Another spindle rpm issue....

You guys were a big help getting my spindle running the right rpm for tapping, thank you for that. Now, I still have an odd issue, my rpm changes from what is called and it's completely random.

Machine is a 1992 Partner 4 Knee mill, with a Centurion 5 "A" version control.

For example, I'm running a job with 5 tool changes, one larger drill I run at 2000 rpm, everything else runs 3000rpm. I would say about once per part, one of the tools will run the full 5000rpm that the machine is capable of, it's never the same tool every time, it varies, sometimes it's the slow drill, sometimes it's one of the other tools that's supposed to be running 3000rpm. It appears completely random, I've counted tool changes between full speed errors, I've check time, like every ten minutes or something.... and nothing seems to fit a pattern.

I thought maybe it was a bad voltage signal off the Acroloop card, and I had a newer "spare" X axis card, so during the tapping issue I swapped out the card, hoping my errant speed issues would go away. No such luck. Now that I have found out I had pots inside the control console to adjust for minimum and maximum voltage on the keyboard circuit board, could my current speed issue be coming from something in there? I have checked voltages when it's in the incorrect speed, and the voltage going to the VFd is 10V even when it should be say 4V (for 2000rpm). Spindle speed over ride pot will decrease the speed for me, even though the screen reads in error (it always reads the commanded speed).

Given I'm machining mostly aluminum at the moment, it hasn't been life or death if the machine runs 3000 or 5000, but when I have a large drill that should only be running 1500 or so, or tapping.... a run away speed is going to be disastrous. Any suggestions would be greatly appreciated.

Hi Brian, The spindle analog is being generated by the front panel K/B card. so I would have to assume that 1. it is having a problem, 2. the spindle over ride switch is flaky and is changing the over rides for some other value. What happens if you try adjusting the spindle over ride does the analog change?
PM me

Sporty bob

Overload switch does what it's supposed to do..... if I get a 5000rpm I can turn it down to 60% and get back to my 3000 rpm I originally called in the program.

As a side note, it has also gone slower than commanded just recently, since I did adjust the voltage pots to get accurate rpm's for tapping. That was the first time I have ever heard a tool go too slow.... now I honestly did think to do it, but I don't know how much the "slow" speed was in relation to what was commanded. I'll get my digital tach back out and next time it runs slow, see what speed it actually is running. That might give an indication of what the issue is.

I do have a complete front control panel from another machine, bought that and the acroloop cards from a guy that gutted and went to Mach 3, but that keyboard card does not have the pots like mine does, I think you said there were 3 or more variations. I don't know what would be involved and if I did change it out, would my speed adjustment go back to the Acroloop card? And would I need software changes to support the different keyboard card?

Hi Brian, If you nent to the newer K/B card in the "other" front panel you would need to run a new "front panel cable assembly. The reason being is that the connector going to the bottom of the K/B card are different, This connector brings the video and K/B commands back to the control. Look at the P2 connector on your current K/B card and conpare to the other front panel you have you will see what I mean.
Using the new K/B card does not entail any software changes just a parameter change, MISC parameters, " "force FP spindle" set to NO. This will then use the "control" spindle section of the "X axis control / acroloop card. still out puting the spindle analog thru Misc connector pins 13 and 14. ( alalog and analog ground)

Sportybob

Sportybob,

OK, having had a wonderful crash yesterday because a 5/8" drill went 3000 rpm instead of 1500 rpm into the part, I need to figure out this errant speed control situation. You mention the front panel cable assembly.... what do I need there. I did retrofit my machine with a flat panel monitor so we pulled a video cable thru the machine. Can I make some sort of adapter to go from the ribbon cable connector to the red connector? Here are some pictures of what I have, the old panel first (on edit, the pictures switched positions, so captions are opposite):

Attachment 231048

and then the newer panel, I'll have to track down flat panel for it too... seems to be a larger opening than the last one

Attachment 231050

Hi Brian, The Milltronics P/N on the cable is 9512-1 ( 10ft) other -X numberts for other lenghts. 9512-1 $232. from Miltro. I have never been able to cross a commercial P/N for the "red / brown" connectors. there is no Mfg or P/N on them. So making an adaptor is difficult.

But it may not be necessarry. I believe you have a -B control. if so, then just change the MISC parameter Force FP spindle to no. this will use the spindle control section from the X axis control/ acroloop card to provide the analog voltage.

Only if you canged the boards would the new cable or an adaptor ( if you could find the connector) be needed. If any one knows what or where I can get these 10 pin connectors. I am ALL ears.

As always PM me if you want.

Sportybob

I have an "A" Revision control. I did try changing the Force FP parameter and then called up M3S1000 and nothing happens, so that's not going to work with just switching the one parameter. Will see what I can do about a cable.... What/where does this cable hook to on the computer side?

HI Brian, Ok I just read back thru the thread. It seems you have a -A control. so, you would have to swap out the X axis acroloop card for a -B version with the acro and control on one board. then there would need to be some wiring changes to the interface panel.

All that being said. intermittant problems can be hard to track down. Looking at the prints I have on the board, I would suspect the U18 chip. It is a LM324 op amp chip, it is readily available. you could try using a "hair drier on it and see if you can get it to "flake" out, conversley try cold air in a can and see what that does. If a chip is flaky it will usually respond to heat or cold, mostly heat, cold is good for finding bad solder joints.
If it goes flaky then change out the U18 chip. if that does nothhing try the U17 chip ( MC1508 ) this chip is HARD to find. but the U18 chip is mostly what controls the analog. Adjusting the R14 ( offset) or R15 ( gain) directly drives the U18 op amp. OUT U18 #14 thru R10 ( 1k) right out to pin 13 on the round J2 plug.
The U18 is soldered directly on the board so you will need some one who is experienced at soldering chips on boards. replacing thee chip you could add a "dip" "carriage solder that to the board and just plug the chip into that. or it can be soldered directly.

Any way there is something to look at.

Sportybob

Hi Brian, busy researching and posting a reply while you were replying. the 9512-x cable connects to the DB 25 connector below the POWEr connector on the card cage. The same place as the exsiting cable.

I will do some looking at the cables in the morning. Also see if I can find the U18 and U17 chips on the keyboard board. I do currently have the X axis Acroloop card from a B revision machine in my machine... I bought the complete set of acroloop boards and the front control panel off of another machine for spare parts. Is there a specific part number on the Acroloop card that I should have? Nothing states that just because it came out of a B revision machine that it is the right card.

As some further info.... the speed changes happen any time, hot, cold, all day long, usually at least once per part (running 5 tools, about a 7 minute long job). I have basically two speeds called for, most of the tools are 3000 rpm, one larger drill is 1500 rpm. What I get is most of the time the speed doubles... The 1500 rpm drill runs 3000, the 3000rpm tools run full out (5000 rpm on this machine). Now, once in a while, maybe once or twice a day, the speeds are half of what is called, so my 1500 is 750, the 3000 is 1500, but this is a lot less common than the over speed condition.

Hi Brian, The U18 chip is just below the round 14 pin connector opposit the side with the R15 adjustment pot. The U17 is located about dead cent on the board.

For the spindle speeds to change, the analog voltage needs to change.
One of a couple things have to happen due to;
1. . the programed spindle speed changes and it varies the analog voltage.
2. the spindle over ride switch changes and it varies the analog voltage.
3. the alalog voltage remains constant but the spindle drive changes the the spindle speed.

check the spindle drive display, it should read a frequency ( Hz) when it is running, check the analog voltage at the drive.( base line tests with your tach)
when the spindle speed changes:
1. check the spindle speed displayed % on the front panel , does it change? say from 100% to 50%. If yes then,This could be a override switch problem.
2. recheck the frequency and the analog voltage. Did either change? which one or both? Keep the rear door open on the machine so you have access to the drive with out shutting it down. If the analog increases / decreases and the drive freq goes up / down then the drive is doing what it is told. if the voltage remains constand with programed speed and just the freq changes then it would indicate a problem with the drive. If the analog is changeing up and or down Then we are back to the analog voltage being generated by the K/B encoder card.

The R15 gain pot is 100K ohm you could check the resistance across the pot ( base line) and then recheck when the spindle speed changes, This is harder to do because you need to have access to the backside of the card. You could try tapping on the side of the pot with yor fingernail and see if that has an effect. If the pot is changing values, replace the pot, if not then we are back to U18 as the most likly source of the problem.

One easy "try and see" would be to reformat the SRAM and reload the parameters, Back up you programs if needed. This is easy and takes 2-3 mins. as long as you have a back up parameter disk.


As mentioned earlier, to use the "new" control/acro card the control would need to be modified to a "-B" This requires an internal wireing changes to the interface FPP connector and a new internal cable(s) as well as a new external 9512-x cable. PM me if this is the route you want to take.
PM me for my #
Sportybob

1. The spindle speed on the display is always what is commanded, if I have called 1500 it reads 1500, whether the spindle is really running 750, 1500, or 3000.
2. Frequency and analog voltage jive exactly with whatever rpm the machine is running, regardless of what is/was commanded, if I'm running 1500 I have 3V, 3000 I have 6V, the VFD is behaving correctly.

I'm hoping this is a result of bad chips at U18 and/or U17 and I can replace chips rather than change out all sorts of stuff.

Hi Brian, if the analog is always correct and the spindle speed override is is always correct and the freq is alwasy correct. Then either the input voltage to the drive is changing ( L1, L2,L3) or the output voltage from the drive to the motor is changing ( T1, T2,T3) That is the only thing left that can effect the speed of the motor speed.
monitor the input and output voltages.

Sportybob

The analog is not correct as to what rpm I have called, but the VFD reacts appropriately to the analog signal. I guess I didn't state that very clearly. The issue seems to be the analog voltage the U18 chip is sending is not consistent.

Bob,

A quick question, is the chip really MC1508? I opened up my machine and found it to say MC1408P8, so is one superseded by the other? I found the 1508 chips on ebay and was already going to buy a couple.... figured I better double check.

Brian, I just found this drawing, I thought it might help.

Thank you ZZZZ,

I have the chips on order, along with sockets to mount them. I'll have to wait to have Marty come over and help me with the replacement, beyond my electronic expertise, or lack there of. Hopefully this fixes the speed problems. The machine has issues on almost every part (average of 3 to 7 different tools) and it's making it a bit stressful if/when a tool can't run double or half the speed and work.

Hi Brian, I did some checking on the 1508 Vs 1408 chips, They are operationaly the same chip except that the 1508 has a broader temp range, -55 - 125 C with a ceramic case. so the 1408 chip is just fine. 0-75C

Sounds good, I have both the 1408 and 1508 ordered, and we will solder in sockets so we can change out the chip at will if necessary.

OK, I received the chips and replaced the LM324 chip (with my friends help) and the results on the machine are the same. I wrote a program with option stops to run 650, 1000, 2000, 3000 and 4000 rpm and would cycle it over and over, sometimes the speeds would be high, sometimes low... 90% of the time the speed in error was the last speed called, like the signal from the computer didn't change.

So, I got to thinking, I know, that's dangerous, but back last year I upgraded the software on the machine because I was having following error problems in the minus direction of all the axis's. That turned out to be a bad -15V voltage regulator on my power supply, so while the cards were getting +15V, they were getting -24V, so it was impossible to balance things. In the process of that, before we figured out it was the voltage, I changed the chips on the Acroloop cards and Keyboard card, along with the Centurion software.

Here is what I had originally:

Centurion V CNC 1.29.09
Front Panel Present 111
Controller Card V0206 0000
X V0208 0000 0100 0000
Y V0208 0000 0200 0000
Z V0208 0000 0000 0000

What it is now:

Centurion V CNC 5.94
Front Panel 117
Controller Card V0240 0000
X V0254 0000 0100 0000
Y V0254 0000 0200 0000
Z V0254 0000 0000 0000

I'm wondering if I have a conflict between the software and the machines capabilities.... this is a 1992 and the control is revision A. I seem to recall the speed issues showing up after the chip and software change, but honestly I was so concentrated on the axis error issues I can't remember. Also, once the speed issues showed up, I never had "slower" than commanded until I tried an X Axis Acroloop card out of a B revision machine I got spare parts from. After I swapped that card into the car cage I got instances of the spindle running slower than commanded.

So, is this a possible issue? A software glitch that the upgrade doesn't jive with the older computer components? I can change the cards and chips back to the original state... but I can't find the emails or remember how I upgraded to the 5.94 software and what the process for going back would be.

Hi Brian, I have the Sim card files off 5.169 if you would like to try them. I obtained them off a sim card I purchased of ebay. I installed the card on my partner 1 which had 5.149 on the hard drive and I had no backup, it ran fine after loading my parameter file and modifying the path. Then I copied to my hard drive and removed the sim card I also may have a msdos backup of 5.149 I would have to see if the backup still works on a old 486 I keep around to play with.

Ron

Hi Ron,

Sure, that would be great, never hurts to have a back up copy of another version. You can always send it to my email at blamb11 at cox dot net

I have copies of the two versions I noted, just wondering if the 5.94 is just too advanced for my 'hardware".

Hi Brian, I just upgraded my Cent V -b partner 1E last month from Milltronics and they sent me 5.404 are you sure its 5.94?

Ron

Brian, Sent zipped files to your email

Ron

Got it, thank you!

Hi Brian, I just realized that the revision numbering system of the software goes thru a number of minor updates IE 5.94 then 5.100 then 5.149 and so on 5.404 is the latest for cent 5. You might want to check with Sportybob or ZZZZ about what you can run on your control with a A serial number acroloop cards. My control is -B.

Ron

Hi Ron,

I was kind of hoping per my questions in post #20 that one of those two would let me know what software revision will or will not work with the A revision machine.

Brian,
Have you made more progress with this?

Well, progress maybe, but not completely fixed. I had my friend come over and we put my original version of the Milltronics software back in, changed out all the chips (on the keyboard card and acroloop cards) I had upgraded back when I upgraded software and while the machine seems better, it's still not correct 100% of the time.

I had a program where the machine steps up and down in rpm with an option stop so I can step it from rpm to rpm and have time to get my tach on the spindle. Previously I could rarely get thru the six or so rpm changes before I had to restart the program without an error. Now, it ran the program maybe 30 times, so maybe 180 speed changes and it was only off twice. I have since only run one program and it was a one tool, one rpm deal, and it ran about 100 pieces and only ran the incorrect speed twice.

So, while it's not gone, it is a lot better. I was in the process of mounting a tachometer kit I got off ebay, display with a proximity sensor and a magnet you attach to the machine. Made all the brackets, wired it all up, to find out Chinese quality control (or lack there of) strikes again.... the tach works to about 950rpm and then goes haywire above that, so not much good for checking the rpm. I stripped it off the machine, and I'll try playing around with it on the drill press or something, where I can test some things easier and see if I can get it to work.

Most of the time running fast isn't a huge issue for me with 99% aluminum parts I make, it's the occasional large drill or when I'm tapping and need a specific rpm that it becomes a problem. If the percentages of wrong rpms stays down like it has since the software change.... and I can get a tach to reliably work, I'll probably just live with it and get on with things. Pretty easy to stop the program and re-input the speed command if I have to to run a specific tool.

How are things progressing with your machine? Hopefully you aren't having speed issues also.....

Slight update, I did mount a tachometer to my machine. Cheap Chinese thing that the sending unit was bad on, purchased a new Cherry brand magnetic sensor from Newark and wired it all up today. A little tough snaking wires thru the innards of the machine, but got it done. Now I have a visual confirmation of the speed commanded, if it's wrong for a tap, I can stop the program before I bust the tap.

Attachment 237430