G320 losing position - Page 2


Page 2 of 3 FirstFirst 123 LastLast
Results 21 to 40 of 56

Thread: G320 losing position

  1. #21
    Registered
    Join Date
    May 2008
    Location
    USA
    Posts
    57
    Downloads
    0
    Uploads
    0

    Default

    I believe I have found the problem. The digital readout I have been using to measure movement is not functioning properly, at least for the X-axis.

    I rigged up a 6" digital caliper to do a quick movement accuracy check. All movements were accurate and repeatable to 0.0005". Next I set up gage blocks and a dial test indicator. Even at 250 ipm all movements were accurate and repeatable.

    The new differential encoders seem to be working fine. I intend to borrow an oscilloscope (and learn how to read it) and check the encoder signal for noise as well as to tune the servos.

    I wish to express my appreciation to all of you who helped with this problem, especially Mariss at Geckodrive.

    Mahalo,
    Randy
    Honolulu, Hawaii



  2. #22
    Registered
    Join Date
    Dec 2006
    Location
    USA
    Posts
    839
    Downloads
    0
    Uploads
    0

    Default

    The US digital "HEDS" incoders is what came with my servos. I guess I can expect trouble. I have already noticed a difference in the way each servo tunes and at idle there is a big difference. I planned to swap parts around and see if I could pinpoint any thing causing it. This is just on a bench test, although it really seems to be working perfect and passes all test.


    One servo will set died quit at idle, and never faults over a wide range of tuning settings. The others tune different and have the hum at idle. One has to have more gain and damp to run without fault. The other will run with out fault over a wide range of tunnning but it hums at idle. Still I can get all three axis to run the Road Runner G-code in 11.26 minutes without fault, I guess that will change when I mount up to the machine.


    Jess



  3. #23
    Registered
    Join Date
    Apr 2008
    Location
    USA
    Posts
    1
    Downloads
    0
    Uploads
    0

    Default EM1 / HEDS Modules

    Quote Originally Posted by Mariss Freimanis View Post
    It's like 40 millionths of an amp (as in 0.00004 A). If you order from them, MAKE SURE you order the 'EM1' optical head and not the 'HEDS' one. They skin you for an extra $3.15 but at least it provides a +/-8mA sink/source capability. It can drive a short cable (<6') while I wouldn't trust the HEDS spec to drive a 1" cable.

    It used to be their encoders were rock-solid. Now we have been dealing with stuff like "The G320 passed 'encoder test' but goes nuts when I connect the motor!".

    Begin rant:

    Something must have changed at USdigital. 40uA is not something in good conscience I would ever use to drive a cable! I would even think twice to even drive a 0.1" long pcb trace with that current. 8mA is barely acceptable but not 40uA.

    Figure on 50pF per foot of cable; a 6' foot (2 meter) cable has 300pF of capacitance. How long does it take to charge 300pF to 5V at 40uA? 37.5uS is how long and it's also the rise time. It also means if CH_A goes from 5V to 0V it injects a 38uS 'notch' in the CH_B wire if it is misfortunate enough to be at 5V at the time (50% chance). That kind of 'notch' guarantees malfunction of the encoder signals. Malfunction on the encoder lines means malfunction of the drive.

    They offer "Line Drivers" at $15 and change. The 'line driver' is a $0.12 IC that should have been included inside of their encoder. They ship a cable with their encoder, the expectation is the encoder can drive that cable so why isn't the miserable $0.12 driver IC inside the encoder where it belongs? Why are they charging more than 100 times what it costs when it should have been for free inside the encoder?

    I can no longer recommend USdigital encoders until they get their act together, stop the cost-cutting and include a proper driver for their cables they supply.

    If you want a really superior encoder for just $29, go to www.digikey.com and tap in 'AMT102-V'. It has a serviceable line driver, has nearly perfect 90-degrees quadrature and get this, it is 16 encoders in one. It has a DIP switch that sets the encoder line-count resolution from 48-lines to 2,048-lines with 14 settings in between. It's a capacitive-type encoder so there is no dirt sensitive optical code-wheel. Finally, the pinout and cable connector exactly matches what's used on the USdigital encoders.

    Why the rant? It is because USdigital is costing us time and money dealing with application support calls that devolve into the short-comings of their encoder cable drivers. These calls wouldn't come up if their encoders could actually drive the supplied cables.

    End of rant.

    Mariss
    Mariss,

    The HEDS module is the original module that was developed by HP and is currently built by AVGO (they bought that division of HP). The specs for sink and source current has not changed since the part was released by Hewlett Packard more than 15 years ago.

    That module is one of the most widely used encoder modules in the world with a yearly production in the millions and is used by many OEM motor manufactures.

    The 1 K pull up on the resistor shown on the input of the G320 and G340 drives provides enough current (5 mA) to pull the input high so the modules source current is not an issue. This of course is a very common setup.

    As you know one reason that single ended cables are not recommended for long runs is for noise immunity. For certain applications it is possible to use a much longer cable, however to be safe we recommend switching over to a line driver past 6 feet.

    You are correct that the EM1 Module will sink and source more current. The ASIC used in the EM1 is a custom designed part and it is tooled to provide resolutions that are not provided by HP / AVAGO. When the part was designed it was decided to upgrade some of the features. Also The EM1 module also has a built in .1 UF filter cap.

    While the CUI encoder is a reasonable choice my testing of the part last year showed that it had a lot more signal jitter and depending on the resolution settings a number of glitches in the signal. I also found that that the quadrature timing and the signal jitter from both the HEDS modules and the EM1 were superior. This really maters if you are using an encoder for closed loop control as well as for general accuracy.

    Also realize that the EM1 is a custom part that supplements the HP modules and is produced in much lower volumes. That said the price difference may not seem all that unreasonable given the capital expenses that were spent to tool the production line.

    If your ever in Vancouver WA let me know I would be glad to give you a tour of the production facilities including the clean room where the modules are manufactured.

    Kevin C
    US Digital

    Last edited by Kevin C; 03-02-2009 at 07:22 PM.


  4. #24
    Gold Member
    Join Date
    Mar 2003
    Location
    United States
    Posts
    2839
    Downloads
    0
    Uploads
    0

    Default

    Kevin,

    Thank you for your reasoned response. My criticism was directed entirely at the HEDS module which effectively is an open collector device. An open collector is not suitable for driving anything but the shortest of cables. A 3K resistor and 500pF mutual capacitance (10' long cable) gives a 1.5uS time constant 'notch' in the adjacent channel conductor when it is high.

    I assume the 100nF capacitor you mention is a power supply bypass capacitor. I would also assume then the HEDS is absent any bypass capacitors. The reason I ask is we have had a spate of support calls that resolve to encoder problems (replace the encoder with a different one, problem disappears). In the future I can recommend customers bypass the HEDS encoder supply with a local capacitor. In effect, the HEDS device is an "incomplete" component requiring supply bypass capacitors and a cable driver IC before it can properly drive a longer cables.

    Your EM1 module works nicely as a cable driver. Its symmetrical +/-8mA output capability drives the same cable (10', 500pF) with 300nS rise/fall times and shows virtually no adjacent channel crosstalk. We have had no support issues with that module.

    I apologize if I gave the impression my criticism applied to all your encoders. I should have made it absolutely clear it only applies to the HEDS module and only in applications where the cable run is long enough to cause significant capacitive loading. The lack of a local supply bypass capacitor is hard to understand though; its absence can create a lot of mischief.

    Mariss



  5. #25
    Gold Member BobWarfield's Avatar
    Join Date
    May 2005
    Location
    USA
    Posts
    2502
    Downloads
    0
    Uploads
    0

    Default

    This discussion between Gecko and US Digital still leaves those with HEDS encoders unclear on what they should do. Lots of servos out there being sold with that encoder.

    Are we saying these encoders are unreliable and should be replaced regardless of cable length? Must they be replaced with cables over 6 feet?

    What exactly is the recommendation?

    Seems like if the word doesn't get out, there will be lots of problems for all concerned.

    Best,

    BW



  6. #26
    Registered
    Join Date
    May 2008
    Location
    USA
    Posts
    57
    Downloads
    0
    Uploads
    0

    Default

    The U.S. Digital single ended encoders (HEDS technology) I purchased simply could not drive even my shortest cable which is less than 4-feet in length.

    One option for those with this type of encoder is to purchase the line driver, cable and receiver. See attachment.

    Randy

    Attached Thumbnails Attached Thumbnails G320 losing position-image001-jpg  


  7. #27
    Registered
    Join Date
    Dec 2006
    Location
    USA
    Posts
    839
    Downloads
    0
    Uploads
    0

    Default

    Quote Originally Posted by BobWarfield View Post
    This discussion between Gecko and US Digital still leaves those with HEDS encoders unclear on what they should do. Lots of servos out there being sold with that encoder.

    Are we saying these encoders are unreliable and should be replaced regardless of cable length? Must they be replaced with cables over 6 feet?

    What exactly is the recommendation?

    Seems like if the word doesn't get out, there will be lots of problems for all concerned.

    Best,

    BW
    I am with Bob on this, what are those of us that have the HEDS encoders to do. Even if we buy the line drivers does this not cause problems with using them with the Gecko drivers (line drivers turn the encoder into a differential encoder? Is this basically correct? ).


    I do not appear to be having any trouble with my HEDS 9100 encoders on my servos at this time. But I have only done bench testing. I just picked up a scope tonight so I will be able to do some better testing. But as I plan my mills build I am starting to see that when I build my inclosure it will cause me to need to run longer cables on atleast one axis (six foot cables at this time). I assume this will lead into trouble, if I am not already having trouble and just done realize it yet. For now they do seem to be working very good and I can say enough about how smooth the system seems to be working. When I add cable lenth, or get the mill running which will have more noise than I am dealing with now, I expect this to change (having problems).


    If I understand right when I go to US Digitial if I buy line drivers first it changes the type cable I need (new cables after already buying cables). Plus it seems they will be 10 pin cables and this will not work with the Gecko . Or is there some type of inteface board that will be needed?


    So whats the anwser here ? It would be nice for me to figure this out while I am building my electronics enclosure and designing my cable harnesses. One way or the other it seems I will be spending more, but I did not expect to need longer than 6' cables untill I got to looking deeper into my mill enclosure, so my bad. It looks like the easy awnser is line drivers but I am not sure that will work with the Geckos.


    Its going to be interesting when I fire my system back up now that I have a scope. That should be in a day or two.


    Thanks Jess



  8. #28
    Registered
    Join Date
    Mar 2007
    Location
    USA
    Posts
    137
    Downloads
    0
    Uploads
    0

    Default

    Hi, I have some experience with the Gecko 320 and 340 drives, and many hours pulling out hair and cussing, due to tuning and positioning problems. With larger motors, one day, they are smooth as silk, the next day, the motors have the jerks. It seems that unless you are using a 4 amp continuous or smaller motor, you lose position. The SMALL motors seem to run great with Geckos, and tune easily. I had this problem with most nema34 and nema42 motors. And yes, I did use the proper encoders, shielded wire, power supply, proper heat sink, etc, etc, etc. The 5 volt fault circuit is a little cheesy also, sometimes it’s on when you power up, all by itself. I would guess you have nema34 (1/2” drive shaft) or larger motors. The solution: sell the Geckos, or use them with small motors, and buy some Larken or Rutex drives. Sorry dude, but I wasted a lot of time figuring this out. I’ll admit, I didn’t use an oscilloscope to tune the drives, like what the Gecko manual calls for, but even if you did, with those plastic pots, how long will it stay in tune? How much will it cost, hiring someone with an oscilloscope to help you? I NEVER had positioning problems with Larken or Rutex, unless the motors are WAY,WAY out of tune, like osculating. If I were you, I would cut my losses, and buy Larken Viper drives. And NO, I don’t work for Larken, I was a Centroid dealer in the 90’s. I have used the Gecko stepper motor drives, and NEVER had a problem with those. I think Gecko should install a heat sink on their stepper drives, stick to selling those, and quit selling the 320 and 340 drives, or simply say, the 340 drive is for a 4 amp continuous motor, or less! The only nema42 motor I ever seen a Gecko 340 drive work good with was those green, Amtek nema42 frame motors, but they are only 4 amps continuous! Now that I have Gecko mad at me for saying this, its ok, cause I use Larken drives! The ONLY bad thing I can say about Larken drives, is the manual isn’t wrote very well, and assumes you’re an electrical engineer & computer geek, not a toolmaker, but that seems to be the “MO” for most all these step and direction items.



  9. #29
    Registered
    Join Date
    Dec 2006
    Location
    USA
    Posts
    839
    Downloads
    0
    Uploads
    0

    Default

    Well thanks for the info, I must admit this sounds very much exactly like the problems I was having. I would get it tuned and the very next day it would go faulting. I do feel like I got it worked out though but again I worked it out one night with something I seen as a problem and I corrected what appeared to be the problem. Although I did not run it the next day because I am changing my enclosure box so I took it all apart.



    As far as scope tuning I picked up a scope last night so I will be able to go that route as soon as I am through with the new install. I believe my motors are around 8.5 amps continues with 38 amps peak. I hope your problem does not turn out to be my problem.


    As far as the line driver thing I believe I figured it out at US digital. The PC3 series driver allows voltage from 9.5 to 32v from a supply to be used, then it converts it back to 5v signal for the Gecko and you use the same 5 pin cabling so it will hook up to the Gecko. This hits me for $20.00 a encoder though but that seems to be the cheapest way out right now as for getting good stable encoder signal. The problem is I need five of them for 4 axis & the spindle. This kinda hurts when your dissabled and thought you had it all worked out already, all though part of it is my own fault becuase I didnt figure for needing longer cabling from the get go when designing my rig.


    As far as Geckos not handling high amp motors I dont know and I will be doing very close testing with my new scope over a few days in a row to see if this pops back up on my system (faulting the next day after tuning). It may be possible that using a scope to get tuned will get this worked out if it is causing me problems, but I want know until I do the testing. My system is not even running on a mill yet so that worries me about finding any problems until I do get it mounted and and cutting on a normal basis.


    I will also have to say that I dont like the slide on screw wiring mount system it has the Gecko has. The first thing that went through my mine when I seen this was high power motors, high voltage, and unstable connections because of the type connectors they have. I do not know that they will cause problems but if something like this does happen ( loosing tune or faulting after its worked out) it very well could be unstable voltage across these connectors. Soldered in connectors would seem a better setup. But I cant say there causing any problems yet either and if they do they are replacable. IT just seems a lot to ask for a connector like this to carry 72v at 8.5 amps (38amps peak though).


    I hope this does not go down this road (Geckos not handling high amp motors) because I really dont have the funds to pull out of such a thing for a very long time. I am strugling enough to complete the system as is. I guess I will find out in time here.

    As far as Gecko getting mad at you I have found the people at Gecko to be very reasonable to deal with and about as helpfull as you could get any company to be. I guess nobody is perfect all the time but it would surprize me if there drive could not run high amp motors no matter what (within there ratings). I have tried to build my system with as matched a set of componets as possible (within my budget). If I have to back up and redesign now it will put my mill in a box for a year atleast. I guess my new scope will be getting a workout here soon (well new to me even though it was a freebie from a company that no longer repaired circuit boards for there work, but just replaces them now).

    I am not sure its a good scope for this type work because I have never used one like this before. I have used scopes in automotive enviroment on ignition and in my last days of still working I was using a hand held scope for working on the on board computors and sensers on fuel injection, but this type of scope use may require something different than I am use to. The scope should be enough to get the job done though because it is still just testing sensors and such. I hope to find out tonight.

    Anyhow I will go forward with it and see where it goes. I think I willl get the line drivers for these encoders just to be sure I can handle what ever cable setup I will have to run. All though I dont really thin I am having encoder signal problems yet. Tonight will be interesting when I get to put the scope in the mix and I can see whats really going on. There is nothng like proving what you supect.



    Jess



    Jess



  10. #30
    Registered
    Join Date
    Mar 2007
    Location
    USA
    Posts
    137
    Downloads
    0
    Uploads
    0

    Default

    Dude, when you say “line drivers” for encoders, I think you mean the conversions to make that little us digital encoder go from 5v. 0v, A & B to having the A,A- B, B-. Right? And then you can have a cable longer than 6’? right? I tried that, it don’t work! sorry dude, your motors are simply to big.



  11. #31
    Gold Member
    Join Date
    Mar 2003
    Location
    United States
    Posts
    2839
    Downloads
    0
    Uploads
    0

    Default

    Jess,

    Something that works one day, doesn't work the next day and then works again a day later is going to be a connection issue. Drives can go bad but once bad, drives cannot cure themselves and work again. That kind of symptom is a bad wire connection somewhere, most likely on a cable that flexes as the machine moves.

    I'm sorry but the G320 being "only good for low currents" stuff belongs in the urban legend category. Every G320 gets tested on a 13.8A, 90VDC NEMA-42 servomotor rated at 1,000 Watts mechanical output. A G320 doesn't ship if it can't deliver an actual 1.33 HP to a load.

    I was called today by a senior technical guy from USdigital with whom I had a nice long conversation. From that conversation and from our own lab tests I can endorse without reservation their EM1 module encoders. We have tested them and they work like they should; they can drive cables without transmission errors.

    The same cannot be said about the HEDS head encoders. After our conversation, USdigital will be checking to see if Agilent (mfgs of the HEDS head) has had a recent process change in the last 6 months. As specified now, they are not capable of driving a cable unless a critical bypass capacitor is added at the encoder end.

    While conducting some unrelated tests a few months ago I noticed a quenched oscillation on the CH_A and B outputs. I took a scope picture, filed it away and shrugged it off as just a bad encoder. That test takes on a new significance now that I've learned there is no 5V bypass capacitor in the HEDS module. I will replicate the test this coming weekend on recent and old HEDS modules we have around here.

    Technical stuff: An encoder uses phototransistors or photodiodes as optical detectors. The signals from these are detectors are analog in nature. The encoder uses comparators to 'square-up' these analog inputs into digital quadrature signals presented at the output pins.

    Comparators are very high gain amplifiers. That high gain amplifies and 'clips' an analog input signal until it looks like a digital output. Oftentimes hysteresis (positive feedback) is used to insure the output doesn't have a linear region. Very high gain also means having some response (gain) on what happens on the 5V power supply voltage. When a digital output has a state-change (0V to 5V or 5V to 0V), there is a change in the supply current. If the encoder is at the end of a cable, that supply current change causes the cable inductance and resistance to give a supply voltage 'spike' or 'notch' dependent of the direction of current change. The comparator reacts to these voltage changes by supplying additional, spurious output state-changes.

    A supply bypass capacitor acts as a very local reservoir or flywheel on the supply voltage. It reacts instantly to supply the needed current with very little voltage change just like a transient load on a flywheel doesn't change its RPM.

    Bypass capacitors are a must for any electronics circuit, the encoder head is such a circuit and it will malfunction if that capacitor isn't there.

    Mariss



  12. #32
    Registered
    Join Date
    May 2004
    Location
    United States
    Posts
    69
    Downloads
    0
    Uploads
    0

    Default

    I have had the same problem as described above and installed the transmitter and Reciever units on the encoders and at the gecko unit and this has solved my lost step problems. My encoder wire is approximately 10' and would get lots of missed steps, Once I installed the reciever and transmitter the problem went away and have been regaining my hair that I pulled out trying to figure this problem out.



  13. #33
    Registered
    Join Date
    Dec 2006
    Location
    USA
    Posts
    839
    Downloads
    0
    Uploads
    0

    Default

    Quote Originally Posted by Mariss Freimanis View Post
    Jess,

    Something that works one day, doesn't work the next day and then works again a day later is going to be a connection issue. Drives can go bad but once bad, drives cannot cure themselves and work again. That kind of symptom is a bad wire connection somewhere, most likely on a cable that flexes as the machine moves.

    I'm sorry but the G320 being "only good for low currents" stuff belongs in the urban legend category. Every G320 gets tested on a 13.8A, 90VDC NEMA-42 servomotor rated at 1,000 Watts mechanical output. A G320 doesn't ship if it can't deliver an actual 1.33 HP to a load.

    I was called today by a senior technical guy from USdigital with whom I had a nice long conversation. From that conversation and from our own lab tests I can endorse without reservation their EM1 module encoders. We have tested them and they work like they should; they can drive cables without transmission errors.

    The same cannot be said about the HEDS head encoders. After our conversation, USdigital will be checking to see if Agilent (mfgs of the HEDS head) has had a recent process change in the last 6 months. As specified now, they are not capable of driving a cable unless a critical bypass capacitor is added at the encoder end.

    While conducting some unrelated tests a few months ago I noticed a quenched oscillation on the CH_A and B outputs. I took a scope picture, filed it away and shrugged it off as just a bad encoder. That test takes on a new significance now that I've learned there is no 5V bypass capacitor in the HEDS module. I will replicate the test this coming weekend on recent and old HEDS modules we have around here.

    Technical stuff: An encoder uses phototransistors or photodiodes as optical detectors. The signals from these are detectors are analog in nature. The encoder uses comparators to 'square-up' these analog inputs into digital quadrature signals presented at the output pins.

    Comparators are very high gain amplifiers. That high gain amplifies and 'clips' an analog input signal until it looks like a digital output. Oftentimes hysteresis (positive feedback) is used to insure the output doesn't have a linear region. Very high gain also means having some response (gain) on what happens on the 5V power supply voltage. When a digital output has a state-change (0V to 5V or 5V to 0V), there is a change in the supply current. If the encoder is at the end of a cable, that supply current change causes the cable inductance and resistance to give a supply voltage 'spike' or 'notch' dependent of the direction of current change. The comparator reacts to these voltage changes by supplying additional, spurious output state-changes.

    A supply bypass capacitor acts as a very local reservoir or flywheel on the supply voltage. It reacts instantly to supply the needed current with very little voltage change just like a transient load on a flywheel doesn't change its RPM.

    Bypass capacitors are a must for any electronics circuit, the encoder head is such a circuit and it will malfunction if that capacitor isn't there.

    Mariss
    Thanks for the reply Mariss.


    I am fairly new to alot of this electronic stuff, but I am fair at diagnossing problems even when there out of my field. My system is still
    on a bench and all connections have been gone over with great care. I have put a extreme amount of time into wiring this unit and testing each connection over several times. Still connection are always a possibility and no one is perfect and can miss something.


    I have broke the system down at least three times though and still I keep having the same repeated problems even after wiring has been changed. I just picked up a scope and I have been using it the last few days. Last night I tuned the drives and things where working quit good. TO night I fired the system up and I was going to do some more testing and sure enough it starts faulting. The strange thing is when it does this it will do it in the - direction always. My scope is older and I doute very much if it is even calibrated to any level of accuracy becuase it was a free unit and most of the time you get what you pay for. But one thing I noticed is when tuning that the drive signals do not mirror each other in the two dirrection. The signal is different and its in the - dirrection that it shows weaker. This is when hooked up to the test lugs in the drive and I am not sure if this is giving a reading of the encoder or some signal the drive is producing. This is my first time using a scope so I don't feel like I really know what I am doing, but according to the scope the drive tuned up just like the paper work showed. The signal is just much weeker when going in th enegitive dirrection when jogging. I am going to hook up to the encoder outputs on the next test and see what I can find becuase what ever it is it keeps showing back up and repeating.
    Even when the system set there and ran G-code for 5 hours without fault, the next day when fired back up it woudl fault on the first jog and stay that way until I backed things down to a lower level.


    With a new system there is a lot going on and without test its going to be hard to pinpoint. I am thinking of buying the line drivers for the HEDS encoders even though my wiring is only 6 foot. Even though its a extra cost I need the encoder to work right all the time no matter what and it might be best for me to just get this possible problem out of the way. The signal I get on the scope looks very clean when I am testing but when it faults the signal dies.


    If this bypass compacitor can be added I would be interested in putting one on if I knew exactly where to install it and how. Even thoguh I am new to this stuff I do understand diagrams and I have even built a few circuits from following someones papers.

    As I learn more about using this scope maybe I can pinpoint what is going on. Until then I can only follow others suggestions. For now I have plenty of time, although I do need to get this Mill going I am still waiting on a few more parts before mounting on the mill itself. So what ever testing or repairs I need to do it want kill me as far as getting any chips made. Any suggestions on what tests could be run would be very helpfull at this point. The signal right out of the encoder is what I had planned to look at next. I have toook the system down to barely enough to run with only 6 inchs of power wiring and two foot of wiring to the servo. The Breakout board is the only other thing hooked up, and of course the 6 foot encoder cables. I will try shorter cables on them as soon as I pick up some more shielded cable so I don't have to cut them.



    Thanks Jess



  14. #34
    Registered
    Join Date
    Mar 2007
    Location
    USA
    Posts
    137
    Downloads
    0
    Uploads
    0

    Default

    Did you get it running to the point of being trustworthy?

    Has anyone ever bought Gecko 320 or 240 drives for the 29 in lb servo that is on Bridgeport’s, and got them running to satisfaction?

    I think this topic should stay open for anyone that’s considering Gecko drives for his or her Bridgeport motors. Sure, Everyone has money problems, and wants to buy the lowest cost drive available, which I believe is the Gecko. At 20 amps, and almost 1/2 the price of other drives, this seems to be the choice, BUT, if its going to be nothing but problems, I think the cnczone forum, which is here to help, should advise people of the potential problems, like positioning, which can ruin an expensive part, not to mention the mill itself.

    I need to have a couple of people, that are actually running the Gecko drives with these fairly large motors, without any problems, to let me know, to keep me quiet. In my eyes, if you have these large servos and Geckos, you’re heading for failure before you even get started. I hope I’m wrong.

    Jess, its been a week, and no more responses in this post, and I think this is a fairly important topic, that should not just go away.

    I'll admit, I got the Geckos running just fine with small servomotors, but not these large ones.

    You have Mariss helping you, and an oscilloscope, which is more than most people have going for them. Please keep all of us informed as to your outcome, if you got them running to satisfaction.



  15. #35
    Member
    Join Date
    Dec 2004
    Location
    usa
    Posts
    1865
    Downloads
    0
    Uploads
    0

    Default Can't we have perfect encdoders for $.99/dozen

    Well here is another vote to throw out the heds encoders and try something else.

    These are ebay purchased taken off of brand new motors.
    I am using them on my bridgeport btc-1 with copley analog servo amps and dspmc/ip controler.
    Since the dspmc only takes differential signals, I built some single ended to differential boards with caps right on the boards. Without these caps, the encoders would fault out way before they should have most likely due to the problem that Mariss describes.
    After the caps, they worked a lot better, but still gave me problems.

    I also had what I will describe as almost random problems with the signals. Some days good, and some days bad. Some times an axis would glitch once early on in the day and then be fine. Especially on the days that the heat was off all night and then I tried to run the machine cold before the room heated up. On the last day the machine was not safe to run at all.
    I then switched over to the renco encoders that are on ebay, using the same line driver on one and changing the line driver on the x.
    Notice I didn't mention the z. The z has been a little less than perfect, but by limiting the rapids to 100ipm, it was very usable. I then starting more problems this week, and thought now what is the problem. I opened up the motor cover, thinking that it was the renco going bad or perhaps the driver board.
    I forgot that I hadn't changed out the last heds. I cut the wires from the heds connector, and soldered in the renco. It works better, and I was able to bump the speed back up to 150 ipm and it tested up to 290ipm, but I am afraid to break it or wear it out at that speed.

    So I now have renco's on all three axis and it has been smooth as silk the last 3 or 4 days, except for my programing skills.

    I did have finger print problems early on with the heds, but the rencos are enclosed units and are working well.
    Time will tell if I have cured the last machine problem.

    Warning: DIY CNC may cause extreme hair loss due to you pulling your hair out.


  16. #36
    Registered
    Join Date
    Dec 2006
    Location
    USA
    Posts
    839
    Downloads
    0
    Uploads
    0

    Default

    SpringLakeCNC, my testing has been limited, mainly becuase of the way I built my enclosure. I am now rebuilding it so that its easier do work on and run test with the system together.


    I did find some answers though. One thing I found what was the limting factor to my system. What I found was either the encoder count, or the RPM of my motors where reaching a point that was maxed out. The most I could get my system to run at was 120,000 encoder counts per second. I was messing around with different motor gearing (belt drive) I notcied no matter what ratio I used I was hitting the same point in encoder counts. Once I found this opoint I backed up to 110,000 counts per second and the system has been stable ever sense. Day to day, heat to cold, no matter what it runs stable now. I am set up with a 3 to 1 ratio and I can get 220ipm. Now this 110,000 encoder counts per second also correspondes to 3300rpm out of my motors. The motors are rated at 4200 and this may very well me the max I wil be able to get out of these motors. TotallyRC brought this up in a different thread I had about encoder count max of the 320 and he might be right about the RPM being the limiting factor instead of it being the encoder count. The only thing I know to do is run more test. My system did find stability though, in fact it runs veyr good where I have it.


    Still I have never been one to just say, OK its working, and let things go. I will do more test and find whats limiting what. Until I finish rebuilding my enclosure I want be able to do any more test. My encoder signals are very good at this 110,000 count limit so I don't think it has anything to do with "my" HEDS encoders. They are simply working as good as one could expect. I do seem to be very lucky with noise in my system though, I simply have no noise in any of my signal lines. So IDK, if they where in a noisy setup they might not be working so good. BUt for now the HEDS encoders are working perfect.


    I do plan to try a higher RPM motor, a different encoder cpr (maybe a 250cpr, and a 1000cpr) , and a different drive here soon. I have a 6000rpm servo that I will mount my 500cpr encoder on. This should be a good test to see if its the motor RPM or encoder count that is maxing out.

    Now my system is still on a bench. One thing I found odd is I know of a person that has the exact same setup as mine, but on a heavier mill. Its odd that his limiting factor fell right within the exact same figures as what I found. I would think that my system being on the bench might reach a higher RPM on the motor than his system does loaded. IDK myself, my testing is slow because of my health, and well my budget. As long as I keep the system at or below this 110,000 counts per second it is rock stable. I have fired the system up dead cold and let it run code at 175ipm without fault. If I let it run 240ipm (or 120,000 encoder counts per second) it will run one day and then the next day it may fault. You can set there and tune it to perfection and run it for hours and it seem stable, then the next day it will change and not run it. But if I back it down to the 110,000/3300rpm point it runs, day in , day out.



    I don't know which motors you are running, if its the 850's from HS I would suggest to do the math and find where this 110,000 count point is and try it. The other system I meantioned is on a Bridgeport mill and he can stay stable upto 330ipm. Now his system has a different screw pitch and gearing than mine but it still comes down to the same point, either motor RPM or encoder count. BTW he uses these 850 servos on the knee even. It has a countering system (either weight or air)but it runs the knee fine with these motors and his system has been cutting chips for some time now. But it is strange that his limiting point is the exact same as mine, and his loaded on the mill with mine on the bench.


    I will have my enclosure box back together here today or the next. I don't have a vid recorder but I will do more test and when I try different hardware setups I will report back with results from the scope. I can post pics of the scope readings but not vids. A higher count encoder on my motors, or a higher RPM motor with my encoders would be a interesting test. But I can do better testing with what I have running now once my enclosure box is finished. I simply could not get to things very good the way it was. I want it to be easy to work on anyway for later if I need to do repairs, or testing to find problems that may come up when I am using the mill. Trying to do these test showed me pretty quick that the way I had the box setup was going to make it a pain to work on later.



    Jess



  17. #37
    Member
    Join Date
    Dec 2004
    Location
    usa
    Posts
    1865
    Downloads
    0
    Uploads
    0

    Default Motor rpm max

    Hi Lucky and springlake.
    I just wanted to update you on how things are going. I have been making parts for 6 hours between yesturday and today and it had been flawless, except fo the ~30 year old home switch that is iffy.

    One thing that you can't see with the gecko drives, at least easily, but can see with my dspmc is a graphic display of the programed move vs the actual move. Plus I can set the error to what ever I want.

    A large error of 1000 counts would let you do things to show how the motor was way out of position during a rapid. What I saw on the bench as opposed to on the machine was that a well oiled axis, not set to too high of an acceleration, would run about the same as bench testing. I could run the accel faster on the bench, but the motor runs out of steam at a given voltage. As a general rule of thumb a brushed dc motor makes max power at 50% of its no load speed. If it will run 4000 rpm on 100v, then at 2000rpm it make max power for a short period. At Max rpm the motor has almost no reserve power as any load will cause it to slow down.

    From my electric flying stuff we used to shoot for about 75% of max speed on the ground which kept the motor in a nice part of the power band and wouldn't burn up the brushes.for some reason If I have trouble with these Renco encoders, and for my lathe conversion, I am going to try the ones from digikey that Mariss recomended.

    Mike

    Warning: DIY CNC may cause extreme hair loss due to you pulling your hair out.


  18. #38
    Registered
    Join Date
    Nov 2003
    Location
    Texas
    Posts
    126
    Downloads
    0
    Uploads
    0

    Default

    Quote Originally Posted by Mariss Freimanis View Post
    ...It used to be their encoders were rock-solid. Now we have been dealing with stuff like "The G320 passed 'encoder test' but goes nuts when I connect the motor!".

    Begin rant:

    Something must have changed at USdigital. 40uA is not something in good conscience I would ever use to drive a cable! I would even think twice to even drive a 0.1" long pcb trace with that current. 8mA is barely acceptable but not 40uA.
    ...
    Mariss
    I have about 20 of the HEDS encoders, but they are all at least 5-7 years old. They have been working fine for the most part (except for a couple that have died in the past), with shielded encoder cable from 3' to 8' lengths. You mentioned that something must have changed at USD.. Does this mean the older HEDS work correctly? These are so old that they have the metal encoder disk (500 count) with physical slots cut in them (quite delicate!)



  19. #39
    Gold Member
    Join Date
    Mar 2003
    Location
    United States
    Posts
    2839
    Downloads
    0
    Uploads
    0

    Default

    Please see the attached gifs:

    These pictures were taken of a HEDS-9100 driving a 25' (8 meter) long shielded 4-conductor cable back to the G320. The shield was unterminated and the NEMA-34 motor was not grounded back to the G320. The supply voltage was 72VDC and the motor load was >10A.

    Picture 1 shows the HEDS signal while the motor is being powered by a DC supply and not the G320. This is an OK signal.

    Picture 2 shows the HEDS signal while the motor is powered by the G320 and under load. Another encoder is providing the G320 feedback. This is a not OK signal.

    Picture 3 shows the HEDS signal after a 1uF 25V ceramic capacitor is placed across the HEDS +5V and GND terminals at the encoder. This is an OK signal now.

    Put a ceramic 1uF capacitor across the HEDS power supply terminals at the encoder. It makes all the difference. I will take and post pictures of how to do this easily on a HEDS-9100 encoder when time allows. Use Digikey BC1151CT-ND capacitors. They cost $1.80 for a package of 10 pieces.

    Mariss

    Attached Thumbnails Attached Thumbnails G320 losing position-1k-pull-gif   G320 losing position-500mv-vcc-noise-gif   G320 losing position-470nf-vcc-bypass-gif  


  20. #40
    Registered
    Join Date
    Dec 2006
    Location
    USA
    Posts
    839
    Downloads
    0
    Uploads
    0

    Default

    Thats pretty good if the bypass compacitor is allowing this signal with the 25 foot cables. I have about 8ft of cable on my HEDS encoders and the signal is every bit as good as the first or last pics. I have never seen a signal get the way the center pics show out of these encoders, of course I am not running 25ft cables either.

    I am glad you posted the pics. The lower encoder signal has the little tick at the midway point of the top of the square wave. What is this about, I am guessing it is normal if you get the same thing but I was wandering why its this way. The pic of the encoder losing its signal is great, I wasn't a &#37;100 sure what I would be looking for when testing if they was giving trouble but that answers the question for sure there. From the info from my scope I am good but I do want to add the caps just to be sure they will stay stable. I did a quick test with some caps and I could not get any better signal but I was not having trouble either. This sounds like good insurance to add the caps.


    For soom reason I was thinking it wasn't a ceramic copacitor that was needed to be used. Caps are cheap is there a better cap that could be used or will these be just fine? I have plenty of caps, even some of the 104 candy coated looking caps they use to bypass the supply for IC's with. My local store has anything I would possibly need so if there might be something that could work or last better jsut list it and I will get them. I do plan to run some even longer cables just so I can run my wiring the best/safest way possible. Plus I just bought some cable to use when I add extra lenth and it is one guage heavier than the factory US Digital cables, I am thinking this will be harder to drive and the caps should help if so.


    Really after looking at the pics closer I might be getting better signals than this. I will post a pics of my scope with it reading the signal at 120,000 counts per second with 8ft cables.


    Thanks Jess

    Last edited by LUCKY13; 04-04-2009 at 06:38 AM. Reason: adding coment


Page 2 of 3 FirstFirst 123 LastLast

Tags for this Thread

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  


About CNCzone.com

    We are the largest and most active discussion forum for manufacturing industry. The site is 100% free to join and use, so join today!

Follow us on


Our Brands

G320 losing position

G320 losing position