[cjmerlincnc]

Hello Dale

Yes I am stupid and I live in an under developed country where we are all stupid. Please do not download my code as it is stupid, you better write your own.
Please stop attacking me. You want something done do it yourself. I'm starting to regret I even made this open source.

Mihai

I too have benefitted from your knowledge, I thank you Mihai for making the project open source. Please don't be put off posting projects just because one individual feels the need to bully people with self-assumed intellect.

[tandel]

I am working with brushed dc and its working with servo tuner but its not working with step and direction
Please suggest me what i am doing wrong I am using logic level converter 5v to 3.3v
Something like this

https://learn.sparkfun.com/tutorials/bi-directional-logic-level-converter-hookup-guide

[Al_The_Man]

I am also in designing a High Current MOSFET drivers based on IRF840 and IR2110. This wll be drived by Arduino. Can you share the picture of your setup and especially the motor you are talking about. How much torque you get from that motor?

I have built a PWM drive using Picmicro and based on this power section by Tahmid.
Tahmid's blog: Using the high-low side driver IR2110 - explanation and plenty of example circuits
Al.

[dwalsh62]

Hello Dale

Yes I am stupid and I live in an under developed country where we are all stupid. Please do not download my code as it is stupid, you better write your own.
Please stop attacking me. You want something done do it yourself. I'm starting to regret I even made this open source.

Mihai

You vanished, stop responding to e-mails so why would I let anyone interested in taking the project to the next level wait for a response from you when you have given an indication by your silence that it does not interests you any more?

It is nice to see you have returned however, my comment is not claiming that you are stupid, according to the STM engineers, your code is pretty much the way it is provided from the CubeMX software with some minor alteration and not something that someone knowledgeable in this area who writes this type of code would create it, I trust what the STM Engineers tell me, they haven't steered me wrong.

STM Engineers classified your work as beginner/student code, I have asked you questions and you were unable to answer them and even ignored some which to me indicates your knowledge in this area is of a beginner type.

I get it, for you this is a learning project and that's great but the push is to complete the project so an inexpensive option is available to the hobby industry.

Yes I can build just about anything machine related but my area of expertise is in spindle shafts, I do it for a living so I know how to make them right, I have good skills to figure out just about any machine related construction and this is good in helping the hobby industry but not cost effective because I work without the thought of costs involved, I have the luxury of expensive tools that most only dream of, I have a good understanding of electronics and have written some impressive computer code and other than a few Arduino projects where I ported and fixed up some ARM code to AVR for myself such as the Chateau Reflow Oven project with an Uno and 3.95in TFT, I have no real programming experience outside of computers and the thought alone overwhelms me that I can't get my head into it.

You can get bent out of shape all you want by thinking that I called you stupid and the fact is I did not, you may not an expert about this stuff but you know significantly more than I do.

I never said your code was bad, according to STM Engineers it's a good base project that has not yet been filled with a lot of cruft and nonsense and like everyone else I have benefited some knowledge from your code but this insight is insufficient for me to make it do what needs to do, minor code correction in casting and such I can figure out, building a cross compiler from scratch because none are available for my platform, not a problem, getting with the STM32 to do what it needs to do is not my area so I seek out those more competent.

Finally realizing (and their acknowledgment of it) that the eclipse CDT is crippled and broken and has been this way since it's inception explains why tool chains are being provided with their own embedded CDT to overcome the issues in the eclipse provided one, the ilignuarm (you use it) is an extremely good example of this and what annoys me most about it is that they know the issues exist but don't want to fix them because CDT is an optional module and not part of the main eclipse program but it's installed blindly for java and only updates the registry so optional is an outright lie.

At no time does eclipse generate a secondary application, it's broken and always has been, I've added console prints to the java code, made the toolchain using the eclipse document example and it never builds or even attempts to process the secondary application and they've know about this for more than 10 years, my solution after the frustration of dealing with the eclipse idiots who've written this stuff but lost some of it along the way due to poor resource management ensuring it can't be fixed was to pay a java programmer $1200.00 to rewrite the broken stuff in my eclipse so I could have it properly working for myself and my platform and let everyone else deal with bloated memory usage due to poorly written and broken code, my running eclipse is using 18 threads and 84mb of memory with your project loaded and in win7 1gb of mem and 37 threads with the same version of eclipse (unfixed because eclipse lost the build scripts for 3.7.2 with CDT 8.0.2/8.4.0 so I can't rebuild for windows).

After this exercise I found it easier to manually compile the sources to generate the files with a script calling the cross compiler tools and my comment about making a cross compiler to jfstockton is based on his environment which I have intimate knowledge and significant experience of.



The layout done by _ID_ even you will have to agree is far better than the one you or I have done so this is not a "who's got the bigger d1ck" contest, this is about making the best project and your work is a large part of it.

If over the last 75 days you have gained more knowledge and experience or you had it but were just too lazy then work with jfstockton to improve the code or at least help him to understand what you have done or were trying to achieve rather than have an attitude because STM Engineers think your knowledge is low.

You will also have to concur, an inexpensive solution is needed for the hobby industry, they are not interested in working on the project, they are interested in using the project because they currently have no inexpensive upgrade path and I think it's about time that one was made available and many others agree with this.

I've made a couple of boards starting with your design, made changes as I thought were needed (some where just wrong and you were less than helpful when I asked questions), then made changes based on STM Engineers, significant improvement in response and performance being able to achieve 92% over your 68%.

Far too many projects have disintegrated or are still incomplete, how difficult is it for this type of project to be completed?

Is it rocket science???

Is it a conspiracy to keep it from becoming an inexpensive viable product for the hobby industry???

I've given jfstockton some updated code, I don't understand the updates or why STM Engineers recommend certain things be added but he seems to understand the logic and reasoning and this is good enough for me.

In my eagle files I have credited you specifically because without your initial works this project would not be possible and I've given the files to jfstockton and posted previous renditions here so yuo can get any idea out of your head that I'm trying to steal your project because that's not the case or my intent.

It is my intention to have this project available to everyone on a tight budget and I would like it to be the best it can be, if this is all you want from it then I'm asking you to help us improve it, after all, it's your initial project and we are grateful for it, if it's beyond your skill level then say so so people are not frustrated by your silence but check your attitude at the door, there is no malicious intent and to infer any is nothing but a fantasy because it just isn't so.

-- Dale

[dwalsh62]

When all is said and done, I may ask _ID_ if he is willing to redo my layout, if he says no, maybe someone else who's really good at it can give it a shot but I think it would be beneficial for this project to be offered with the utmost quality and integrity, I hate settling for "it works" kinds of stuff and this forces me to seek out better options.

I want this to be that better option, I want the Mihai project to be the standard for others to attempt to rise to rather than be a sub-standard used for comparison and I think others feel the same way.

Overall, this is a good project base to start from and this has been confirmed from several sources, some of which I trust so can we get past the pissing contest and back to focus on the project please.

I would really like it if you continued to contribute Mihai but if you have no interest then perhaps someone like jfstockton can take over where you left off and you can help him get up to speed by answering his questions, privately if you have no public interest, it's your call, after all, it was your project to start with and credit goes to you for it.

[dwalsh62]

As some of you have seen, I've built some nice small and powerful servo motors for all kinds of applications including an 12,000RPM, 1.5KW output with 2500ppr encoder servo spindle motor in a small package.

While I have several motors on hand now, I've been building and testing with the help of International Motors so I have several candidates that would make for a suitable match for this driver.

I've tried 350W with Mihai's driver but the peak current of 34A kills the driver so we should consider something with a peak current in the 10A to 12A range

One I'm very happy with is a Nema23 size employing a Tamagawa type A/B/Z optical encoder and from talking with my preferred motor production house, the price for this 150W servo motor is around $55.00 with a 1000ppr quadrature encoder in quantities of 100pc (minimum order) and $45.00 in quantities of 500pcs.

My testing has shown 150W exceeds the power and RPM capabilities of all Nema23 stepper motor I tested so what do you guys think of this as a suitable match for this 36V driver?

For the cause I would be willing to send one of my prototype motors to the production house (I trust them and they would need a sample along with the technical data) but moving forward without funding would be foolish so I'm open to discussing possible options or do you think that everyone should be left to create/find their own motor solution?

[tandel]

This my current situation I am using mach3 and as you can see in video that first I am giving it one rotation and afer that ten rotation and last I am giving one thousand rotation and as you can see servo is losing position ! I dont understand what is happening with it I am using 5 volt to 3.3 voly logic level shifter not optocopler

https://learn.sparkfun.com/tutorials/bi-directional-logic-level-converter-hookup-guide

and using smps for motor power and avago hedl encoder and inkjet printers motor. please somebody please solve this puzzle

This is my video showing problem :

https://m.youtube.com/watch?v=ymQ-GOiosBE

Please!!!!

[dwalsh62]

This my current situation I am using mach3 and as you can see in video that first I am giving it one rotation and afer that ten rotation and last I am giving one thousand rotation and as you can see servo is losing position ! I dont understand what is happening with it I am using 5 volt to 3.3 voly logic level shifter not optocopler

https://learn.sparkfun.com/tutorials...r-hookup-guide

and using smps for motor power and avago hedl encoder and inkjet printers motor. please somebody please solve this puzzle

This is my video showing problem :

https://m.youtube.com/watch?v=ymQ-GOiosBE

Please!!!!

Too much informaiton is missing to determine what your problem is.

I find it doubtful that the driver itself is missing steps and not signaling an error but let's assume for a moment you are correct, the driver is missing steps and not signaling an error.

Just looking at your test-bed I see some potential issues (see first and third) that would cause lost steps without generating an error.

First, use shorter wire jumpers, the noise introduction on those wires would be horrendous and could account for the error.

Second, I hope your using really high speed opto-couplers for the encoder inputs or you have switched to a line receiver or this could account for your errors.

Third, your HEDS encoders are optical and exposed so at high speed ambient light might be affecting the encoder, wrap it in shoe-box cardboard to (shiny side out).


You have Mach3 motor config set so that 25mm = the real encoder CPR (or 1in = the real encoder CPR), if you tell it to move to 25mm (or 1in), it should make one complete revolution, if you tell it to go to 250mm (or 10in) it should make 10 complete revolutions and so on and so forth and at 25000mm (1000in) it should make 1000 complete revolutions and it should stop at the index line.

First try it at slow speed, feedrate of say 5in a min, then 50in a min then 500in a min, the final results should be identical if the driver is working correctly.

If you are still having positional errors and your jumper wires are 3in or less, you have covered your encoder and using high speed opto-couplers, I would connect a scope and look for noise and I would also check the shifters to ensure they are working correctly and fast enough for the application and personally, I would have stayed with the high speed opto-couplers for the step/dir signals and I really recommend the line receiver for the encoder inputs, their extremely fast and require nothing but the IC to integrate into the existing circuit.

[dwalsh62]

This my current situation I am using mach3 and as you can see in video that first I am giving it one rotation and afer that ten rotation and last I am giving one thousand rotation and as you can see servo is losing position ! I dont understand what is happening with it I am using 5 volt to 3.3 voly logic level shifter not optocopler

https://learn.sparkfun.com/tutorials...r-hookup-guide

and using smps for motor power and avago hedl encoder and inkjet printers motor. please somebody please solve this puzzle

This is my video showing problem :

https://m.youtube.com/watch?v=ymQ-GOiosBE

Please!!!!

Your output stage test-board doesn't look familiar, is this something you made or is it something you purchased and information about it's characteristics might be a contributor to your missed steps?

[_ID_]

When all is said and done, I may ask _ID_ if he is willing to redo my layout, if he says no, maybe someone else who's really good at it can give it a shot but I think it would be beneficial for this project to be offered with the utmost quality and integrity, I hate settling for "it works" kinds of stuff and this forces me to seek out better options.

I want this to be that better option, I want the Mihai project to be the standard for others to attempt to rise to rather than be a sub-standard used for comparison and I think others feel the same way.

Overall, this is a good project base to start from and this has been confirmed from several sources, some of which I trust so can we get past the pissing contest and back to focus on the project please.

I would really like it if you continued to contribute Mihai but if you have no interest then perhaps someone like jfstockton can take over where you left off and you can help him get up to speed by answering his questions, privately if you have no public interest, it's your call, after all, it was your project to start with and credit goes to you for it.

Hi Dale. I can redo your layout, no problem. It will take a bit because i'm quite occupied with my job at the moment. Just send me the files or put them hire in the thread. But then I ask my self if it's the right thing to redo the board now as some of you claim that it's possible to take the whole thing to the whole new level by using different pins from microcontroller and rewrite a bit of code. If this is going to happen then I will wait for the update of the code and than redo the PCB. Who has the knowledge to go through the code and make changes!? Till now only roivai did something on the code-that I know of. Anyone to cover this part?

[dwalsh62]

Hi Dale. I can redo your layout, no problem. It will take a bit because i'm quite occupied with my job at the moment. Just send me the files or put them hire in the thread. But then I ask my self if it's the right thing to redo the board now as some of you claim that it's possible to take the whole thing to the whole new level by using different pins from microcontroller and rewrite a bit of code. If this is going to happen then I will wait for the update of the code and than redo the PCB. Who has the knowledge to go through the code and make changes!? Till now only roivai did something on the code-that I know of. Anyone to cover this part?

Fantastic _ID_, thank you.

Currently we are looking at some additional changes, removing a no longer needed optional line due to another option choice and adding an encoder index signal so as soon as all the changes have been finalized I'll shoot you the eagle schematic.

I'm waiting to hear back from ST on the best pin choice from the available pins so we'll see what they have to say about it.

We've decided to use the F103T8/F103TB based on ST's recommendation due to it's package type, everything will be onboard and isolation protection on the USB plug will prevent any ground-loop shorting issues.

When you're all done routing the board you can just directly post the eagle files since it is all open giving everyone easy access to them.

[tandel]

Your output stage test-board doesn't look familiar, is this something you made or is it something you purchased and information about it's characteristics might be a contributor to your missed steps?

I have made this output stage same as mihai had given I am not using optocoupler because I have pc817 I think its not fast inough i using logic level shifter insted of optocoupler
This is my lovic level shifter for step dir and encoder interface i have connected en pin to 3.3volt so its always remais enabled

https://learn.sparkfun.com/tutorials...r-hookup-guide

I have hedl 500cpr so it becomes 2000 ppr so I have setted 2000steps in mach3 so g0x1 gives me one revolution

I am powering power bridge with 7812 because my motor is 12v and 7812 is getting power from 24volt smps 3amp
My current pid tunning p:10 I:1000 D:10000 intregal limit:2000
Max current limit:1000 mA

I have compiled code in eclipce and I got hex but I got 100 to 200 warnings
I also used pecomipled hex of mihai but same result

[dwalsh62]

I have made this output stage same as mihai had given I am not using optocoupler because I have pc817 I think its not fast inough i using logic level shifter insted of optocoupler
This is my lovic level shifter for step dir and encoder interface i have connected en pin to 3.3volt so its always remais enabled

https://learn.sparkfun.com/tutorials...r-hookup-guide

I have hedl 500cpr so it becomes 2000 ppr so I have setted 2000steps in mach3 so g0x1 gives me one revolution

I am powering power bridge with 7812 because my motor is 12v and 7812 is getting power from 24volt smps 3amp
My current pid tunning p:10 I:1000 D:10000 intregal limit:2000
Max current limit:1000 mA

I have compiled code in eclipce and I got hex but I got 100 to 200 warnings
I also used pecomipled hex of mihai but same result

A lot of the warnings are casting issues which can be ignored, a couple are global type casts which resolve at link time but I chose to fix them in the headers as I hate to see so many warnings but these wont contribute to your problem, only make compiling cleaner.

I think your PID tunning is a little off but I'll put a motor together to match what I see in the vide and see what I come up with, it may be OK.

I've ordered a couple of those bi-directional logic level shifters which I think are not the correct way to go but as soon as I have everything to duplicate your test-bed we can dig a little deeper into the cause of your problem.

[dwalsh62]

This my current situation I am using mach3 and as you can see in video that first I am giving it one rotation and afer that ten rotation and last I am giving one thousand rotation and as you can see servo is losing position ! I dont understand what is happening with it I am using 5 volt to 3.3 voly logic level shifter not optocopler

https://learn.sparkfun.com/tutorials...r-hookup-guide

and using smps for motor power and avago hedl encoder and inkjet printers motor. please somebody please solve this puzzle

This is my video showing problem :

https://m.youtube.com/watch?v=ymQ-GOiosBE

Please!!!!

I've pretty much collected all the parts including the motor which looks exactly like yours robbed from a defunct Lexmark printer but you don't seem to be too interested in solving your problem or you've already solved so into the junk-drawer they go.

[tandel]

I've pretty much collected all the parts including the motor which looks exactly like yours robbed from a defunct Lexmark printer but you don't seem to be too interested in solving your problem or you've already solved so into the junk-drawer they go.

no please i need help i am still getting problem need your help

[dwalsh62]

no please i need help i am still getting problem need your help

A detail of the problem requesting assistance publicly is OK but solving your problem should be private to avoid a lot of nonsense post that would cause confusion to other people.

Look at your private messages here on CNCZone, I sent you some, respond privately, when the problem is solved an update to your problem post will help anyone else experiencing similar issues.

[dwalsh62]

no please i need help i am still getting problem need your help

tandel, I can't help you because you don't read your private messages or reply to me and I can't sit around waiting for you so you're going to have to get someone else to help you, good luck with solving your problem and I hope you find the help you need.

[paul3112]

Is that a soft limit issue in Mach?

[dwalsh62]

I fixed up _ID_'s layout, replaced two opto couplers with a line receiver for the encoder inputs and stopped the important traces, unfortunately _ID_ went ahead and produced 10 boards before I fixed up the board so he's going to make an adapter to add the line receiver but I've already tested the altered design and it is an improvement over the Mihai board but it still doesn't give you optimal performance due to pin selection.

As well, I've completed two designs using the STM32F103T8U / STM32F103TBU, one design utilizing an encoder with an index signal so other than pin changes, jfstockton is looking into the dual current sense, the dual back-EMF sense and the encoder index code which is all that needs to be done and a high performance commercial grade inexpensive driver is a reality.

The design with encoder index signal support is using the MAX3097ECSE / MAX3098ECSE A/B/Z encoder IC so one thought is to use this single design and add a config option to enable/disable the encoder index signal so a simple basic RENCO or HEDS type encoder can still be utilized.

While I'm not making use of the IC's encoder error output signals, this is the only improvement that can be added which is basically a line that signals one of the encoder lines is in error and it's possible to add 3 LED's to show which line has generated the error so I'll talk to jfstockton and get his take on it.

At this time, there is little else to do with the electrical design so progress with the code is pretty much mandatory to finish off the driver.

[dwalsh62]

Is that a soft limit issue in Mach?

Now that's funny...

No, his driver is missing steps and I've tried to help him but he doesn't seem to have the time to fix it now so he'll have to find someone else who can work with his lethargic time schedule and his failure to communicate in a timely manner.