Development of best servo motor for leather sewing machine

[GerryR]

7 hours ago, CowBoyOUTLAW said:

Our motor do not use buit-in gear box. No need of it

So, you have a high torque, slow start-up speed motor, which means you have a special controller for it (?).  What are the power requirements for it?  How big is the motor?  What are the output shaft dimensions?  What are the mounting requirements for the motor and controller?  Is the speed controlled by a foot pedal or a dial on the control box?  Do you have provisions for a position stop (needle up-dwn)?  These are just some questions I have for any new system I would consider.  It sounds like your system already exists, so should be easy to answer.   

[dikman]

Gerry, my understanding is that it's simply a Hightex 750w servo with 12 coils to give it lower startup speed and low-end torque, nothing more (other than a higher price because of the extra coils). I still believe that unless someone goes down the path you have to obtain effective slow-speed control and torque then a speed reducer is needed with a servo used for leatherwork (particularly with the big harness stitchers).

[GerryR]

27 minutes ago, dikman said:

Gerry, my understanding is that it's simply a Hightex 750w servo with 12 coils to give it lower startup speed and low-end torque, nothing more (other than a higher price because of the extra coils). I still believe that unless someone goes down the path you have to obtain effective slow-speed control and torque then a speed reducer is needed with a servo used for leatherwork (particularly with the big harness stitchers).

I know, I just wanted him to say it.  I still prefer VFDs with 3-ph gear motors over DC systems. JMO, YMMV.

Edited Tuesday at 11:18 PM by GerryR

[AlZilla]

This is usually such a friendly place. He's made no secret that he's the marketing guy, not the tech guy.

Would torque increase the same way speed reduces? As in 3:1 on speed equals 1:3 on torque?

I'm definitely skeptical that a motor in this size range can achieve the same increase as 750 Watts with a 1:3 speed reducer.

[dikman]

Its ONLY advantage over other servos is the claimed startup speed of 50 rpm. Just how practical that is in the real world of leather sewing I don't know. All I do know is that with my limited abilities I need to be able to sew slow and know that there is enough torque to punch through heavy leather - which means a speed reducer.

[AlZilla]

10 minutes ago, dikman said:

Its ONLY advantage over other servos is the claimed startup speed of 50 rpm. Just how practical that is in the real world of leather sewing I don't know. All I do know is that with my limited abilities I need to be able to sew slow and know that there is enough torque to punch through heavy leather - which means a speed reducer.

I've got a machine with a servo that does 100RPM startup and a speed reducer. Today I counted 14 stitches in 30 seconds.  50RPM would give me 7!

But, I'm slow enough.

[GerryR]

Torque from a motor depends upon current in the windings and the number of poles (basically, I'm not a motor expert) so without a speed reducer of sorts, the motor has to be very robust to maintain high torque at slow rpm and cooling is paramount.  The speed reducer gets the motor rpms up where cooling is better, current is lower and torque is higher because of the gearbox.

As far as the OP is concerned, I don't understand why he can't just tell us his new motor/system specs; why all the mystery?

[dikman]

http://hightex-solution.com/blog/technology/2024/0425/315.html?1714026959

This is the Hightex site with info on the 750SH model, which I assume is what we're discussing here. There is a video at the end showing it in operation. It does appear to have reasonable torque at slow speed, but no better than running a cheap servo and speed reducer. For a factory maybe it's an advantage not to have to fit a speed reducer so paying a higher price for a servo may make sense? I said earlier though that for the hobbyist is it cost effective? Not for me, particularly with more than one servo, if a cheapy servo dies I'll simply replace it with another cheapy servo (and save $300+).

Interestingly, the link shows a 9-coil motor. Gerry does raise a point about heat generation when running a motor very slow.

[kgg]

My experience with a cheap digital brushless servo motor that came with my TSC 441, 16" cylinder arm clone has a speed range of 500 - 5000 rpm on it's 800 watt, 9 coil and with no reducer pulley on the machine:

Calculated:

1. Top speed of the hand-wheel pulley should come in at 156 rpm using:

i) a 160 mm hand-wheel pulley size,

ii) a 50 mm motor pulley

iii) with the motor speed set to the minimum speed of 500 rpm's. This particular servo motor is a 800 watt brushless 9 coil servo motor with a speed range of 500 to 5000 rpm's.

Measured:

i) The hand-wheel pulley measured out to be 159mm.

ii) The small pulley at the motor measured out at 49mm.

Measured results:

i) with 500 rpm's showing on the controller screen

ii) with the foot pedal fully depressed the rpm's with my digital tachometer showed 154.5 rpm's with no material under the needle

iii) with the foot pedal feathered to get the minimum hand-wheel pulley speed the digital tachometer showed a low rpm reading of 61 rpm's with no material under the needle.

From that I can assume a no load ramp up speed of 61 rpm's.

I'm assuming the 50 rpm speed they are quoting is done in a similar manner. 

6 hours ago, dikman said:

Interestingly, the link shows a 9-coil motor. Gerry does raise a point about heat generation when running a motor very slow.

With everything being equal a 12 coil should get rid of the heat better then a 9 coil operating under the same load conditions.

kgg

 

[GerryR]

6 minutes ago, kgg said:

 

With everything being equal a 12 coil should get rid of the heat better then a 9 coil operating under the same load conditions.

kgg

 

Power is power, so if it requires a certain amount of power to pierce the leather, both motors, whether 9 coil or 12 coil, will generate the same amount of heat.  It all depends upon the heat management of the motor design, so "should get rid of the heat better" is up for grabs until the longevity of the motor is proven.  12 coils will distribute the heat better, but it is still within the confines of the motor housing.

[kgg]

54 minutes ago, GerryR said:

12 coils will distribute the heat better, but it is still within the confines of the motor housing.

Agree.

One of the members on the forum has a brushless servo motor that has a fin system moulded right into the small pulley. I am assuming that was done to try and help keep the housing cooler. Lets face it brushless servo motors are designed to last too as long as the warranty period and utilize a design to save on manufacturing costs.

I doubt any brushless servo motor will ever come close to lasting as long as the big old heavy clutch motors that weight about 40 lbs. with real metal housings verses about 18 lbs of plastic and aluminum. Sort of like a Zippo lighter lasting a life time verses a BIC lasting a week or two. 

kgg

[Cumberland Highpower]

1 hour ago, GerryR said:

Power is power, so if it requires a certain amount of power to pierce the leather, both motors, whether 9 coil or 12 coil, will generate the same amount of heat.  It all depends upon the heat management of the motor design, so "should get rid of the heat better" is up for grabs until the longevity of the motor is proven.  12 coils will distribute the heat better, but it is still within the confines of the motor housing.

Actually there are a few other bits at play, such as the elemental composition of the iron used in the stators.  Chinese motors often use some really low quality sheet steel for that.   You'll get oversaturation and more heat as a result.

I'd gamble an Efka motor generates a good bit less heat to dissipate based on the Watt rating than any of these bargain Amazon servos.

Seems like I remember a guy on ebay selling some 4:1 speed reduction bolt on (planetary?) gearboxes for servos.  I don't remember too many details other than they were somewhere around $100.  Anyone here have one or try one out?

Edited Wednesday at 01:11 PM by Cumberland Highpower

[GerryR]

15 minutes ago, Cumberland Highpower said:

Actually there are a few other bits at play, such as the elemental composition of the iron used in the stators.  Chinese motors often use some really low quality sheet steel for that.   You'll get oversaturation and more heat as a result.

 

Yes, I stated that both motors would generate the same amount of heat and meant power; that is where the heat management comes into play, and like you stated, the choice of materials used.

[Wizcrafts]

On 7/5/2025 at 10:06 PM, CowBoyOUTLAW said:

WOW, are you servo engineer?
Few people can talk about optical controls deeply. Now we use 600 line encoder, the minimum speed on CB4500 is about 10 stitch per minute. If use 1200 line encoder, in theory the speed can down to 5 SPM.

I'm not an engineer; just a knowledgeable user. Optical speed control is not new technology. I've had volume pedals for my pedal steel guitars from decades ago that used a small light bulb or LED to emit a focused beam against a photo-resistor. The sensor/resistor controlled the signal transfer between the input and output jacks. When backed off, heel down, a small phenolic board blocks the light beam, so there is no output at all, yielding full resistance. As one moves the pedal with his toe down, the phenolic board is brought out of the line of light, gradually increasing the amount hitting the sensor, decreasing its resistance. This results in an extremely smooth transition from zero to the maximum volume that the volume pedal is able to transfer to the amplifier (~95%). My current electronic volume pedal is still optically controlled, but uses a variable density filter instead of a solid piece of board. It has a small amount of amplification so that it doesn't lose any signal at full volume.

Fast forward out of the world of pedal steel guitars and into the world of digital servo motors on sewing machines. In 2011 through 2013, our main 441 machine suppliers sold them with a digital motor that had a 45mm pulley. Some started with a jolt at 200 rpm. Others started with a similar jerk at 100 rpm. So. somebody took their motor apart and saw that it was optically controlled, just like my old volume pedals. That person had a good printer and some transparent "paper." He went and printed out some strips that had a varying density of gray ink, smoothly ranging from 100% black to 100% transparent. By playing with the spread of the printed filter, he came up with a simple optical filter that anybody could install inside their motor. He sold them through the forum for a modest fee. I installed one in my Cowboy cb4500, in my buddy's Cobra Class 4, and on a Techsew I acquired a little later. This modification worked 100% as designed. The only thing that caused a variation was if the machine hit a bump in the floor when I moved it to sew a log project (webbing strapping or a long rifle case). It was a simple matter to reposition the filter, which I later glued in place.

So, is this new KineDyne motor optically controlled? If so, have you considered using an optical variable density light filter to control the speed? I think I still might have a spare filter in an envelope. Bob Kovar might still have a couple left. We could probably get one to you to experiment with. I'm pretty sure there is a variable density filter in my original digital servo from my cb4500, which is on the floor in a closet. If you try the variable density filter control system and perfect it, then secure it to survive shipping and man-handling, you will be able to offer a motor that smoothly transitions from absolute zero through the top speed as set on the control panel, with good low speed torque from the numerous coils.

A side benefit of having a digital servo motor running this slowly is that there seemed to be much less heat being generated. Originally, we felt the heat as we sewed at 5 to 9 stitches per second. It was even running hot at slower speeds due to it starting at 200 rpm. I also disliked the non-linear speed change which ramped up after I pressed down on the pedal. It wasn't an instantaneous change that I preferred. I grew up using clutch motors on sewing machines, and still do on a few of them.

I have since moved away from touchy digital servos to the analog motors sold by my Cowboy dealer. They contain brushes and replaceable carbon signal transfer blocks on springs. They reliably start at zero and speed up linearly as I press my toe down; just like my volume pedal on my steel guitar.

[GerryR]

Those optical filters are nice, but they defeat the purpose of starting the motor at the so-called required minimum speed, which is probably not necessary in most instances.

Edited Wednesday at 04:59 PM by GerryR

[AlZilla]

1 minute ago, GerryR said:

Those optical filters are nice, but they defeat the purpose fo starting the motor at the so-called required minimum speed, which is probably not necessary in most instances.

If you wanted a certain minimum speed, couldn't you just start the gradation at that point? Then have a smoother ride up to max?

[Wizcrafts]

5 minutes ago, GerryR said:

Those optical filters are nice, but they defeat the purpose of starting the motor at the so-called required minimum speed, which is probably not necessary in most instances.

I didn't find that. I sewed with the optically modified motor on my cb4500 and a couple of walking foot machines.  They worked very well. The Cowboy had a 3:1 reducer. My long arm Singer machine fed directly from the motor's 45mm pulley. They had no hesitance to start turning over.

[GerryR]

31 minutes ago, AlZilla said:

If you wanted a certain minimum speed, couldn't you just start the gradation at that point? Then have a smoother ride up to max?

Yes, but that would defeat the purpose of a softer start, which you are trying to achieve.

[Cumberland Highpower]

4 hours ago, GerryR said:

Yes, I stated that both motors would generate the same amount of heat and meant power; that is where the heat management comes into play, and like you stated, the choice of materials used.

It's easy to mistype these days, especially with all the autocorrects, I do it all the time! 

A person could have an oversaturated motor w/stators made from Chinese bean can steel that draws 600W and a quality built Efka or Misubishi that draws 575W and actually have "more" output shaft power put out by the later.  

When it's a $200 servo vs one a few times more expensive, nobody really seems to care though.   I bet all the ROC made clutch motors I have aren't as efficient as my old Belgian and USA made versions, but I don't seem to care either.

I'm just rambling on GerryR.

Edited Wednesday at 05:59 PM by Cumberland Highpower

[GerryR]

Keep rambling, good info!!

[sbrownn]

Best solution is to buy a commonly available servo and double up the speed reducers if you need to.

[dikman]

2 hours ago, sbrownn said:

Best solution is to buy a commonly available servo and double up the speed reducers if you need to.

I did that once - a snail was fast in comparison!  An interesting experiment but not very practical.

Bottom line here is the Kinedyne (Hightex) motor is doubtless a good motor but if I had one, which isn't likely given the cost, I would still fit a speed reducer. And I don't need all the other gimmicks fitted to it.

Edited Wednesday at 10:58 PM by dikman

[AlZilla]

6 minutes ago, dikman said:

I did that once - a snail was fast in comparison!  An interesting experiment but not very practical.

Of course, I had to run the numbers. A 30MM motor pulley, (2) 2:6 reducers and a 3.5" hand wheel gives a 1:31.5 reduction. 100 RPM becomes 3.2 at the hand wheel. 20 seconds per stitch!

[Wizcrafts]

Despite all of the improvements made in electronic motors over the last almost 20 years, I can still feather a clutch motor so it barely turns over the machine, and hold it at a slow and steady speed. Yes, they are noisy and generate a bit of heat. But, they are like the Ever Ready Bunny: they just keep on going!

If anybody reads this and is having trouble controlling a clutch motor, try backing off the big screw on the clutch housing. This adds a certain amount of slack movement before the disks engage. You can also tighten the coil spring that holds back the clutch lever. That keeps the lever up and away from engaging accidentally. Finally, you can smear a little grease on the internal cork disk that mashes into the main disks. The final adjustment is to replace a large motor pulley with a 2 inch pulley, or to add a 2:1 speed reducer.

[Wizcrafts]

1 minute ago, AlZilla said:

Of course, I had to run the numbers. A 30MM motor pulley, (2) 2:6 reducers and a 3.5" hand wheel gives a 1:31.5 reduction. 100 RPM becomes 3.2 at the hand wheel. 20 seconds per stitch!

I try to keep my minimum sewing speed to about 1 stitch per second, except on intricate turns and patches. I slow down to get the best needle placement, then speed up on straight-aways.