) many thanks for taking the time to explain above. Very informative. I've started work on a 1" Minnie and currently turning wheel rims. I figure I'll need a rotary table before long and when it comes to cutting gears which I'd like to try myself then dividing plates could be very useful.

I've learnt a lot from this thread to help make an informed purchase.

Dave

[Ian Skeldon 2Participant @ianskeldon2]

Does anyone here use a Soba roary table without any issues? As already covered it depends on the level of accuracy the user is trying to achieve but it would be interesting to hear from owners of such tooling.

[JasonBModerator @jasonb]

Posted by Ian Skeldon 2 on 13/03/2019 18:29:17:

Does anyone here use a Soba roary table without any issues?

See my post towards the top of the first page.

[Ian Skeldon 2Participant @ianskeldon2]

I have seen lots of the things you have made, so in the right hands they're very useable.

Thanks Jason

[HopperParticipant @hopper]

Posted by Ian Skeldon 2 on 13/03/2019 20:04:58:

I have seen lots of the things you have made, so in the right hands they're very useable.

Thanks Jason

Well the one Jason bought eight years ago was. But see Ketan's posts about declining quality from certain factories/suppliers over the years and you see what a difference time can make. Then see Geoff Walker's experience with a rotary table from another supplier that had 47 turns of the handle instead of the requisite 48. Seems like unless you go for the higher priced items (eg 400 quid) these days you pays your money and you takes your chance.

 

Edited By Hopper on 14/03/2019 05:36:09

Edited By Hopper on 14/03/2019 05:38:33

[PeroParticipant @pero]

My 4", 6" and 8" rotary tables are all older Vertex so I cannot add comment on the Soba or recent Vertex quality.

However, like Geoff Walker I also purchased the RDG Tools 2 3/4" rotary table, in my case to use with a Proxxon mini mill. The relative scales are just about right – the 8" Vertex is definitely overkill with this machine.

Unopened until today when I read this thread, I opened the plastic bag and lo and behold – 47 turns. Checked it twice counting full turns (47) and once counting half turns (94). Same outcome each time – yes I can still count – hooray.

I checked the RDG website where it is still available and still advertised as 1:48. I have emailed RDG advising of this and asking for a response. Perhaps they will send the missing tooth!

Alternatively does anyone have a recommendation for an itty bitty stepper motor and the necessary program to convert it to a useful bit of kit?

Pero

[Michael GilliganParticipant @michaelgilligan61133]

Pero,

Your proposed alternative certainly seems the best work-around

To keep things compact, I suggest you look at NEMA14 size stepper motors.

for example: https://www.pololu.com/product/1208

https://www.pololu.com/product/1208

MichaelG.

.

P.S. it would be helpful to know the height of the worm axis on that table

Edited By Michael Gilligan on 14/03/2019 07:30:27

[JasonBModerator @jasonb]

Hopper have you got a link to the £400 tables and what sort of spec do they come out at compared to what the OP measured?

[Neil WyattModerator @neilwyatt]

Posted by Michael Gilligan on 13/03/2019 15:13:04:

Posted by Neil Wyatt on 13/03/2019 15:04:22:

 

… to the nearest 0.1 of a degree (ten seconds)

.

Oopsy! Six minutes…

Edited By Neil Wyatt on 14/03/2019 08:34:48

[Neil WyattModerator @neilwyatt]

Posted by not done it yet on 13/03/2019 16:11:38:

Posted by Michael Gilligan on 13/03/2019 15:13:04:

Posted by Neil Wyatt on 13/03/2019 15:04:22:

… to the nearest 0.1 of a degree (ten seconds)

.

Nearer 2 thous if 1/10 th, and almost 3 thous if 10 seconds – to the outer edge of a 2" circle?

One might only need a mirror, JB?

Actually, if you work to the nearest 0.1 degree the maximum error is 0.05 degrees, so 1 thou.

In practice I have been known to work to half a division so 0.5 thou.

And yes I had a brian fart…

Neil

[Neil WyattModerator @neilwyatt]

When I made my first rotary table by free hobbing, I got 61 teeth then I used the approach of gashing the blank by indexing it with a change wheel.

I think I still have the duff gear. Pero, I could cut the extra tooth out and send it to you if it's useful…

Neil

[PeroParticipant @pero]

Thanks Michael

I'm not quite sure how much holding torque I would require and whether I could possibly go down to a NEMA11.

I haven't dismantled it yet – I have to, it feels a bit gritty – so cannot advise on the worm height. I am guessing it will be very low ( probably a bit less than 20 mm ) as the whole table is only 38 mm in height. I suspect with either a NEMA11 or NEMA14 I may have to fit a sub-plate onto the bottom of the table for clearance. That said, the table on the Proxxon MF70 is quite small so the motor might comfortably hang over the edge and its size may then not be a problem.

Further investigations for tomorrow.

Cheers

Pero

PS A check on the Pulolu web site indicates that the centre height on a NEMA14 is 17.5 mm so it should be right on the mark or very close to.

Edited By Pero on 14/03/2019 08:55:45

[Author]

Posts