[1hander]

hello again guys ... i got the ring made...and took a guess with the width of the slot... any body have any advice...maybe if you have the same kinda setup..the problem is that above 800 rpm or so, the rpm readout becomes erattic and the rpm climbs and keeps climbing until i hit estop..i have done spindle calibrate yet as i want to get a reliable rpm readout first..

the ring is 3 inches in diameter ...the slot is about 20mm wide....if you have a similiar setup.... my rpm maxes out at 10500...but ive decided to to only run at 8500..thats plenty.. maybe you could help me out with the following

1. how wide should the slot be
2. index debounce setting..

i saw a document withe recomendations on the width of the slot but i cannot find or remember where

[Kenny Duval]

What do you have the switch debounce setting at currently? If it's set too high the RPM index won't trigger at higher RPM's because MACH3 will treat it as a bouncing switch and filter it.

[1hander]

thankx for the response kenny...... i have the debounce set at 1 and have tried 0.... the lower i went the higher the rpm it could read but right now 2500 or so is highest it can read without going crazy on the rpm readout..

my settings are as follows

index debounce 1
using input 14 on port 2 on the index input
active low set on the c3 card
i reset my max rpm to 8500
ratio is 1.9 and max rpm set to 8500
the index ring is 3 inches in diameter and the width of the slot right now is 16mm wide
the led on the c3 lights on when the space in the disk comes around and goes off when the disk slot passes the trigger sensor IR
the pot on the c62 gives 0v at off spindle and 10v at 8500 rpm
i have use spindle averaging checked to on.....
closed loop spindle control checked to on......
i tried the spindle calibration but thats not going to work until i get a clean rpm reading all the way to max rpm of 8500 but the linearity is totally screwed up now... is there a linearity file i can delete

i really think its just the width of the slot that needs to be wider but im not sure... i cant find the document to calculate slot width..

[Kenny Duval]

Ok..I get that. Mine is a magnetic pickup and single neodymium magnet. Once I went about 10 on the debounce than the spindle would run away above about 1500 rpm. It could be a kernel speed issue as well. Perhaps you are generating more pulses at a certain point than MACH 3 can read at a given kernel speed.

[1hander]

im not sure about that...my kernel is set to 35000.... i read somewhere that in order for the rpm to read right ...that the kernel speed had to be set to 25000 do you know about that by chance... im looking sat the settings for the lathe indexing to see if thasst helps me out.

[1hander]

i found the formula for figuring out the slot widths ... buut its for setting up indexing on a lathe... im not sure if that will work on a mill setup...does any one know

[mikegiraud]

I'll chime in with my RPM sensing setup on my G0704:

1. G0704 Index Ring from Scott at Benchtop Precision modded to have Qty 1 0.150" diam hole rather than the stock 16 holes. Scott can provide this to you for $15 plus shipping. See his site, and email him. The single hole is needed as the parallel port can't handle multiple holes or slots well. The $15 was cheaper and faster than designing, and turning a custom ring on the G0602 lathe.
2. Panasonic SUNX PM-T44P photoelectric switch. These come in many form factors, I chose this SKU and machined a bracket for it. This model does not require a pull up/down resitor when wired to my C11G BOB. Its wired directly to the input and the 5V supply.
3. My Index Debounce is 4. This provides stable and accurate RPM readings to 6500 RPM in my setup. Kernel Speed is 25KHz. Above 6500 RPM I see some noise spikes in the RPM reading, but my motor/VFD limit me to 6200 rpm for sustained operation so I consider it a non-issue.

Link to the stock 16 hole index ring. Again, you need a 1-hole version. G0704 Index Ring - BenchtopPrecision.com

Attachment 360070


Pic shows the sensor and the single 0.150" diam hole.

[Kenny Duval]

That's essentially what he has with a single slot in the disc.

[1hander]

thanks for the response mike....my indexing ring has basically 1 hole in it... in the form of a slot 16mm side at the outer edge.. so am i undrstanding you correctly...your not using a c3 index card... you have the photoelectric switch wired directly to 5v and the sensor ouput is going directly to the BOB...

also im running at 35khz... im wondering if that is whats causing the problem... i ewas thinking about trying a different kind of sensor as well...
based on your setup..im going to change to 25khz and try that first.. b back in a second..

[1hander]

Not sure what happened but.. Now I'm not getting any voltage from the 10v analog output.. Ahhjhhhhhhh

[1hander]

im using MSM.... i was tring to delete the linearity files...so i went into the mymsm template and deleted the curve1.dat...curve2.dat...3,4,5 and the linearity.dat files.....would that cause the 10v analog output not to give an ouput when i hit the CW button... i replaced the files where they were and im still not getting anything...any ideas anyone..

does anyone know if cnc4pc does repairs or ami screwed i just bought this card not long ago...

[mikegiraud]

thanks for the response mike....my indexing ring has basically 1 hole in it... in the form of a slot 16mm side at the outer edge.. so am i undrstanding you correctly...your not using a c3 index card... you have the photoelectric switch wired directly to 5v and the sensor ouput is going directly to the BOB...

also im running at 35khz... im wondering if that is whats causing the problem... i ewas thinking about trying a different kind of sensor as well...
based on your setup..im going to change to 25khz and try that first.. b back in a second..

Correct, no C3 index card. It's wired directly to an input on the BOB and 5v. I would assume that your 35kHz kernel would make your sensing more robust at higher rpm rather than less. On the index debounce, I started with 1 then nudged up the value until I got a reliable and stable reading through the full rpm range.

Sent from my SAMSUNG-SM-G870A using Tapatalk

[1hander]

Correct, no C3 index card. It's wired directly to an input on the BOB and 5v. I would assume that your 35kHz kernel would make your sensing more robust at higher rpm rather than less. On the index debounce, I started with 1 then nudged up the value until I got a reliable and stable reading through the full rpm range.

Sent from my SAMSUNG-SM-G870A using Tapatalk

Thanks for the clarification Mike.. I will probly try your way as soon as I can figure out why the 10vanalog output is no longer putting out voltage.. I'm super passed off right now

[CL_MotoTech]

Try a larger slot. You aren't getting a clear sharp edged signal. I've had this issue with wheel speed sensors before.

[1hander]

ok, im going to try a wider slot.. i think your right because the more i widen the slot the higher i get a steady rpm readiing... its looking like its going to be half the diameter of the disc though...
i would really like to get the index working...and have a shot at rigid tapping...

if the rpm is reading correctly..my spindle is 400 rpm slower than actual... 4250 is half of the setup rpm..and the analog voltage is not showing 5v at half like it should but something around 4.78volts..


i had ordered a new c62 card for the 4 axis router im building...and i decided to hook it up in place of the or the vurrent c62 thats not putting out an analog signal... and voila i have a working
analog signal and spindle again... so for some reason my old c62 cards analog output has taken a crap..i didnt change any of the settings on mach3 and the jumpers are all setup the same way
the old one was....

frustrating and expensive to say the least..im in contact with art and johnny and really hoping they can repair it or replace it.. fingres crossed..

[DonKes]

Can you reverse the high/low logic? The recovery might be better in one direction than the other.

Don

[1hander]

i thought the same and tried both ways ...the index card itself may be part of the problem...maybe its just one of those ...IT DOESNT NEED TO BE THAT COMPLICATED DEALS... another member posted his setup and hes not even using the index card..just a sensor with 5v to power it and the output going straight to an input and hes not using a slot as wide as mine is currently..before i cut the wheel more im going to try his setup..

i have not however tried reducing to 25khz from my current 35khz to see if that does anything..

thanks for chiming in guys,, i know this stuff has been beat to death and it gets old answering the same ole questions...i do appreciate it and for that reason..i will always help someone else anytime i can ....even when its the same old stuff..
rick

[mikegiraud]

i thought the same and tried both ways ...the index card itself may be part of the problem...maybe its just one of those ...IT DOESNT NEED TO BE THAT COMPLICATED DEALS... another member posted his setup and hes not even using the index card..just a sensor with 5v to power it and the output going straight to an input and hes not using a slot as wide as mine is currently..before i cut the wheel more im going to try his setup..

i have not however tried reducing to 25khz from my current 35khz to see if that does anything..

thanks for chiming in guys,, i know this stuff has been beat to death and it gets old answering the same ole questions...i do appreciate it and for that reason..i will always help someone else anytime i can ....even when its the same old stuff..
rick

Note that the sensor that you're using will play a big part in the slot size. The Panasonic sensor I use has no issues resolving the single .150" diameter hole up to 6200rpm. If you're needing a huge slot, it may be time to look at a sensor with more capability. I spent under $20 for a 2 pack of the Panasonic sensors. My guess is a larger slot like .5" would allow you to go up to the 9k rpm that you need.

Sent from my SAMSUNG-SM-G870A using Tapatalk

[1hander]

i thought maybe the ensor may be the weak link too... so i ordered a couple ...are the wires easy to figure out... black-ground...... red 5v......white or blue signal

[1hander]

well after talking with art..he has decided that there is something wring with the analog port... and his advice was to add a spindle card...which is stupid because i paid
180.00 to not have to have seperate hardware...so i could have everything on one board.. so i guess hes not going to replace it..and said that fixing it is not cost effective..

if i had hooked it up wrong then im pretty sure the second c62 i purchased would have dies as well.. i have a spindle card that never worked and was supposed to
getting replaced but never showed up... ...and an index card that stopped working and replacement never showed up..
that i replaced parts on to get working again..

im not sure what to say or think.. the c62 i bought the other day was going into the router im building.. this sucks..i guess ill have to try and fix it... follow traces, one component at a time starting from the analog
ouput and work my way back looking for the voltage thats missing...

maybe ill do a write up on it...wish me luck guys..its just too much money to just give up on.. especially for a dis. vet.
rick