Machsupport Forum
Mach Discussion => General Mach Discussion => Topic started by: Katoh on January 22, 2013, 09:04:31 AM
-
G'Day Friends
Just going through little problems on this lathe build one by one. Have solved many a problem but there is one recurring at present.
I have a slotted disk on the spindle with an opto, (recommended by Hood) in a previous thread, all is good there wiring is fine and all shielded to BOB. I'm using a Homman BOB and the little led next to the input flash's every time that slot passes through the opto without fail, even to turn by hand, 1rpm lets say it will flash on the board. Now I don't know why but when you go into Mach, on the diagnostic page it wont show the index input the way it shows up on the board? The spindle input light is actually always activated it shows the rpm but erratically, Now my spindle can turn right down to 30rpm, there is no chance I can get that reading in Mach, anything under 100 is unstable, I have tried the debounce settings to some very high numbers, it will then read down to 46 rpm but then at higher rpms the readings play up, and I have read its best to keep the debounce as low as possible.
I have read also to turn off spindle relays, this is unacceptable as I will loose my M3/M4 command to the Homman DC03 Digispeed which is working correctly. I have had to add a capacitor to the E-Stop and Limit switch inputs to stop noise setting them off and also to the 0-10v control to the VFD which gives me a very stable motor speed controller.
Question
Firstly when the led on the BOB registers the slot passing the pto, should it not show up on the inputs on the Mach screen on every pass not just every now and then?
Would adding more slots to the spindle disk help?
Is this a noise problem ?
Or is a config problem?
I realize this has being posted before but there solutions in prior threads have not worked in my case.
Thanks
Katoh
-
Dont worry about the index LED on the diagnostics page, that is due to the refresh rate of the screen and the update rate of Mach, it is just an indication.
Regarding Debounce, it is the Index Debounce you adjust, probably what you were meaning but just saying in case ;)
Set the Index debounce to 0, also go to ports and pins, spindle setup and disable averaging and see if that helps.
Hood
-
Sorry for the late response, I have not had a chance till now to test those things.
Disabling/ enabling speed-average did not do much if anything, the Index debounce if set to 200 will read down to 42rpm (great) but will not read over 150rpm (bugger) actually the faster you go the slower it says it is, Go figure. Go the other way with the debounce minimum 50 will give me good rpm higher but nothing under 100rpm.
tried a cap between input pin 15 from opto to 0v think it made things worse. I have wired the opto as described in previous thread,
Black and green connected to pin 0v
blue input pin 15 (index)
White +5v from BOB
Red + 5v from common power source with a inline 220Ohm resister to limit current.
Thanks
Katoh
-
I am not sure how the parallel port behaves regards reading the Index at low revs nowadays. It has seen a lot of changes since I used it many years ago so it may just be these changes have made it react that way.
I tested out yesterday with the smoothstepper and set the spindle to 1rpm and the DRO read 1 then 0 then 1 then 0 etc, same up to 25 rpm, it went 25/0 but over that it sat at a steady rpm. When however I enabled averaging then I got the numbers jumping around when under 30, eg the 25rpm would move around from maybe 12 to 25 continually counting up and down.
This however likely has no similarities with the parallel port way of doing things especially if you say averaging makes things worse.
sorry cant be much more help other than to say its unlikely a problem with your input.
Hood
-
Hood many Thanks again for input, from what you describe I don't seem to have the same issue with the low rpm, mine just is stable and 42rpm, that is the lowest it goes, even though as mentioned earlier the Led on the BOB flashes without fail ever time that slot goes past the opto.
Do you think adding extra slots would help?
Or did I put the capacitor in the wrong place Maybe should have been between the red and Blue or the red and black or the White and blue, I should just check the lot and see what works.
Thanks Muchly
Katoh
-
Adding more slots will not help as you can only use a single slot in Mach now, well not strictly true so I will elaborate.
Previously you could use a disc with many slots but have one wider as the index, this was set up as a timing signal rather than the Index. It was found however to be no better and likely even worse with regards threading so it was abondoned and now the only valid setup is the Index.
You could have a disc with two or more accurately milled slots and use the spindle pulleys ratio to make Mach read the rpm as 1 per rev. Would it help? I have no idea but it may at the lower revs, the problem you may have however is at higher revs the parallel port input is not fast enough to see the multiple slots and thus you get erroneous readings.
I do not think it is a problem with your input, I think it is likely the way Mach does things in the driver but hopefully Rich or someone else that uses the parallel port can confirm or deny that.
Hood
-
Adding more slots will not help as you can only use a single slot in Mach now, well not strictly true so I will elaborate.
Previously you could use a disc with many slots but have one wider as the index, this was set up as a timing signal rather than the Index. It was found however to be no better and likely even worse with regards threading so it was abandoned and now the only valid setup is the Index.
You could have a disc with two or more accurately milled slots and use the spindle pulleys ratio to make Mach read the rpm as 1 per rev. Would it help? I have no idea but it may at the lower revs, the problem you may have however is at higher revs the parallel port input is not fast enough to see the multiple slots and thus you get erroneous readings.
I do not think it is a problem with your input, I think it is likely the way Mach does things in the driver but hopefully Rich or someone else that uses the parallel port can confirm or deny that.
Hood
I can understand that, cannot really understand how one slot makes it just as accurate as multiple slots, but that's not why we are here at the moment. You would think with the Bob identifying every pulse sent by the opto switch that it really should be a piece of cake, but something is screwy somewhere.
I read that earthing the computer chassis to the machine earth helps also in dealing with noise, but maybe this is not a noise problem and the pulsed signal is not actually even reaching the Mach?
Its all speculation at present but something has got to work surly.
Cheers
Katoh.
-
I dont think its a problem with your signal it is, I think, likely to be the way the parallel port driver processes the signal.
Hood
-
Hood
Just doing some further research I came across this thread from 08, with particular reference to page 3 post by SimonD,
http://www.machsupport.com/forum/index.php/topic,6707.20.html
Do you think I'm looking down the barrel of the same sort of thing, maybe?
Cheers
Katoh
-
I am not unerstanding what you are meaning? Your index is being seen fine abve 30rpm and you will not be threading anything at anything like these RPMs so dont think you will have a problem in that respect. Maybe that is not what you are meaning?
I will see if I can find the info Art posted regards the Timing via Index.
Hood
-
Hood I apologize don't mean to confuse the issue, It's probably more me being pedantic more than anything else. In my thinking, if it works at 40rpm it should at 2000rpm but doesn't and vise versa.
For some reason I'm assuming that something is not quite right in the setup, so I go off reading all these other threads where they used multiple slots and timing and capacitor's with bells whistles and jam & scones for morning tea to make it work.
If I'm being unrealistic, please say so, maybe it just wont give me low and high rpm or the full spectrum.
I still have not tried at max RPM which is 2000.
Thanks
Katoh
-
Set Index debounce to Zero and see if you get the RPM correct above the 100rpm up to top speed.
I dont think there is a problem with your input, I am thinking it is the parallel port driver and the way it does things.
Hopefully Rich or someone else that uses the parallel port will say if they have the same thing.
Hood
-
It's been a long time since actual testing but need to find my notes on the subject.
I use 53 rpm on one of my lathes and i know that The Sherline was used below that.
more after i do some looking into my vaults......
RICH
-
Hood I tried what you suggested, setting the debounce at 0 and and seeing what I got .
I could not get a constant reading at all, then I started from 1 and work my way up on the debounce, it took to 41 to acheive a stable reading at high rpm.
Something sounds a bit screwy now.
Thanks Again
Katoh
-
Is this Index Debounce?
Yes that does sound weird if it is as on any system I had anything over 0 would degrade the reading and anything near 10 would not be readable at all. Maybe things have changed in the driver since I use the parallel port but cant see it.
Tell me a little about your wiring, ie I think you said its shielded, where is the shield connected? Can you try another input on your breakout?
Hood
-
Hi Hood
That's definitely Index Debounce, I'm using pin 15 as the input on the bob, the wiring is shielded, actually its 4core shielded microphone cable, nice cable you can buy by the meter from Jaycar in Oz. The grounding of the shield is from one side only, the other is capped off as close as I can to the the opto, say 2" or 50mm. The grounds are all common as with the neg side of all power supplies all run from separate buss bars or points.
Ill include a couple of pics.
The only thing that I can think off at the moment, is my spinning disk is made from a CD, maybe reflection, refraction, or just do the locomotion (bit of humor here) but I have read that with the opto I'm using it makes no difference if its a reflective surface or not. I have painted my black anyway.
As I said in prior posts the BOB picks up without a problem every pulse sent by the opto, just mach cannot define it. Maybe I have to go for timing instead of index, maybe a different input pin would do better, maybe Peter Homman could shed some light on this its his bob's I'm using?
Cheers
katoh
-
Timing can not be used now as far as I am aware. Changes were made in the driver quite a while ago. I have looked for the info Art posted about Timing not being any better than an Index but as yet can not find it, I was sure I had copied it so I wouldnt forget, seems I have forgotten where I stored it ;D
We will see what info Rich comes back with, hoepfully he will be able to test.
I think you have said but just to make sure, are you using the parallel port for this machine?
Do you have a second parallel port that you could input the signal via so that it goes direct rather than through a breakout board? Just be carefull if you do that that your power supply is either taken from the computer or the 0v are tied to computers 0v.
Regarding DC supplies 0v being grounded, I have never done that although I hear of people doing it.
Hood
-
Timing can not be used now as far as I am aware. Changes were made in the driver quite a while ago. I have looked for the info Art posted about Timing not being any better than an Index but as yet can not find it, I was sure I had copied it so I wouldnt forget, seems I have forgotten where I stored it ;D
Hood
Yes I don't what it is but I seem to have the same problem lately, "do you think we are just getting old" heaven forbid, because I still feel like I'm 28, only problem last year I felt like I was 21?
Regarding DC supplies 0v being grounded, I have never done that although I hear of people doing it.
Hood
I think you misunderstood me there, the ground and the neg or both separate but all power sources have a common Neg source through a common buss bar. The same way you would wire a boat if that explains anything?
If you look at the photo control box, you will see a group of contacts say 1/3 of the way up, the green is the earth bus bar connect to earth, the red is isolated 5v power supply, the blue is isolated 12v power supply, the black is the neg bus bar containing all the neg sides of each power supply into one point, then the neg side to each BOB is taken from here. All positive sides or active sides are separate for each power supply.
I forgot to post the pics in the last post so here they are.
Yes I'm using 2 parallel ports and 2 BOB's one is for inputs only and the other for motor outputs including spindle.
Thanks Again
Katoh
-
Ah ok I see now what you are meaning and that is how I would do things with maybe one exception, I would be using normal Earth Din rail terminals which would connect the base plat of the cabinet to Earth, might be worth trying connecting an earth wire direct from the Earth terminals to the base plate.
Hood
-
Katoh,
What version of Mach are you using?
Found most of my notes.....quick look at over 100+ pages of stuff.
Good grief.......I will post when sorted out.
Hope to try low rpm's tonight.
Later,
RICH
-
Hood, you got it spot on now, but yes from the earth buss bar if I can call it that it has a a way larger lead back to the cabinet and the actual lathe frame which is also connected to the main Ac earth wiring also, if that makes sense. I reckon (believe) I could easily confuse myself at times!
RICH, I'm using the latest lock down version of Mach, I believe it to *********.066 I downloaded and reconfigured the lathe about 3 days ago, after hood pointed out I'm using the wrong screen-set. Problem fixed now.
Gentlemen do you think maybe possibly it could be an input problem, say active high vs active low, I realize this in reality is the same as all it is meant to is be a trigger but maybe just maybe index needs to be opposite? The other thing could be, for example on my router the touch probe I could not just connect to the the BOB, I had to make a up a separate little board for it work correctly, using a few resistors and cap if I remember right. But the point with that is I followed some else s design to make it work. I'm a Civil Engineer, not an electronics engineer, I can't work this stuff out, thank goodness there are people who can.
Cheers
Katoh
-
Easy way to see if active state is correct is look at the diagnostics page and rotate by hand. If the LED is lit when the slot is allowing the LED to pass to the reciver then your active state is correct, if the LED is lit and goes out when you are blocking the signal then that is not right. Would still probably work but may not get a long enough off period for it to work effectively.
How wide is the slot on your CD? Also dont think a CD is ideal but more for the flex it could have rather than the reflective nature.
Hood
-
Hood, you may be onto something here, as I have it setup with the index light on in the diagnostic screen, then switching it off when the notch passes the opto. Ill have a chance later this evening to change it round the other way.
The notch on the disk is 7mm wide. I used a formula that I found in the setup tutorials for notch size vs disk diameter vs max speed. The CD is actually great it spins true and is quite rigid, once bored out to fit the spindle, there's not really that much if it left.
Cheers
Katoh
-
I always had my slot 10mm wide, why I chose that I am not sure, probably just the cutter I had ;D
Hood
-
A little update.
I changed the settings on Mach and also on the board so now I have the index light on when the slot passes the opto, no slot no light. hold the slot on the opto and the index light stays on, move the slot it goes off. all good I hope.
It may be an improvement but still a problem in there, somewhere.
What I'm finding now is at max RPM 2000rpm I can get a stable reading with the debounce at 8 lowest value, but any rpm lower forget it. At lower rpm say 100 we need a debounce of 80, but still it will not give me an rpm over 500 or under 100.
I'm starting to run out of ideas, any more suggestions?
Cheers
Katoh
-
I am afraid I dont, it is just so strange that you need the Index Debounce so high. As mentioned any time I have used the Index pulse via the parallel port anything above 0 Index Debounce and it has made things worse.
Is there any way you can temporarily bypass your breakout board and connect direct to the parallel port? Even if that means temporarily disconnecting other I/O.
Think you mentioned how you were powering the opto but would have to read back to see so just make sure your 0v to the opto shares the 0V of the computer (or breakout if its isolated)
Hood
-
I'm back to stage 1 here. the opto is code 121-7280 can be found on this page, about halve way down.
http://au.element14.com/jsp/level5/module.jsp?moduleId=en/206215.xml&sku=1173631
wiring as follows, black & Green are connected to one wire to pin 0v BOB,
Blue connected to pin 15 Input / Index
White is connected +5v
Red is connected to +5v with an inline 220Ohm Resistor.
Again I am confused as the BOB led for the pin 15 input, registers each pulse without fail.
Maybe its simply a matter of increasing the size of the slot in the disk? It's not that hard to make another disk, maybe I should increase the size and add another 3 to 6 smaller slots and see what happens, maybe add 7 more go for Octo Disk, can only be defeated by "007" you never know your luck in the big smoke , it just might work!
Apologies talking stupid again, I have just run out of puff, maybe time to walk away from this problem for a while.
Many Thanks Again
Katoh
-
Only thing I can think of is signal is not passing through the BOB correctly.
As said in the previous post I would try and connect direct to the port and see if that helps, remember power must share the computers 0v if you do it that way.
You can add as many slots as you want but as far as I know Timing is no longer in use so you would be wasting your time, that however may not be the case and maybe I am remembering incorrectly.
Hood
-
Timing is no longer supported, don't activate it, use single slot. Single slot is just as accurate.
Use the formula for the slot width as a guide. I found i had to
make it wider if the higher rpms were not seen. Use a flat black painted CD for the disc on the Sherline, in fact have one disc with multiple slot widths of varying widths just to play with. Put black tape over the slots so only one is seen.
Debounce here was usualy 10, below that was not good and above, say 100 that took to long to confirm the index signal.
Each system is different so just need to play. Lower the better. I changed to Hall sensor and found that to be far supperior. BTW, not all Hall sensors are the same for sensing time, some are slower and you won't be able to use for high rpm's.
By pass the BOB, make things as simple as possible so you know if something is affecting the index.
I'll be out and about for and back in a few days. Didn't get a chance to play around last night nor get a few definitons
asked about.
RICH
-
Thanks Hood
Thanks Rich
What Ill do, is bypass the board, and just connect pin 15 input and 0v straight up, see what happens. I'm lucky in the respect that I'm using two boards from for inputs and one for outputs, Ill leave the outputs the way it is so I can still control the spindle.
I will post my findings over the weekend, after having the chance to do this.
One thing Ill ask before surgery, would the computer kernel, speed have anything to do with this?
I have a big Bridgeport mill sitting under my verandah, my next NC project. Goal is to have the router, lathe, and the mill, but seriously I'm thinking that using Parallel Port on the mill is going to give me grief like this, I'm be looking for a better system.
Once again Thanks Muchly
Katoh
Just to add I might send Peter Homman an Email regarding the situation, he might shed some light on this matter, its his BOB's I am using.
-
Parallel port setup is usually easy and straightforward, why you are having problems I am not sure but it could be the Bob and that is why I suggested connecting to the port direct if possible. As you say your setup of having one BOB for outs and the other for Ins makes that easier.
If your Bridgeport is going to have servos then you may be better going with an external controller, just depends what you will have and what you are wanting as to which one to get.
Hood
-
It's quite easy to see why I'm having problems, its called Karma! :D
Laughing with a friend how said no one has got there machine to run not mention just to move, straight of off the bat, I boasted this one will, wiring is all spot on, no problem's here Mate! :'( Famous last words
My router is a bit of a Bitser, bit of this and that, CNC4PC boards, Homman speed controller, Probe by (?) forgot his name, but each of them had a problem to make them work together, but There was a solution.
I thought this time make it easy on me I'm going all the one brand items that are made to work together, All good till now, and to be honest I think it might be a signal issue from the bob to Mach also.
I will try the bypass of the BOB. If it works I know where to go from there.
The Mill is still in planning stage, I am thinking maybe smoothstepper system, with a servo spindle? Plenty of time to think about that one.
I really appreciated all your help! :)
Thanks
Katoh
-
;D serves you right ;D
If going servos then I would say an external controller would be best simply due to the PP's limited kernel. Which one is personal choice. The Smoothstepper works well but for me I like my machines to have differential step/dir signals and 24v I/O. That meant I needed to make interface boards up but now the CSMIO controllers are on the scene they suit me much better and make things simpler.
Hood
-
Hooray :)
Finally something that actually worked. I disconnected the Bob and went straight into the cable pins, located pin 15, and pin Common, and just moved the wires from the board to those pins. Big Big improvement, Still must say I have nothing but inaccurate readings below 40rpm, but with a index debounce setting of 5 I can go from 40 to 40000 without a problem, how accurate I'm not sure but at least stable.
Culprit was the BOB, so what to do now and how to bypass for one pin, or is it good idea to bypass?
I think I must have have a talk to Peter Homman first before more drastic measures are taken.
Cheers
Katoh
-
Ok some progress, talking to Peter is definitely the thing to do. If you do need to bypass you could cut the cable and get a new plug and solder it on but leave one wire unconnected for your opto connecting direct to. No problems connecting direct, especially if you power from the computers own 5v. I ran mine on the Bridgeport for years direct to port 2.
Hood
-
I sent Peter an Email just now, describing the situation, hopefully he will get back to me soon.
Just speculating maybe he has an assigned pin on his board specially for this, but I see an input as an input regardless of pin assignment, I on the offer could be wrong, hasn't happened before, apparently I was just mistaken.
I do like the CSMIO kits, I'm scared to ask how much, but the six axis unit looks like a great bit of gear.
Cheers
Katoh
-
Price is on their website priced in Euros €,
http://www.cs-lab.eu/en/sklep-dzial-1,-,-online-shop.html
Looks expensive at first but when you consider its all in one then it doesnt look so expensive, for example there are analogue inputs and outputs so you dont need a spindle motor board if using +-10v to a VFD. The analogue inputs can be used for Feed and Spindle overrides so no need for any additional means of getting analogue inputs. There is no need for breakout boards as the I/O is all isolated, so thats basically 3 or more breakout boards not needed. (32 Digital ins, 16 digital outs, 4 analogue ins, 2 analogue outs and six axis of step/dir signals)
The additional MPG unit is obviously extra but it has the best MPG motion I have seen and it works like it should, ie has x1 x10 x100 (x10 x100 x1000 setting for imperial) so each click will either be 0.01mm (0.001") 0.1mm(0.010") or 1mm (0.100") and is silky smooth and instant, ie wind the MPG fast in any of the 3 modes and stop winding and the axis stops, no buffering of motion like happens with a MPG via Mach.
Hood
-
I will definitely look into one of those units for the mill, very nice indeed! Just need to finish the lathe, the boat, re-seal the bobcat, do a bit round the house, yard, go to work, ah! almost forgot keep the family happy, so I maybe planning sometime around May-June. :'(
2020
But seriously I going to look into one. The Mills going to be another kettle of fish altogether, I'm also looking into steppers with feedback, now won't that be nice. Enough said on the mill I will start a new thread when the days get closer, or Ill ended hijacking my own thread. :D
Katoh
-
Steppers with feedback have always tempted people but I dont see any benefit in them. If a stepper loses position it is because it has stalled due to being overloaded, so no way back except E-Stop.
Hood
-
I will be looking forward to Peters reply .
-
Hey Katoh, I was playing around with my little MicroMill today and hooked up a C3 index card. It works great all the way down to 36 rpm. Below that & my KB speed control gets jerky. 36 rpm is VERY slow by the way! The top speed is a bit under 3100 but I'll get it up a bit higher later when I take the back cover off the enclosure and dial up the top speed pot on the KB. They have quite a bit of adjustment range and I don't think I'll hurt the motor running the spindle at 4000 or so since my cycles will be short.
I checked the accuracy by warming up the spindle/motor/control for 10 min or so and carefully adjusted the pot (the mill has a manual speed control pot) to 60 rpm. I clicked on my stop watch and counted the hole in the disc going by a mark 60 times and clicked off the stopwatch....just about dead nuts 60 seconds showing. Haven't checked the top speed with my digital tach yet but I'll bet it's right on.
At 35-40 rpm the LED on the diag screen pulses every 2nd or 3rd rev even though the C3 LED is dead solid so what Hood said about Mach/screen refresh is definitely true (as if there was a doubt if Master Hood pronounced it to be so.) ;D
I cut down a CD to 3 1/2" and drilled a #12 hole about 3/16" in from the edge of the disc to the center of the hole and painted the disc black. I couldn't tell any difference between 0 and 1 on the debounce setting. I'm using the parallel port on an old Dell P4 800mhz and a cheap, non opto-isolated B.O.B. My ORAC lathe has a CNC4PC C11 B.O.B. with a C3 index card and it works great down to a crawl as well. I have tested the top speed on it with the tach and it's within 10 - 20 rpm at 2000.
-
Dickeybird,
Thanks for taking the time and posting as I have been out and about for a number of days.
RICH
-
No prob Rich, it feels pretty good for a change to successfully get a project working that I set out to complete. Pretty simple stuff to ya'll but dangit, I can find more ways to screw up things!! I'm sure next weekend I'll be back to business as usual. ::)
-
GDay Dickeybird, Rich
Sorry for some reason I never got notification that you had posted. weird?
Dickeybird if I bypass the BOB I get much the same results as you, which only points to the BOB. I'm waiting to here back from Peter Homman to see if he can shed some information on this or maybe he has a special pin that needs to be used or some other setting that needs to be set. But without going through the BOB I get an accurate reading anywhere from 41 to 2000rpm tested fine but I still need a index debounce of 5, not to shabby I thought. Considering I had to add a cap to a few inputs to stabilize them, its not to bad.
Cheers
Katoh
-
Katoh have you heard from Peter yet
-
No all quite on the southern front! Maybe he's on holidays still?
I don't really want to go cutting and moving single cables yet,to bypass the BOB especially if there is an easier work around,
Cheers
katoh
-
Gentlemen
It has come time and cannot wait any further, so I took the easy fix method. I removed the parallel port cable from the BOB and and simply removed cable 15 and the last common cable, drew them back through the plug and simply bypassed the BOB.
I reconnected the the opto cables to those I removed, but I re-connect to the Bob only the common, just to make it common? Does that make sense?
Anyhow with the Index debounce value of 1 I have accurate rpm feedback from 41rpm to 4000rpm. Fantastic that's all my wiring done! "hooray"
Now I'm simply left with two jobs on the lathe itself, one being the cover for it, and the other looking at coolant. Once they are finished I will post some happy snaps of my latest project, all working.
Tooling I'm looking at the moment at quick change toolpost, and indexable carbide tools positive rake only, preference would be DCMT inserts. What would you suggest for a CNC lathe 10x22" the only hitch spindle power is only 1hp, the machine takes 12mm/1/2" holders, plus an internal and external threading bar.
Many thanks
Katoh
-
Any reply from Homman re the board wireing . I think I am having the same type of problem
-
No nothing, I will send him another email, actually I would like to now the solution also, I don't think the bypass as I have done is correct method.
Cheers
Katoh
-
Hi Katoh,
Sorry, I had not been monitoring the forum for a while and have been a bit busy.
I'll have a read through the thread and see what I can find.
What index sensor are you using?
Cheers,
Peter
-
Hi Peter
That's Ok, this time of year with school holidays and everything else, life gets a bit crazy.
I'm using this switch "recommended in a prior thread"
http://au.element14.com/optek-technology/opb916bz/opto-switch-slotted/dp/1217280?Ntt=121-7280
My wiring I have described in the thread.
hope this helps.
Cheers
Katoh
-
Hi Katoh,
Looking at the datasheet for the OPB916B sensor, it has an open collector output with a 10Kohm pull-up resistor. When activated, the open collector transistor pulls the output line (blue wire) to 0V.
So you MUST have the MB-02V6 inputs pull-up jumper set to 5V (pos 2-3). Is this the case? If, not it won't work.
Cheers,
Peter.
-
Hi Peter
So you MUST have the MB-02V6 inputs pull-up jumper set to 5V (pos 2-3). Is this the case? If, not it won't work.
That's where I have the input set to, to 5v.
I can get the signal from the opto to the BOB and it does reach Mach, The problem is to get low RPM, under 100 I need an index Debounce of 80, but for high rpm 2000 the index debounce is set to 8, but wont give me a reading under 1500.
If I bypass the BOB which I have done now the index debounce is set to 1 and I have stable rpm reading's from 40 to 4000 rpm. The little led next to pin 15 that I used, flash's without fail every time the switch is triggered, without fail be it 1 rpm or 1000rpm, just something goes wrong sending the signal to mach.
Everything thing else works fine, I have 12 other inputs on that board all are good, maybe its loosing a bit through the board, enough just not to give a good reading.
Thanks
Katoh
-
Hi Katoh,
Is the opto exposed to fluro light?
If so can you test it in the dark?
Cheers,
Peter.
-
No the Opto is the lathe gearbox / pulley box. Covered with the steel door.
Katoh
-
Hi Katoh,
Try changing the 220ohm resistor with a 180ohm or 150ohm one as the MB-02 has 1K strong pullups on the inputs.
Cheers,
Peter
-
Hi Peter
I will give that a go, but please bare with, it will take me a couple of days to test at the moment.
Thanks
Katoh
-
Hi Katoh,
OK. In the meanwhile I'll see if I can get a few of the optos and do some teting myself.
Cheers,
Peter
-
Hi Peter
Just had a chance to test now, went out today and bought a new 180Ohm and a 15Ohm resister and replaced the 220Ohm.
Had to go through pin 13 is pin 15 is now a direct input to the computer via the cable, I was really hoping for good news but it did the same as before. There was no difference in readings whatsoever. The same scenario, again bypass the BOB and I get stable readings.
Again the little led corresponding to pin 13 flashed every time the switch was activated without fail, but for some reason the message to mach was scrambled eggs.
The funny thing here is even through the board if I turn the spindle by hand and the index light gets triggered on the diagnostic screen, but only if I turn slowly (something to do with screen refresh rate as pointed out by Hood), It works exactly the same when you bypass the BOB. But still the readings are everywhere through the BOB but stable without it.
Got me beat!
Cheers
Katoh
-
Hi Katoh,
The only thing I can think off is that the BoB has 1Kohm pullups on the input lines which may be too strong for the sensor to pull to ground.
I should receive my sensors on Monday or Tuesday so I'll test it out using a Scope and see what is going on.
Cheers,
Peter
Cheers,
Peter
-
Thanks Peter
If you work it out it be great!
I still have the Bridgeport mill to convert next, once the lathe is running and finished.
Cheers
Katoh
-
Hi Katoh,
OK, I received my sensors and hooked it one up to a MB-02V6, I used pin 10. The doide was connected between 5V (red wire) and 0V(Black wire) with a 390R resistor.
For the output, VCC (white wire) was connected to 5V, Ground (green wire) was connected to 0V, and output(blue) was connected to pin 10 input terminal.
I connected the oscilloscope to the input pin 10, and there was a very clean signal right through the rpm range. I then connected the oscilloscope to the parallel port input for pin 10 (via the 26pin box header on the board) and the signal was identical, very clean.
Do you have an optivisor or magnifying glass? If so have a look at the resistor packs on the board especially around the idc26 pin box header. Look and see if there are any solder bridges. It's the only thing I can think of that would cause a problem.
Let me know how you go?
Cheers,
Peter.
-
Hi Peter
Sorry for the late response, Work always seems to get in the way of a good project :'(
I could not see an problems with the board, I get the same results through the other board as well, so I don't think the boards to blame here.
Here is a little diagram of the wiring, I tend to read your wiring differently to what I have done.
What do you think?
Cheers
Katoh
-
any luck with it yet
-
Otpco
The only way I can it working is still going back to bypassing the BOB, through the bob things get scrambled, maybe someone will work out why. The signal gets through but Mach goes crazy?
Got the lathe cover built, looks good so far, and then its just a matter to sort the coolant side out. Then I'm off and running.
A small divert from the original thread, just something to do with tooling for the lathes. I made a mistake when I bought my tooling, as all my tooling takes the same insert a dc style but size 07. I should have spent a little more time and found all the tooling that takes as before the dc inserts but size T3. can't understand it but you can buy top quality brand name inserts in that shape and size on Ebay for near nothing and they are plentiful, where the slightly smaller sized 07 are fewer and at a premium price.
go figure!
Cheers
Katoh
-
The DCMT0702 come up now and again on eBay here but the DCMT11T3 are certainly more common, I am finding however that nearly all inserts appearing on eBay are starting at prices that are getting close or often even higher than I can buy new from a dealer. Not sure what kind of prices you pay down under but here its usually about £30 to £40 for 10 and I have seen crazy starting prices on eBay recently as high as £60 ::)
Hood
-
Hood
The prices I'm finding here are around $5 for 10 inserts with shipping around the 4-5 dollar mark. Say $10 Au all up that's about 6-7 pounds in your language.
The T3 sizes are quite plentiful but the 07's don't come up around that price very often. The average price is between 40-55 Dollars Au depending on breed, still a lot cheaper than what your paying.
Katoh
-
Ebay used to be like that here but no more :( Occasionally you will get a good deal but its getting fewer and fewer.
Hood