[joeybagadonuts]

Websrvr

There are many people here capable of coding however you may have to wait until they have time unless you are capable of doing it yourself.
Live from downtown Burbank home of the comedy club doughnut shop.

JoeyB

[websrvr]

As I understand it, the OP is in charge of the thread. He takes it however far he wants to. If others assist and the OP wishes to push it further, then that is what happens.
There is no obligation for the OP to "develop it further" if he chooses not to.
Anyone can take the bulk or all of the open source project and start a new thread to push it further, or in a different direction.


The OP has not responded for some time, so I take it that further work is not planned, or available, at this time.

Not exactly the news I was hoping for, I scanned back and the OP (Mihai) hasn't posted anything in about 6 months.

SInce I do know him by e-mail (although we haven't spoke in several years) I'll get in touch with dwalsh62 and see what additional information he can provide to me about the pin changes and the updated schematic that may be beneficial.

I after studying the schematic I figured out a wire patch that allows me to use the Mihai code with minor modifications to the current sense code to compensate for the current sense amplifier and the DC-link, encoder index and encoder error lines are ignored so the board is usable with the old code until new code that takes advantage of all the changes is available.

I've come across several servo related projects at github which make use of the encoder index signal so I'll spend the next couple of days studying these code snippets as I delve into the STM32F103 datasheet for additional information that may make it clear for the required code changes.

[websrvr]

Things are going from bad to worse, I got an e-mail back from dwalsh62 which wasn't exactly worded with any positive encouraging phrases which I can't repeat and it may be a some time before Mihai responds back to me so it seems we're on our own.


I've downloaded a few of the STM32 primer courses and motor related tutorial projects and over the next few weeks I'll attack them and see where it leads me.



My assembled board is using IRF740's and they seem to generate a lot of heat at 4A (can't touch them), the fets listed in the schematic are not a local stocking item and Arrow wont sell them in quantities of less than 1000 so I'm looking for a small quantity source if anyone has one.


The changes to the schematic give me a flood of questions which may or may not be related or important to the project and I have no clue who to turn to for answers.

Is anyone knowledgeable about the recommended pins who can at least give some advice, explanation and/or code fragments that may be helpful?

Of specific interest is why the change to dual current sense lines on B and C channels and not triple current sense lines since there are three PWM channels and the purpose of the DC-Link sense line which doubles as Back-EMF sensing.

A quick glance at the Back-EMF sensing example appears to be nothing more than a high-voltage sense with the mean voltage being 90VDC (we're using 36VDC) but the example code doesn't do anything based on the results of this over-voltage sensing except turn on an LED which we don't have so what exactly are we supposed to do in this case and at what trigger voltage are we supposed to take action at since the example targets 100VDC as the trigger voltage?

One of the motor tutorials does some PWM wave shaping to generate a sine wave output feeding an L6234 triple bridge motor driver IC so AC/DC operation can be obtained, should we be considering such shaping or it doesn't apply at the voltage and/or driver stage we're targeting since we turn the output stage on and off as hard as possible to produce a constant torque type of environment?

[websrvr]

One more thing to note, I found an ST motor tutorial and it's not using the PWM compliment channels, it's using gate driver IC's which have one non-inverting and one inverting input to drive the motor output stage, this would free up three MCU lines, is this something worth considering or moot based on the ST recommendations?

[websrvr]

I am interested in this project, I would like to see it blossom into the swan it can be but with the loss of dwalsh62 and no apparent interest from Mihai, this looks to be an uphill battle and quietly sitting on the sidelines is no longer an option that I've got to become involved to aid in it's progress.

I'm prepared to assist to the extent of my abilities and I'm willing to work with anyone who has the same goal/interest in this project and I have no compunctions to learning whatever is required to aid in achieving this goal and have already put a plan together.

Despite anyones opinion of dwalsh62, I will be consulting him if he is open to assisting but I'm not holding my breathe in the hopes he still has an interest and if no one likes the idea of this, sorry, but the project is more important than a biased personal opinion of a man who has been unjustly treated.

[paul3112]

I hope this link is not "off topic".
https://github.com/rene-dev/stmbl

As I see it , the scope of motors that this will drive is very nice ( steppers being the exception)
I have not seen or used this driver so the link is based on the info @ github / stmbl

Paul

[websrvr]

I hope this link is not "off topic".
https://github.com/rene-dev/stmbl

As I see it , the scope of motors that this will drive is very nice ( steppers being the exception)
I have not seen or used this driver so the link is based on the info @ github / stmbl

Paul

Hi paul3112, welcome to the discussion.

I've already grabbed this project for examination in the hopes that some of it would apply such as the encoder index signal and DC-Link sensing but having dual SMT32's in a low-voltage BLDC/DC Servo Motor driver seems like a lot of overkill and expense not to mention the MCU inter-communication timing requirements for smooth operations.

Even though the STMBL project seems to be operating under significantly higher voltages, I think it's still useful reference material if nothing else.

I did some searching at github, seems to be quite a few unfinished servo motor projects over there, found roivai's project and was wondering why he hasn't jumped on this one, moving the code into his repo as an alternate build would give it a stable home and I don't believe Mihai would object (but I could be wrong).

I'm also noticing a trend of the lesser experienced or beginner driver designers with the use of HALL sensors and dual channel encoder and the better quality drives are using triple channel encoders without HALL sensors, should we consider adding HALL sensors or stick with the ST provided schematic and do what the better drivers are doing?

I have started the STM32 primer course, estimated time for completion is 16 hours so it may take me a couple of weekends to get through it before moving to some of the motor tutorials and I'm considering skipping the ESC based tutorial since it doesn't specifically apply to motor positioning applications but I'll still glance over it for relevance just in case.

Is ST pretty much the only source for STM32 tutorials or is there another source I should be looking at as well?

It's unfortunate I only have the weekend to work on this stuff and I hope I'm not the only one making any effort in this project which at this time is an unknown and, since no one seems to be talking about any coding I'm wondering if my efforts are pointless, some encouragement and assistance would go a long way right about now, I feel like I've jumped overboard and looking for someone to latch onto without realizing the amount of work and time that is going to be required which seems to be mounting and I've just started looking at things and already overwhelmed.


On another note, I got an e-mail back from dwalsh62 and he has emphatically stated he has had no contact direct or indirect with joeybagadonuts since his banning and any claim of contact made by joeybagadonuts is an outright lie (I really wish I could repeat verbatim his response).

He has also informed me he has no interest in helping complete this project on the Zone based on the treatment recently received so if I get a similar response from Mihai we're going to have some serious problems with this project and a project leader/manager will become a priority.

I can't manage the project, I'm a solder in the field doing grunt work and have no experience leading or managing any projects so I wouldn't be a good candidate but, roivai, if you've set your project aside for a rainy day project and we don't get a favorable response from Mihai in the next two weeks (gotta give the man reasonable time to respond), would you consider taking the lead on this project as you seem to have some idea what's going with a motor driver and I think you're a viable leader candidate because of your github project?

[paul3112]

Thanx for the reply Websrvr,

I came to this tread looking for bldc drivers that will work in torque mode.
If this thread was to set some boundaries for the type of motor that was to used and set some voltage / current limits, the type of control signals then I think it might be advantageous.
There is a great deal of info for ST micro dev boards ( code and circuits) I am looking towards the stm32f4 dev boards.

At the moment the task/ scope of this thread is a little too broad,at beast, and undefined at worst. I don hold that against anyone at all. The zone is a collaborative thing, not a production design studio.

My requirements are well outside what I see in this tread so I cant assist based on the fact that I can just follow the bouncing ball not throw the ball!

Regards

Paul

[TOTALLYRC]

I would love to help but I have no experience in any of whats going on but I am enjoying(if that is the right word) following along.

[cjmerlincnc]

I also wish I had time to contribute but I am very interested in the project and watching with interest.

[roivai]

I can't manage the project, I'm a solder in the field doing grunt work and have no experience leading or managing any projects so I wouldn't be a good candidate but, roivai, if you've set your project aside for a rainy day project and we don't get a favorable response from Mihai in the next two weeks (gotta give the man reasonable time to respond), would you consider taking the lead on this project as you seem to have some idea what's going with a motor driver and I think you're a viable leader candidate because of your github project?

Thanks for those words. Actually I have very limited knowledge about motor control. I've been trying to study something, but the more I read, the more confused I get.

Unfortunately I have no interest in leading any sort of a project, I have plenty of other ways to make my spare time very minimal. I found it even a bit strange that this thread which started as one guy presenting his work should now be forced into a project (led by others)?

Anyway I'll probably do some work in the BLDC world in the near future as I now have some real hardware and a router to play with. But that is completely based on my own needs and schedules. I'm happy to share if I manage to make something spin and if someone finds it useful it's nice, and if someone makes it better and contributes back, it's even nicer. But that's it, there is no need for a project from my side.

Pekka

[KOC62]

I had a small project to control an old electric drill for very slow turning but used a triac type controller. I joined this thread mainly to grapple with PID. But I finished my task and have no need for mihai's BLDC controller, except as an educational lesson. I don't have a BLDC motor to practice on.
I would put further time into this project if I were to replace my CNC's four 425 oz-in steppers with BLDC servos, but that isn't in the works in the foreseeable future.

[websrvr]

I just received an unfavorable reply from the OP Mihai, he has no time or interest to work on the project which means we're on our own.

I've noticed a trend in people agreeing to aid in this projects development but within a 30 day period they vanish leaving everyone high and dry.

dwalsh62 has commissioned firmware for his driver and the source wont be publicly available and has arranged recouping this cost by selling the design and code to a china production house and I find it a little frustrating that he's gone this route but it does show me that a coding solution for all the changes is just a matter of figuring out what needs to be done.

I went over to vist with dwalsh62 tonight, he's pretty bitter about this whole Zone nonsense and this makes it hard for me to take sides when I have a personal interest in this project.

I had a look tonight at dwalsh62's board, very impressive, sectioned and outlined silk screen, pretty translucent red color and not the sickly purple OSH-Parks is known for.

If you price out the cost of the parts in quantities for assembling less than 500 boards such as a quantity of 10 and you produce your own high quality boards you will spend more than this, OSH parks was $75.00 for three 1oz two-layer boards and $90.00 for three 2oz two-layer boards, the parts to assemble 5 was just over $190.00 excluding fets and I have a spare board suitable for framing but no where near the quality of the board dwalsh62 produced.

dwalsh62 has a far better layout than the one I did and he's offered myself and a few others his populated boards without his code or fets for $350.00 for a 10pack, far cheaper than it cost me to produce the smaller quantity I did so I've placed an order for two packs for personal use which he says will most likely be run within the next two week so if anyone else is interested in some boards I advise you to contact him before he does the run but be prepared to purchase a pack, he already stated to me he wouldn't be selling then individually or with his code which means you buy a pack or two and wait for the code to be completed or you buy a pack (or more) to use for development and personal use or make a group type purchase if you don't need a small quantity on hand or want to reduce your personal costs.

I just started the PWM tutorials and it has given me some insight on how to tailor the pulses specifically for DC motor application to increase starting torque but since no one here seems interested in the projects development and as a public project, it's clearly not a one man operation, I'm wondering if my effort is warranted or I should just go off and do development away from the Zone for my own personal use.



So I have questions after completing some of the tutorials that are causing me some problems and confusions.

Based on a HALL sensor servo tutorial I found, one area I am struggling with is the use of NMI's and MI's in the PWM code and why both are used, would it still work if both the general PWM and PWM-A interrupts were NMI (or both MI) or is there a specific reason that the general PWM interrupt is an NMI and the PWM-A is an MI and why does Mihai's code only have the general PWM NMI?

Looking at Mihai's code, the default PWM interrupt registers are default set to trigger on the peak leading edge and the current code does not modify their condition, since this is a reversed sawtooth signal with no wave shaping, wouldn't it be better to trigger on the falling or trailing edge of the pulse?

Since I can see from the motor driver stage a sine wave would not be beneficial in this DC only design application, would there be any advantages to use a normal sawtooth with a hard off or does this defeat the gate driver off transition which deactivates at 55% (below 2.5v) of the input signal?

What about the use of a square wave, while none of the examples/tutorials use a square wave, in this particular case, wouldn't a hard on/hard off be more in line of the requirements of the gate driver IC in a DC only application?

And last, the Mihai PID code is also confusing, it seems to be doing most of it's work in the PWM dead zone (PWM signal dead time) based on the NMI's timing, wouldn't it be better to do the work during the peak zones (above 55% signal or above 2.5V on primary channels) or peak inverse zones (above 55% signal or above 2.5V on compliment channels) or does this create more unnecessary stringent timing issues that are beyond the scope of the requirements?

It seems the more I learn about this project and it's requirements, the more confusion is thrown into the mix.

[paul3112]

sorry Web I missed the stm info request for tut's
One should keep in mind that most of my info is based on the STM32F407 disco board ( heaps of grunt ..cheap as.. usb onboard.. and a couple of free ide's around < uvision prefered)
STM32F4 Discovery - Libraries and tutorials for STM32F4 series MCUs by Tilen Majerle. Working with STM32F4xx series and Standard peripheral drivers (STD, SPL) or with STM32F0xx, STM32F4xx or STM32F7xx using Hardware abstraction layer libraries (HAL) f
STM32F4 Boards - Arduino for STM32 Arduino ide for the f4

Paul

[jimmkaz]

websrvr,

Please post a picture of the board you are selling, I might be interested in purchasing a few dozen.

[websrvr]

websrvr,

Please post a picture of the board you are selling, I might be interested in purchasing a few dozen.

First, I'm guessing you really didn't read what I had written and you made the incorrect assumption that I'm selling boards, I'm not selling anything so why would I have a picture to post of a board I'm not selling?

Secondly, I had a preview of his sample boards and my opinion is that they are way more professional in design, layout and quality than the board I did so I'm using the ones I made as my sample/dev boards and will use his boards as my personal driver source stash because I know he wont be stocking them for later sales, he's a person not a business.

Third, if you think that you can wait until others have received any boards then decide to purchase some, I wouldn't bank on that if I were you, why would he invest money to stock a product he has no intention of publicly selling in any business capacity when he's a private individual, the production will be based on the quantity ordered/needed and the price being offered is less than cost of producing a small quantity for yourself and it is very unlikely he will order 100pcs to stock just in case six months or a year down the road someone like you wants to buy a pack of 10.

Since you're obviously an untrusting soul and my explanation of what I had seen while brief is from a person who was a trusted public figure around here, making available his board as a courtesy is his choice and if it were me selling the boards, I wouldn't bother dealing with someone who has trust issues, I have no interest or desire to make you or anyone else feel warm and fuzzy about buying anything I would be selling, that sounds like the job for a professional sales person although, I might be more inclined to post a pic for inquiring minds since I'm not a high profile person as the person who is offering the boards and my trust is yet to be established at the Zone because I'm not a high profile public person.

I have no need or desire to bother the man who is offering the boards to demand he go take a picture of his board just for you or anyone else and seeing the eagle layout pic wouldn't satisfy your morbid curiosity so my advice to you is that you just go make your own, the schematic is in post #296 and since I was able to do my own layout, I'm sure you can do a layout that is acceptable to you then you can offer your board in the manner in which you prefer under the conditions you find acceptable.

I certainly don't recommend anyone contact the man if they have any trust issues, see his posts, he's not really average people friendly, they call him Admirable Einstein for a reason, he's harsh on a good day, pummeling on a bad day and not tolerant to insecure people who have no clue or knowledge about they want or what is being offered so if you're thin-skinned run for the hills is your better option.

While he stated that _ID_ can do a layout well, in comparison of the two layouts, several people have told me his layout is far better and while I wont be getting the mating eagle layout file (I don't need it if I have the boards), someone else does have the eagle image file for the layout and if I really need it so badly, I would probably spend the time placing and routing the traces based on the pic if I was really desperate enough and had a month or two to play with it to recreate the mating file.

I've seen a post from another user in this thread but don't recall who or the specific post as it wasn't relevant at the time, they praised him about their own layout which he adjusted and I'm sure they would agree any layout he does work on would be outstanding but hey, don't take my word for it, make your own and you wont have anything or anyone to worry about criticize or blame.

A user sent a private message to me, one of the things they pointed out is that I seem to be a very long winded individual, more-so than anyone else in the thread so I've gone over the posts I've been making and I see what they are basing their conclusions on, unfortunately I don't know how to make my posts any shorter and have people understand what I'm trying to get across or explain and as we've already seen, one person has missed the boat completely and unlikely he's the only one so, the lesser of two evils is long winded posts where the majority can comprehend the content, the rest can move on to the next post.

Thanks paul3112, I settled on eclipse because it's mentioned in this thread and the OP's source is an eclipse based project and installation in win7 was painless and simple.

My concern about the F4 based projects and tutorials is the compatibility the resulting code will have with F1 based products, basing a preference for F4 dev boards based on onboard USB makes no sense since all of the F1 dev boards I've seen have onboard USB so it makes no sense to limit oneself to just F4 based dev boards, of course this is just my opinion.

I've noticed that Tilen Majerie tutorials are F3/F4 based and some are not at all compatible with the F1 so I'm trying to avoid such tutorials to prevent any working/learning issues.

[paul3112]

STM32 Motor COntrol

[jimmkaz]

First, I'm guessing you really didn't read what I had written and you made the incorrect assumption that I'm selling boards, I'm not selling anything so why would I have a picture to post of a board I'm not selling?

Secondly, I had a preview of his sample boards and my opinion is that they are way more professional in design, layout and quality than the board I did so I'm using the ones I made as my sample/dev boards and will use his boards as my personal driver source stash because I know he wont be stocking them for later sales, he's a person not a business.

Third, if you think that you can wait until others have received any boards then decide to purchase some, I wouldn't bank on that if I were you, why would he invest money to stock a product he has no intention of publicly selling in any business capacity when he's a private individual, the production will be based on the quantity ordered/needed and the price being offered is less than cost of producing a small quantity for yourself and it is very unlikely he will order 100pcs to stock just in case six months or a year down the road someone like you wants to buy a pack of 10.

Since you're obviously an untrusting soul and my explanation of what I had seen while brief is from a person who was a trusted public figure around here, making available his board as a courtesy is his choice and if it were me selling the boards, I wouldn't bother dealing with someone who has trust issues, I have no interest or desire to make you or anyone else feel warm and fuzzy about buying anything I would be selling, that sounds like the job for a professional sales person although, I might be more inclined to post a pic for inquiring minds since I'm not a high profile person as the person who is offering the boards and my trust is yet to be established at the Zone because I'm not a high profile public person.

I have no need or desire to bother the man who is offering the boards to demand he go take a picture of his board just for you or anyone else and seeing the eagle layout pic wouldn't satisfy your morbid curiosity so my advice to you is that you just go make your own, the schematic is in post #296 and since I was able to do my own layout, I'm sure you can do a layout that is acceptable to you then you can offer your board in the manner in which you prefer under the conditions you find acceptable.

I certainly don't recommend anyone contact the man if they have any trust issues, see his posts, he's not really average people friendly, they call him Admirable Einstein for a reason, he's harsh on a good day, pummeling on a bad day and not tolerant to insecure people who have no clue or knowledge about they want or what is being offered so if you're thin-skinned run for the hills is your better option.

While he stated that _ID_ can do a layout well, in comparison of the two layouts, several people have told me his layout is far better and while I wont be getting the mating eagle layout file (I don't need it if I have the boards), someone else does have the eagle image file for the layout and if I really need it so badly, I would probably spend the time placing and routing the traces based on the pic if I was really desperate enough and had a month or two to play with it to recreate the mating file.

I've seen a post from another user in this thread but don't recall who or the specific post as it wasn't relevant at the time, they praised him about their own layout which he adjusted and I'm sure they would agree any layout he does work on would be outstanding but hey, don't take my word for it, make your own and you wont have anything or anyone to worry about criticize or blame.

A user sent a private message to me, one of the things they pointed out is that I seem to be a very long winded individual, more-so than anyone else in the thread so I've gone over the posts I've been making and I see what they are basing their conclusions on, unfortunately I don't know how to make my posts any shorter and have people understand what I'm trying to get across or explain and as we've already seen, one person has missed the boat completely and unlikely he's the only one so, the lesser of two evils is long winded posts where the majority can comprehend the content, the rest can move on to the next post.

Thanks paul3112, I settled on eclipse because it's mentioned in this thread and the OP's source is an eclipse based project and installation in win7 was painless and simple.

My concern about the F4 based projects and tutorials is the compatibility the resulting code will have with F1 based products, basing a preference for F4 dev boards based on onboard USB makes no sense since all of the F1 dev boards I've seen have onboard USB so it makes no sense to limit oneself to just F4 based dev boards, of course this is just my opinion.

I've noticed that Tilen Majerie tutorials are F3/F4 based and some are not at all compatible with the F1 so I'm trying to avoid such tutorials to prevent any working/learning issues.

So you are suggesting we send money to someone for a product sight unseen? That doesn't sound very appealing, can't you show us a picture of the board you have in your possession so we can evaluate the quality for ourselves? It only takes a few seconds to snap a picture, that's not asking very much. Typing long winded retorts require a lot more effort than that.

[KOC62]

So you are suggesting we send money to someone for a product sight unseen? ....

Yes. Or do you have a "trust" issue?

[jimmkaz]

Yes. Or do you have a "trust" issue?

KOC62

As far as I am concerned this is not a trust issue, it would be nice to see the quality of the finished board is that not a lot to ask for?
Secondly, this project is supposedly open source. Please correct me if I am wrong shouldn't the gerber file for the board's suggested be posted to this thread?