Rotary table indexer

Quote:

---

Originally Posted by vishnu

Thanks, i am planning to make a UHU controller and connect them. I have a long cycle and have to run my rotary indexer Non-Stop for 10 Hours a day, so i think a servo would be better choice over stepper. Like to hear your views, i am a novice in electronics ....... awaiting your comments to start your Rev 2.0 design.

---

Hi, I'm afraid I'm no expert on steppers vs servos. I've made dozens of stepper related things (The rotary controller was initially a box-joint cutting jig!).

What I would say is at low speeds there's probably not much in it and steppers are dirt cheap and very easy to implement.

Whilst servos are better at high speeds, for indexing high speeds aren't that useful (most of the time the table hasn't reached max speed before it starts to decelerate).

The maximum step rate available from the controller is 10khz - you can calculate the theoretical time to do 1 rev from this.
If your stepper/encoder is 400 steps/counts per turn and the worm ration is 90:1 then the time to make 1 revolution would be:-

Time (seconds) = (steps * worm) / 10000

In this case (400 * 90 ) / 10000 = 3.6 seconds.

Using a 24v ex-laptop supply and cheapo 2A stepper, I find realistically I can drive a six inch table at half that speed so it takes 7.2 seconds to do a full rotation. Might be worth plugging some numbers into the formula and seeing what it means for you.

Another point is that when steppers are stopped they're still drawing power - this is bad because - well they're still drawing power! (Although most stepper drivers I've seen lately reduce the amount of power used when there's no motion) however it's good because it 'locks' the stepper and helps prevent movement.

In terms of reliability I can't vouch for either - my gut feeling is your rotary table will wear out before the stepper though.

In terms of construction, it's really simple requires perhaps an hour - maybe two... You'll spend far more time making the hardware to connect the motor to the table (although these can be bought - at a price) and making a suitable case.
Read the construction notes and if you have any questions my private email address is in documentation.htm.

Obviously I'm coming at this from a 'hobby' level - if your indexer is industrial, then it's probably no help at all. I'm sure at an industrial level the servo/stepper dynamics are completely different...

Regards
Steve.

Hi Steve,

thanks for your views. The rotary indexer i am planning to convert is for a particular job. A brass rod with center bore and eight holes have to be drilled. For one index currently i have to do 10.5 turns * 8 holes = 85 turns approx for one piece. My production quantity is 1000 nos per month thats approx 85000 turns manually. For that particular production job alone is was in an idea to converting my indexer all other jobs i do like them doing manually as they are very low production jobs. I have a helper who is doing that job currently and he feels really tired.

The reply was very informative and would like to continue with a stepper itself as cost is a factor and its not of industrial use other than the job i mentioned which runs up for a day.

I will start up with the project next week and hope to post the developments in this thread.

How much did it cost you to build one. Have to source components as its not easy in India where we can find them all in one place.

Your design looks realy simple and easy to build with a good documentation.

i would like to know about ic programmers as i am a novice in electronics. Please guide me on that issue only.

Regards,

Vishnu

Vishnu,

For that volume, get rid of the rotary table and build a fixture. Or just get rid of the drive reduction and use the table. Fasten an index plate with eight holes onto the table. Setup a pin that can drop into a hole. Drop the pin; drill the hole; pull the pin; manually rotate the table; repeat seven more times.

Ken

Quote:

---

Originally Posted by vishnu

Hi Steve,

thanks for your views. The rotary indexer i am planning to convert is for a particular job. A brass rod with center bore and eight holes have to be drilled. For one index currently i have to do 10.5 turns * 8 holes = 85 turns approx for
{snip}
Vishnu

---

Hi Vishnu,
As with any production run as Lerman says you'd probably be better off using a jig (if it's possible). However if complex sequences are desired the 'program' mode would allow these to be carried out simply and error free.

For programming - software is available here:- http://www.ic-prog.com/index1.htm it supports various pieces of hardware - a common one being the JDM programmer - circuit diagrams are available on the web and completed programmers available cheaply on eBay.
In terms of cost a quick tot up suggests building it on a piece of stripboard would cost around £30 (UKP) although it should be possible to do it for less. This includes £12 for the keypad from Farnell which you may want to replace with discreet buttons. You'll also need a case.

Obviously stepper and controller depend a lot on what you are going to drive with it. For small 4 to 8 inch tables you can get a motor for around £18 and driver for around £25, with these you can also use an old laptop PSU.
For larger tables you'll obviously need a bigger stepper, controller and PSU. In this case using a small wallwart PSU for the controller might be better than driving everything from the same PSU.

Hope this helps.
Steve.

If anyone is interested I can provide a V2.0 programmed 18f452 at cost.

Posted to the U.K this is £4.60
To the U.S £5.58
(other countries please mail me)

You can paypal me using the email address in the documentation. (If using paypal please add 0.40ukp to cover their costs).

Note prices are in UKP (U.K pounds) currently 1 UKP equals 2 U.S dollars.

Just in case anyone has missed them V2.0 firmware, diagrams, construction, manual etc is available on the bottom of page 4 of this thread.

I am designing a stepper controlled faceting machine and want to control the index and angle and this is exactly what I need. I need to drive one stepper nema 17 1.8 deg motor thru a 72 to 1 worm gear for the index and the second axis is for angle setting but I do not have the drive math done. I am not a electronics person as far as board design and construction. So is there anyone that can help me with this unit. Please contact me off thread if possible.
Thank in advance,

Jim Walthall [email protected]

I think this is a really cool project, and I would like to build a 5C indexer. This morning I sent the parts list to a few electronics supply houses to see if they could quote on the component prices. They all
asked for more info on the components. Is there a more detailed description of the parts?

dan k

Quote:

---

Originally Posted by kmcwhq

I think this is a really cool project, and I would like to build a 5C indexer. This morning I sent the parts list to a few electronics supply houses to see if they could quote on the component prices. They all
asked for more info on the components. Is there a more detailed description of the parts?

dan k

---

Hi Dan,
More detailed description? They're all pretty basic parts... Did they mention any parts in particular?

I'd assume there's no problem with the resistors,caps, transistors etc - simply buy the same part numbers.

The 18F452 is probably available in a few different case styles in which case you want the 40 pin DIL version.

The LCD is a standard HD44780 based device - in fact I've never seen a 20x4 display that wasn't suitable. Just type HD44780 into eBay then look for a 20x4 version - there are hundreds!

The only awkward bit is the keypad - you can get this in the U.K. from Farnell or R.S - I'd assume a subsidiary of theirs exists elsewhere.But the keypad is just an arrangement of 16 buttons into a 4 by 4 array, so it's fairly easy to make your own.

Hope this answers your question, if not PM me with more information about what you're having trouble sourcing and I'll see what I can do.

Steve.


P.S. For anyone coming to this thread late - this is the post that explains it all:-

http://www.cnczone.com/forums/showpo...5&postcount=45

Kwackers, I built the board ( point to point wiring on Veroboard ) and all I get on the LCD display is a row of squares any ideas? I have checked and double checked everything I used a JDM programmer to burn the .HEX file onto the chip .... anything I may have missed?

Quote:

---

Originally Posted by Ben Olson

Kwackers, I built the board ( point to point wiring on Veroboard ) and all I get on the LCD display is a row of squares any ideas? I have checked and double checked everything I used a JDM programmer to burn the .HEX file onto the chip .... anything I may have missed?

---

Could you have the contrast bias wrong? I have not looked at the schematic so I don't know if that is called out specifically or not. Of course, the standard adage also applies.. check your connections!!

Also, are you POSITIVE you have the device properly programmed? Is the xtal oscillating? pwer and ground to the cpu correct? the variable resistor for contrast on the lcd, is it adjusted properly? if you have a scope, can you see the lines to the keypad toggling, which would indicate the keypad is being scanned and therefor the micro is actually running?

Just a few things to look at.


Horsedorf

Quote:

---

Originally Posted by Ben Olson

Kwackers, I built the board ( point to point wiring on Veroboard ) and all I get on the LCD display is a row of squares any ideas? I have checked and double checked everything I used a JDM programmer to burn the .HEX file onto the chip .... anything I may have missed?

---

HI Ben,

A row of squares means the LCD isn't being initialised (if both rows are full, check the contrast).

Assuming the chip is programmed properly there's not much can go wrong.
Check the area around the crystal.
Check the connections to the LCD.
Check pin 1 (mclr) is at 5v.

I assume you verified the chip when programmed to see if it matched?

Finally check the config bits on the PIC - it's important they match otherwise the PIC may not start.
Look at the image attached to this post and see if your config matches this.
http://www.cnczone.com/forums/showpo...9&postcount=23

Let us know how you get on.

it looks like I have some checking to do , I did verify the chip ( using IC Prog ) but I was unsure about the 10k variable resistor ( I am an electronics newbie )

I have to turn the pot all the way one direction to see the squares

the power supply to all the pins on the IC seem to be OK

I am a little unclear on checking the config bits is this the same as verifying?

thanks , Ben