Compact and Simple ...
Neil Wyatt | 07/08/2018 19:23:01 |
![]() 19226 forum posts 749 photos 86 articles | I have to be honest, I wouldn't use a PLL, I'd just time the rotations accurately with a hall sensor and do the rest in software. a 20MHz timer is going to give you plenty of resolution. Neil |
Joseph Noci 1 | 07/08/2018 20:24:25 |
1323 forum posts 1431 photos | A hall sensor and a 20MHz timer just won't cut it I fear... The easiest and most optimal method is to use the DDA algorithm - an extension of Bresenham's line draw method. It's simple, works for ANY combination of encoder count value ( sensible encoder values, that is - use at least a 1000PPS encoder) and any number of gear teeth on the blank. Very easy to implement in software, but a 16MHz Arduino will not cut it either! Hence my use of the Nucleo with a small STM processor @ 120MHz - The Nucleo module cost around US$15.00. I can give you all that software if you wish Michael. A PLL implementation 'does' work, but it's a nightmare working out all the diveders, etc, if done in discrete logic. Also, PLL's are a pain to make stable - phase lead and lag can kill it. A digital PLL is better, easier to stabilise, but needs a really fast processor to up the sample rate. The DDA method is used in all modern CNC threading and gear cutting implementations. Joe Edited By Joseph Noci 1 on 07/08/2018 20:33:13 |
Michael Gilligan | 07/08/2018 22:00:20 |
![]() 23121 forum posts 1360 photos | Posted by XD 351 on 07/08/2018 19:05:33:
https://youtu.be/ZhICrb0Tbn4 Doesn't give any plans but does show his general set up and software he is using . . Another thank you MichaelG. |
John P | 07/08/2018 22:24:59 |
451 forum posts 268 photos | Hi Michael |
Michael Gilligan | 07/08/2018 22:50:34 |
![]() 23121 forum posts 1360 photos | Posted by John Pace on 07/08/2018 22:24:59:
Hi Michael
. That's another very helpful response, thank you John I'm doing rather well on this thread ! My hob was just a one-off purchase from a German supplier on ebay ... perhaps rather whimsical, given that I don't yet have a machine; but whilst I was browsing for ordinary MOD_0.8 or 32DP cutters, it cried 'Buy Me' MichaelG. |
John Haine | 08/08/2018 09:37:30 |
5563 forum posts 322 photos | I can't agree that a PLL is tricky in this case. There will only be a limited range of spindle speeds - possibly only one, and a fixed division ratio between the VCO and the spindle pulse - following JS's practice perhaps 4000, or 4096 to be convenient in binary. The 4046 has an edge triggered frequency/phase detector which makes locking pretty reliable provided the VCO frequency is well centred. It may jitter rather, but there is also a pretty large division ratio between the VCO and the spindle carrying the blank so the phase errors will be reduced by this ratio. It's true that one will not be able to stop the milling spindle and restart it in sync but I'm not really sure that's a requirement - anyway it may be possible to add some such synch as a refinement. The programmable divider is simple, the MEW article had a circuit for one, or it comes from the application notes for the ICs - 4000 series CMOS has programmable dividers, in fact the 4059 chip will divide by any number between 3 and 15999 but I think has to be programmed in binary which is a bit tricky. Yes the whole thing is possible in software if you are skilled at real-time programming and using a processor with high enough clock speed, but such skills are rarer than building fairly simple logic circuitry using 4000 series devices on strip board. |
Neil Wyatt | 08/08/2018 12:29:09 |
![]() 19226 forum posts 749 photos 86 articles | Posted by John Haine on 08/08/2018 09:37:30:
Yes the whole thing is possible in software if you are skilled at real-time programming and using a processor with high enough clock speed, but such skills are rarer than building fairly simple logic circuitry using 4000 series devices on strip board. I'm not sure that's true these days. Show a 'maker' a quad NAND they will (a) ask where the SPI pins are and (b) if they do know what it is, they will say that you can emulate one on a microprocessor chip at a fraction of the price... |
I.M. OUTAHERE | 09/08/2018 03:41:23 |
1468 forum posts 3 photos | Wouldn't just be a matter of dividing the spindle speed by the number of teeth to be cut on the gear blank ? Isn't a hob is basically a thread with cutting teeth gashed into it ? If so one turn advances what ever it is engaged with by one pitch or in the case of a hob one tooth , if the spindle is rotating at 250 rpm and we are cutting a 25 tooth gear the gear blank would be rotating at 10 rpm to match the hob . What if the spindle was driven by a stepper motor & controller that is driven by a square wave generator ( crystal controlled ? ) couldn't the gear blank drive be exactly the same set up and use the same square wave signal that is controlling the spindle but fed through a microcontroller to do the frequency division then fed to the gear blank drive stepper controller and motor ? Or would the gear blank drive be better served if its square wave drive was generated by the encoder on the hob spindle ? That way it would sort of be able to sense the speed of the spindle and any variation because of the load when cutting . Just some thoughts ! |
John Haine | 09/08/2018 07:25:35 |
5563 forum posts 322 photos | You could drive the spindle with a stepper as you say, but normally the spindle is rotating rather fast for a stepper to cope with and give reasonable torque/power to drive the cutter. You mention having an encoder on the hob spindle - that's just what a number of the schemes described above do. The problem is that you need a very fine encoder to make the numbers work, so JS for example used one with effectively 4000 lines - 4000 pulses per hob rotation. Other schemes described here try to avoid this by having for example 1 pulse per rev and electronically multiplying this up using a phase locked loop, implemented either in hardware or software, then dividing down to drive the blank. |
Michael Gilligan | 09/08/2018 07:52:51 |
![]() 23121 forum posts 1360 photos | Please feel free to continue debating the electronic options ... Its.s all good, interesting , educational stuff. I should just mention though, that Joe's solution appears to fit my requirements very well. MichaelG. |
Mark Rand | 09/08/2018 11:53:03 |
1505 forum posts 56 photos | Quite a few years ago, I wrote the programs for some racing motorcycle (and car) combined rev-indicator lights and rev limiters. With a 4MHz PIC controller, I could accurately measure frequencies up to 200kHz (had calibrated correction factors to allow for the time taken to service interrupt routined etc). We even made a modified version to indicate the speed of steam turbines on barring gear down to 10 rpm, with an RS232 interface!
Anyhow:- I've got some ideas along the divider route to work as a modification to the recent lathe based gear hobbing setup described in MEW. I've got all the ingredients in various caches around the house and shed, just need to find the time to do it! |
Another JohnS | 09/08/2018 12:32:06 |
842 forum posts 56 photos | Andy Pugh is one of the LinuxCNC'ers who cuts gears with a hob - one of his videos is linked to above. He's very active on LinuxCNC (old name was EMC2) and is great at answering questions. Both on the mailing list, and on the online forum. A link to a wiki entry: Another youtube video: The software's all there, and it's proven, if that is the end goal you want. |
Martin Connelly | 09/08/2018 13:22:47 |
![]() 2549 forum posts 235 photos | I knew I had seen something about this on YouTube. It is a series of videos titled "Home Hobby Hobber" by F Cleff. Martin C |
John Haine | 09/08/2018 15:52:10 |
5563 forum posts 322 photos | Another reason to use LinuxCNC! It looks like he is using a gear as a spindle encoder with 3 sensors offset presumably each by 1/3 tooth and x-ored to get a x3 pulse rate as well - but impressive how LinuxCNC is effectively implementing a gearbox between the hob and the blank. |
Joseph Noci 1 | 09/08/2018 17:20:38 |
1323 forum posts 1431 photos | As some may know, I have made many add-ons for my machines. ELS on both my lathes, my NC Shaper, my 'computer' controlled Spot welder, My CNC Mill, etc...And That I use Mach-3 for the CNC mill, the CNC Router, the CNC PCB Engraver, the CNC Foam-cutter, etc..... It would seem that I am therefore not that dense... But I give up with LinuxCNC, in fact with Linux in all its glory, at that kind of level. I am pretty computer literate, do well in C and assembler, but you have to be born with that special (mis)gene to work Linux. Most of the time getting stuff to work and debugging processes is spent getting to grips with linux's obscure terminology and syntax and has one greps-ing all over the place. And in the end, it is but a tool, would it only realise that and behave accordingly... Anyway, the intention is not to hijack this thread with a linux group tussle - I just don't believe Michael needs the consternation of setup for the simple implementation he is after! Joe |
Another JohnS | 09/08/2018 19:55:04 |
842 forum posts 56 photos | Hi Joe; Thank goodness everyone has different strengths - I enjoy (and, am envious of!) your strengths that you show here. One of my strengths is the ability to develop software that runs internally on graphics chips, which means that I have to be able to easily write programs on Linux, Windows, Android, MacOS and IOS. Of those, I find Windows obtuse, confusing, and hard to do anything on, compared to the others! I think we should focus more on the ease of the end results rather on the specifics of getting there - there can be many paths to a destination. Regards - John.
|
Tim Stevens | 09/08/2018 20:00:18 |
![]() 1779 forum posts 1 photos | It seems to me that in essence, what you need is a drive linking the rotation of the workpiece to the rotation of the cutter. So, a different solution would be to fit a tooth-belt sprocket on the cutter shaft and a (much larger) one on the workpiece spindle, choosing the numbers of teeth to suit the gearing, of course. Then work out a system of pulleys so that the drive is tight, bearing in mind the wonderful flexibility of these belts, and the need for an adjustment of length, as well as the relative movement of cutter and workpiece. In other words, a meccano-type drive which I can understand rather than all this smoke and mirrors electric stuff. Ho hum Tim |
JasonB | 09/08/2018 20:10:54 |
![]() 25215 forum posts 3105 photos 1 articles | Tim, isn't that essentially what a Jacobs machine does which is what Michael said he did not want, the electronic solution avoids the need for pullies or gears in the case of the Jacobs. A couple of presses of a button and you have a setup for whatever tooth count you want rather than having to work out a gear train or pulley ratio then obtain suitable gears or pullies |
Mark Rand | 09/08/2018 20:12:35 |
1505 forum posts 56 photos | Only problem with that approach is similar to the problem with a conventional hobber:- you need a supply of sprockets/gears with the right number of teeth or which can be arranged to have the right ratios to generate all the gears that you want to make, including any prime numbers that might crop up. The electron shuffling approach allows you to get those ratios via the magic of arithmetic. |
Bazyle | 09/08/2018 20:59:10 |
![]() 6956 forum posts 229 photos | We have discussed encoders to provide pulses but, especially in a stand alone machine, at what point doe it become a better idea to drive the hob spindle from a stepper as well? This also enables the system to ramp up to speed in a controlled way which may help. |
Please login to post a reply.
Want the latest issue of Model Engineer or Model Engineers' Workshop? Use our magazine locator links to find your nearest stockist!
Sign up to our newsletter and get a free digital issue.
You can unsubscribe at anytime. View our privacy policy at www.mortons.co.uk/privacy
You can contact us by phone, mail or email about the magazines including becoming a contributor, submitting reader's letters or making queries about articles. You can also get in touch about this website, advertising or other general issues.
Click THIS LINK for full contact details.
For subscription issues please see THIS LINK.