[KOC62]

KOC62, so you're a win10 lover, ...

Nope. When my old laptop died a new one came with WIN10.

...In case you're wondering, I answered your PM in public because the questions will help others who are contemplating the same questions and I believe they are relevant....

No problem.

...I wasn't impressed with KICAD myself, it's a really different environment and if you're comfortable with eagle or have good help (like me) switching to something else is frustrating and unless you absolutely have to, switching isn't appealing....

I didn't know either package well, so it wasn't a switch. Many of us are reluctant to change once we have overcome the basic hurdles. I get that.

...I can rename D10 to D4 and renumber the resistors and capacitors if it makes you feel better

Don't need to. It was that at the time I wasn't sure if I had a complete schematic and thought to check with you. I did it private because it didn't seem to pertinent to the mihai board discussion. My further perusing the schematic made me think of the reason for the gaps as you just explained as a possible scenario. I do appreciate your feedback. Thanks again...

[websrvr]

It doesn't matter what you say to Mihai, he was clear in his communication to me that he's done with the project and has no intentions of continuing it so I find it highly doubtful he will return any time soon just because jimmkaz or someone else asks him nicely.

Mihai had an idea, he turned it into a proof of concept as an educational aid, here's the basic status of the project, does the project work, yes, does the project have issues, yes, are the issues correctable, yes, has the electrical design been updated, yes, has the firmware design been updated, no, can the firmware design be updated, yes, is someone developing the firmware now, no.



_ID_ hit the nail on the head with his comments about updating and I full heartedly agree, no matter what you do, you need to update the electrical design and you need to update the firmware if you wish to correct the flaws and you can't do one without doing the other.

I don't get why dwalsh62 didn't write the firmware himself, he's a competent programmer and I've seen his work but this STM32 thing causes him to mind-freeze like it's the plague or something and he couldn't get past this block, probably age plays a big role in this thought concept of "too old to learn".

For most, updating the electrical design is a possible option based on adding features or functionality and fixing issues related to performance and/or functionality but the best possible electrical update has already been done by the producers of the MCU while keeping it inexpensive so it is unlikely you could improve upon it without making it way more expensive and way more complicated than it needs to be.

Oh, the switch to the current sense IC's is a definite performance enhancement and this is how the original design should have been done.



dwalsh62 has been building motors for a couple of years now with the help of a motor specialists and has made some very impressive motors, you've seen some pictures floating around here of some of the motors he's made, he uses 4 pole encoder on 2 pole motors, 6 pole encoders on 6 and 12 pole motors and 8 pole encoders on 4 and 8 pole motors and until I met the motor specialist I wasn't aware that an encoder had poles or that it even mattered in a motor but he showed me it does.

He gave me (let me take from a trash pile) a spindle motor he said was a prototype that he claimed didn't make the grade, the following is the confirmed specifications by the motor specialist who performed the tests, dim 72mm x 72mm x 220mm (total length including encoder cover and shaft), 4-poles, Inertia: 84450cpt (no clue what his is but it's on the dyno printout), Rotaitonal MASS: 57oz/1KRPM, Magnetic GAUSS: 116/50%PWM, Signal: Trapezoidal, Power Input 220V/9.17A, Power Output: 11,562RPM/1.56KW (2HP), it failed because it was supposed to generate 1.875KW (2.5HP) and >12,000RPM but considering it's size is smaller than any 1/6th HP single phase motor and weighs less than 8lbs, how can it be a failure when a Shirline mill or lathe motor is 60W and weighs 6.2lbs and is larger so I yanked it from his trash pile because it looked like a commercial made/grade motor and not one someone made and the output more than adequate for any small machine.

There were also concerns raised by the motor specialist regarding the commutation table, he stated it isn't optimized for or compensates for any changes in motor poles and really only works with 2, 4 or 8 pole motors which covers most BLDC servo motors and anything else is operating under a 3rd or 4th gen harmonic and his opinion was it didn't need to be changed, just select an appropriate poled motor or bloat the code with multiple tables.

The motor specialist is a friend who deigns and builds motors for a living and I consider him to be GOD when it comes to motors and don't question his expertise anymore since he's proven everything he's every stated when I did question them, he also says in BLDC and AC motors, the encoder must be matched for the poles in the motor, I don't get how an encoder is affected by poles so I just accept it as the word of god and move on because I don't see the correlation.

_ID_ is close to determining the number of poles in a BDC motor except the rule is, count the number of magnetic poles in the rotor or stator, or, apply a small voltage to a single coil and count the number of pulled positions in 1 revolution.

If a BLDC motor has a non dividable by 3 slot count and induction type then without the design data it would be impossible to determine the number of poles visually and test equipment will be required if it is an induction type.

3-PH BLDC motors are a little more complicated then BDC motors and if the motor has no magnets then even harder (no magnets is induction type and can be operated from AC or DC), you have to know how it is wound, the coil connections (WYE half, WYE full or DELTA), the number of stator slots and the number of rotor bars, the only known value that is consistent is 3-PH but could be wound in either lapped or basket in sequential, split, staggered or offset to make it fit a particular count of stator slots or to achieve a specification outside the norm.

Lapped wound staggered coils are primarily used in explosion proof motor design and high efficiency 2 and 4-pole induction BLDC motors, are more time consuming to wind and install thus reflecting the higher cost and must be manually installed in the stator, each coil is bifilar wound, the leading coil sides are all first inserted into their respective slots, the tail coil sides are overlapped into a slot that already has a coil winding in it until all the coils are installed and then the exposed coil tops and bottoms are shaped with a non-metallic former to give it that nice symmetrical/balanced appearance, coated then baked but the end result is more torque due to even distribution and saturation since every coil is identical in size and physical position in the stator relative to the rotor location reducing hot spots and uneven saturation.

This coming weekend I will be assembling my first set of three BLDC lapped wound induction motor, while I didn't design them or even contribute in any way, he wants me to experience the effort required to assemble a lapped wound induction servo motor so I can appreciate the results of such labor and see the benefits of the design over the traditional basket wound types which he claims will be prevalent.

Basket wound can be machine wound or even threaded directly into the stator reducing build time and cost but a tradeoff in performance since the coils are not all identical in size, shape and position relative to it's location in the stator slot since the entire coil is placed in it's slots before moving on to the next coil.



A simple 2 pole to 4 pole conversion you can make to most BDC motors such as ElectroCraft BDC E240, G730 etc... (if it looks like these ones it cna be modified), is to disassemble it, mark the position of the magnets for later reassembly, measure the gap between the two magnets at both sides, sum, divide by two, this is your pole gap, remove the retainer, remove the two magnets, cut them so you have 4 magnets and the gap between them is the pole gap you calculated, cut some phenolic material the same width as your pole gap plus a tongue on both sides of 1/16 x 1/16 width and at least 80% of the magnets height, cut a 1/16 x 1/16 groove in the edge to accept the tongue, assemble, if you want to increase the low-end torque, drill two holes matching the current ones for the brush holders at 90deg from the existing one and don't forget the set screw hole to hold the brush holder in, wire them in parallel and now you have converted it to a 4-pole version of the motor without giving up anything but a couple of dollars and an hour of your time. I've done several using a knee mill and it took about 47 minutes for each motor from start to finish.

Adding the second set of brushes only helps with the low end torque, when it's spinning fast enough the gain is negligible and the no load current is the only real affected specification but of course you can also hit double the load current for more power but I found the tradeoff of turning a 250W motor into a 500W motor is heat so I limit to the 2 pole load specifications or 25% above it without heat issues.



While I hate to see a project struggling and begging for development, I still think this project has potential as an open source project but you have to set up an environment that is conducive to it's development, someone who knows how to lead a team is mandatory, someone who just wants to be boss but lacks the knowledge and experience are a dime a dozen and as we've already seen, it's easy for one of those types to quickly kill it's development in a heartbeat by making a poorly thought out decision.

At this time the best you can hope for is to find someone willing to work on the open source firmware AND YOU LET THEM WORK ON IT GIVING THEM WHATEVER HELP THEY NEED OR REQUIRE.

This is only going to happen if someone steps up to the plate to lead the project and of course those who want to lead now shouldn't be but no one with the skills is offering so stagnation until this changes is what your doomed to accept.



Boards suitable for development are already available, two were assembled using a hot-air reflow pencil, three using and IRDA station and five using a toaster oven (Reflow Mansion firmware, 3.95in TFT touch, MAX6675 + K Thermocouple and an Uno), the toaster oven was the fastest at 3 board in 8 minutes and gave the best/cleanest looking finished result.

The only problem I see with people who are working on or wanting to work on the firmware is that those who should be breadboarding are using boards because they can't even make the effort to stick to the electrical design and are degrading the design but lack the capacity to realize it's wrong and trading out values and bypassing IC's should not be done and those who should be using boards are breadboarding because they have this innate sense of "saving money" by not buying a board without realizing they're spending more cause their piecing it together as they go and taking a lot longer to get to the point of a useable development platform.



This project now has no personal benefit for me, it would be nice to see it done but if it's not it doesn't affect me, I already have working firmware and aside from a little kindness to those who I feel deserve it, I have nothing to really offer the public open source project except advice and opinions, some of which may not be welcomed.

[KOC62]

websrvr:
Since motor and controller design is not going to be my day job, I can't become a master. But as a hobbyist, I can make something work to my satisfaction.
"Piecing it together" is part of the journey for some us, as that is how one can learn. It's not for everyone.
Overly protecting a loved from pain isn't always the best pathway for a person's developing years - but it can be good advice.

[Al_The_Man]

_

he uses 4 pole encoder on 2 pole motors, 6 pole encoders on 6 and 12 pole motors and 8 pole encoders on 4 and 8 pole motors and until I met the motor specialist I wasn't aware that an encoder had poles or that it even mattered in a motor but he showed me it does.
The motor specialist is a friend who deigns and builds motors for a living and I consider him to be GOD when it comes to motors and don't question his expertise anymore since he's proven everything he's every stated when I did question them, he also says in BLDC and AC motors, the encoder must be matched for the poles in the motor, I don't get how an encoder is affected by poles so I just accept it as the word of god and move on because I don't see the correlation.
_ID_ is close to determining the number of poles in a BDC motor except the rule is, count the number of magnetic poles in the rotor or stator, or, apply a small voltage to a single coil and count the number of pulled positions in 1 revolution.

If a BLDC motor has a non dividable by 3 slot count and induction type then without the design data it would be impossible to determine the number of poles visually and test equipment will be required if it is an induction type.

.

I think I posted earlier, the Method previously to detect the commutation point for BLDC motors was to use Hall effect sensors, these have been replaced by commutation tracks on the quadrature encoder and are totally separate, this method is much more accurate and very reliable.
The way I detect the No. of poles in a BLDC motor is to short the three stator connections and turn the rotor and count the number of 'bumps' per rev. this is the No. of poles.
It can also be done with a ''scope which if a double beam type, the encoder commutation tracks can be aligned per PDF. The sine wave is the rotor generation when the motor is back fed about 200rpm is sufficient.
There is no induction BLDC, an AC induction motor is a total different animal, the rotor fields use magnetic induction, hence induction motor.
BTW 2 pole BLDC motor are virtually useless for servo control, I would even hesitate to use one for a spindle etc, as I believe Sieg did?
Al.

[jimmkaz]

Mihai has and will help if he is treated with respect, regardless of what others are suggesting.

[_ID_]

I made a short video. You can see that servo moves in steps instead of smooth rotation. Also there is a dead spot visible. You can see the video here:



I tested the servo motor for the number of poles. I hooked 2 coils to 15V and turned the motor. I could feel 2 bumps per one revolution. I also made second suggested test by shortening all three windings together and turning the shaft with the hand. I felt no bumps, just some constant resistance to turning the shaft.

[websrvr]

Al_The_Man, you're wrong about DC induction motors, they do exist and a quick google search would have shown you several companies producing them.
DC_Induction_Motors_Can_Be_Found_On_This_Page


KOC62, yes I get it but this project isn't about piecing it together any more, Mihai already did this as his proof of concept, it now about making it work right so it's truly a desirable product.

You can probably cut some corners and drop the current sense IC's and use a resistor network but the encoder IC, it's advantages far outweigh going to a different solution.

I like to hear of developers using an STM32F429 to develop something that will be built and run on an STM32F103 and then later running into issues when the code fails to run correctly on the device.

Yes, STM32F103TBU6 dev boards are not very common but I have a source for them, they had manufactured a discontinued product just for developing this project and I have been offering it to anyone in the US and the board is $12.00 shipped if they want one so they can at least work on the intended platform.

Making a reflow oven dirt cheap that works well is a good solution for assembling SMD based boards and if you want/prefer a hands on experience, a hot-air pencil is a good option so fear of SMD should not be a contributing factor to decide on through-hole construction.

Through-hole products are a thing of the past and really only good in the initial stages of development before settling on a good final design, speaking from experience, this design is pretty sweet and it looks good as well.

The changes I made to my new layout are basically connector changes and just about the only through-hole devices aside from the fets on the board and still in the 100mm x 100mm board size.

The motor specialist suggested reworking the board for top and bottom components and SMD fets would reduce the size to 60mm x 80mm but I didn't want to make this effort.

Others who are developing their own firmware have made this decision based on the drama surrounding this project and those who are acting like they are in charge and I don't blame them but this doesn't help the public project.

All I can suggest is that someone needs to be responsible and start leading this project if it is desired or expected to survive and it better happen quickly.

Even Pommen has enough sense to stay away based on the current self appointed decision maker and I think he would gladly reconsider if proper leadership is present so what happens until then is a crap-shoot.

Decisions need to be made and someone needs to make them and they need to be made for the best interest of the project, a trait found only in a real project leader and a couple of people spring to mind.

I think _ID_ or KOC62 would make good leaders, they are inquisitive and research information, they have displayed behavior that would promote development and attract developers but the decision is theirs to ask everyone for permission to lead it and I doubt any sane person would refuse either of these candidates, I tried to promote development and incite assistance but no one wanted me making any progress as I was clearly informed I was not longer welcome in the project and best that I leave on my own steam essentially kicking me out.

If the project does move in a forward direction, I may consider contributing some code but under it's current regime, not much chance of that.

[KOC62]

Perhaps there is a "semantic" component to the "DC induction" motor?
Induction requires a changing magnetic field. To get a changing magnetic field requires a changing current.
Isn't a changing current similar to an AC current, or AC induction?

I'm not arguing... but also not understanding the "DC induction" term.

[websrvr]

KOC62, I used to think like you and Al_The_Man but I was told by GOD that the trapezoidal drive signal which has a changing amplitude will induce a magnetic field in a stator and transfer it into the rotor which in turn generates flux in the squirrel cage and this induced electro-magnetic energy (now voltage and current) creates a matching magnetic field in the rotor causing attraction much like it does in an AC induction motor and the shaft rotates, this works with three or more phases so the electrical energy can introduce a shift in the magnetic phase angle and also the way it is wound has something to do with it, I still don't really get it myself but I no longer question GOD about motors and how they work, I learned quickly he knows what he's talking about.

At the time this was confirmed by finding supporting information on the internet and substantiated by the fact that DC inductions motors are being sold so I accepted the reality and moved on.

[websrvr]

Someone sent me a PM who just had an epiphany and I think it's worth sharing, why can't both _ID_ and KOC62 lead the project together as a team, they would discuss directions, issues and options and make much better decisions as a cohesive leadership team than they could make alone and it would lessen the responsibilities and accountabilities of an individual person by shifting it to the collective team.

So guys, talk amongst yourselves and let everyone know but I think it's a great idea and sad I didn't think of it myself.

[websrvr]

_ID_, why does it look like your encoder has only 20ppr or is it your MPG giving too many pulses?

[Al_The_Man]

Al_The_Man, you're wrong about DC induction motors, they do exist and a quick google search would have shown you several companies producing them.
]DC_Induction_Motors_Can_Be_Found_On_This_Page

I

No 1 that link is Asian in origin and I have read all sorts of weird descriptions from these sites.
I would certainly be interested to learn of the technical details and where their definition of DC induction motor originates!
There is one motor that uses an induced voltage method to transfer external power to a rotating armature but it is not termed as a DC induction motor.
Al.

[Al_The_Man]

I looked out my old Motor text books from way back during my instruction years, There is one motor that was called a Synchro-pac and it used a rotor with a AC exciter winding which was coupled externally by induced magnetic (AC) energy with a full wave DC rectifier all mounted on the motor shaft to excite the rotor and turn the motor into an AC synchronous motor.
This is the only reference I could find that would fit this description.
But were basically AC motors in design and usage.
Al..

[websrvr]

Al_The_Man, I posted the first link from the search results and someone questions it's validity, no problem, they can do their own search and pick another site but I'm curious, is the information contained at wikipedia also considered false cause I wouldn't want to quote another false information site.

[KOC62]

What-if the "DC induction" motor is really a low voltage AC induction motor, say 24VAC.
We now design a DC-to-AC converter, using PWM techniques, to drive this AC induction motor.
Could we now market this as a "DC induction" motor?

If we add encoders, we could now make a "DC induction" servo motor. How cool is that?

[Al_The_Man]

Al_The_Man, I posted the first link from the search results and someone questions it's validity, no problem, they can do their own search and pick another site but I'm curious, is the information contained at wikipedia also considered false cause I wouldn't want to quote another false information site.

Having worked with motors for some decades, I am also curious in the construction of these motor, so in doing a comprehensive search of the internet in order to find this elusive motor, it has come to appear that the promoters of these motors are using semantics, the very description is misleading, DC & Induction in the same sentence.AC motors have used DC injection to enable synchronism for Many decades, but these are still AC motors, it is very difficult to create effective 'Induction' with DC!
I could find no information on any 'new' motor that I had not already come across?
Al.

[jimmkaz]

WOW James Newton is all ready backordered!

Slowing 69.16.243.61&c=1&t=42501.4766215278

https://www.youtube.com/watch?v=rnwYqGB3bXw

[KOC62]

I currently have 4 steppers, nema23, -- http://www.automationtechnologiesinc...kl23h286-20-8b and a Gecko G540 that is working just fine.
What would be the equivalent in a BLDC motor of nema23 size? It appears to me that any BLDC is going to cost a lot more with little return in my scenario. Or do you see a significant benefit?
Maybe I just need to add an encoder to my stepper and get good results, as a wannabe BLDC?

[jimmkaz]

KOC62,

There are literally thousands of people using the Gecko G540 with steppers very successfully. Why squander money if you don't need to?
By the time you build what Websrvr is promising you might be out of business. Right now it's just conjecture with no proof of concept.

I currently have 4 steppers, nema23, -- http://www.automationtechnologiesinc...kl23h286-20-8b and a Gecko G540 that is working just fine.
What would be the equivalent in a BLDC motor of nema23 size? It appears to me that any BLDC is going to cost a lot more with little return in my scenario. Or do you see a significant benefit?
Maybe I just need to add an encoder to my stepper and get good results, as a wannabe BLDC?

[jfong]

I currently have 4 steppers, nema23, -- http://www.automationtechnologiesinc...kl23h286-20-8b and a Gecko G540 that is working just fine.
What would be the equivalent in a BLDC motor of nema23 size? It appears to me that any BLDC is going to cost a lot more with little return in my scenario. Or do you see a significant benefit?
Maybe I just need to add an encoder to my stepper and get good results, as a wannabe BLDC?

It would be very difficult to choose a BLdc motor without knowing your machine specs. You need to know

All leadscrew pitch
Machine weight for each axis. If you have a gantry cnc, how heavy is it. This is used to calculate your required acceleration.
Do you want pulley reduction or direct drive. If pulley reduction, what ratio.
The required maximum feedrate.
How much cutting force is required for each axis. Cutting metal usually requires more than wood.

Probably a few things I forgot to list but there really isn't some easy way to say that this servo motor is equivalent to that stepper. A servo motor of equivalent size to a stepper will have lower torque but run at much higher rpm. This usually requires some kind of gearhead or pulley reduction to utilize a servo motor to its fullest potential.