Designing an Integrated AC Servo Motor (Drive, Encoder and Motor in a single package)

Dear Friends,

I am in the final stage of designing an Integrated AC Servo Motor (Motor, Drive and Encoder in a single package) with the following features. Trying to get your opinion on the demand and features expected.


1. Low cost. NEMA 23 around 250$ to 300$ NEMA 34 around 300$ to 400$
2. Highly energy efferent Vector Control or Field Oriented Control based motor driving algorithm at the core. Uses 33 KHz PWM switching of MOSFET bridges
3. Very high resolution possible 32000 steps/rev is easily achievable. 10000 or 20000 steps/rev by default.
4. Smooth vibration and noiseless operation due to Sinusoidal PWM output driving stage
5. Operates in torque, speed and position control modes
6. Very high torque density in much smaller motors than general 8 pole PMSM servo motors at low speed due to the use of a 100 pole count PMSM motor (i.e. a stepper motor)
7. Significantly higher toque in high speed range compared to conventional and micro-stepping drives due to closed loop operation
8. Due to high torque, the motor can be directly connected to ball screws or machine shaft without any speed reduction/torque expansion mechanism like gears, timing belts and pulleys etc. which with reduce product/machine development cost and increase accuracy and precision of the drive system
9. Due to high torque, high resolution and direct driving capability, simpler and cost effective belt driven linear motion components like this can be used directly in high precision applications. Linear power driving without the ball screws reduce the product cost further more. Direct belt drives are generally not suitable for stepper motors with conventional drive because of their low resolution and ball screws are not suitable for long linear motion.
10. PC based controller parameter programming software using USB port
11. Two encoder ports (one for future implementation as dual-encoder drive)
12. CAN port for making the drives work in/through a network (for future implementation)
13. Operates from a 24V or 48V switching power supply depending on the motor model

Your comments and suggestions will be highly appreciated as I am going to kickstarter after several months.

Cheers!!

marifhossain

Sounds like an interesting project, great job as you have put alot into this

I bought existing servos of nema 23 size at 290€ per axis, including 22% VAT taxes, with 5000 count encoders.
Price includes cabling, servos, and drives.
Your target price needs to be about half, or 170€ without tax, to be succesful, in qty 10-20.

What is the actual benefit, and have you demonstrated working precision at 20-30.000 steps / turn ?

Quote:

---

Originally Posted by hanermo

I bought existing servos of nema 23 size at 290€ per axis, including 22% VAT taxes, with 5000 count encoders.
Price includes cabling, servos, and drives.
Your target price needs to be about half, or 170€ without tax, to be succesful, in qty 10-20.

What is the actual benefit, and have you demonstrated working precision at 20-30.000 steps / turn ?

---

Thanks for your inputs!

Yes, the working precision is 20,000 steps / turn or 10,000 steps / turn, we have two models. Other advantage is significantly high torque than that of a traditional 8 pole PMSM servo motor in low speed range (below 500 rpm, at least 4-5 times more torque), so you'll need much smaller motors compared to traditional servos. Could you please share the maximum torque of the NEMA 23 model you bought?

- Price/resolution and price/unit-torque would be the most significant benefit. The price above is not fixed and we are actually collecting data to understand the demanded price-range. We are also looking towards understanding the need for features that are demanded.
- Compact size, all-in-one for space saving
- Due to high torque and high resolution direct drive is possible in many application as it may eliminate the need of gearbox, belt-pulley etc. mechanical drive components to reduce the system cost further
- We'll include some application specific features if needed (customization) for bulk order quantities.

Price depends largely on who you are targeting imo. At the prices you have listed that is on par with ClearPath and Leadsine Easy Servo's; only you have greater than an order of magnitude better resolution (One of the main drawbacks to ClearPath is the poor resolution at 800 counts per rev). So I think it's actually an excellent value if it is as capable as you say.

Also I would stress making it easy to integrate. While some people here have professional level capabilities, most of us are advanced hobbyists at best. One reason I went with steppers vs 220v AC servos on my build was lack of build logs utiilizing AC servos. While stepper build logs were prolific to say the least. That gave me the confidence to take the leap. Knowing what I know now I wouldn't hesitate to tackle and AC servo build, but the point is you want to make it as close to plug and play as possible when targeting the hobby/prosumer market; which you obviously are if you are on Kickstarter. This includes user friendly documentation and support.

Lastly, if these are low volt servos I would at least include some (affiliate?) links to capable compatible power supplies. Again, making it as easy as possible for someone to integrate these into their build. Bare minimum that will cut down on the 'will this power supply work?' emails.

eta: What is the rated rpm of the motors? Do you have a torque speed graph available? If you don't you should imo. I can't stand when there isn't sufficient performance documentation.

You are completely correct.
If you actually deliver stuff with 20.000 steps/turn resolution, and the typical 1-2 step positioning, it WILL eliminate the need for gearing.
It will also make you option very valuable.

The torque on mine, iirc, without looking it up, was about 3.9 Nm peak.
Dont really recall, atm.

If you actually significantly increase the resolution, and the torque, you will have a huge success.
I would be happy to be an EU distributor (not dealer).

Speaking, again, from a commercial pov.
Timeframe ?

Quote:

---

Originally Posted by marifhossain

Thanks for your inputs!

Yes, the working precision is 20,000 steps / turn or 10,000 steps / turn, we have two models. Other advantage is significantly high torque than that of a traditional 8 pole PMSM servo motor in low speed range (below 500 rpm, at least 4-5 times more torque), so you'll need much smaller motors compared to traditional servos. Could you please share the maximum torque of the NEMA 23 model you bought?

- Price/resolution and price/unit-torque would be the most significant benefit. The price above is not fixed and we are actually collecting data to understand the demanded price-range. We are also looking towards understanding the need for features that are demanded.
- Compact size, all-in-one for space saving
- Due to high torque and high resolution direct drive is possible in many application as it may eliminate the need of gearbox, belt-pulley etc. mechanical drive components to reduce the system cost further
- We'll include some application specific features if needed (customization) for bulk order quantities.

---

Eddard_Stark,

Thanks for your inputs. Although professionally I am a firmware engineer with control engineering background, I am an electronic hobbyist from my childhood. This servo will be priced and simplified considering the advanced hobbyists for sure.

Any power supply is fine i.e. linear or switching, some suggestions of the ratings and models will be made later. We'll also prepare simpler documentation for commissioning and tuning of the servo drives. The motor firmware will be up-loadable through USB for bug fixes and feature extensions. Some preliminary torque testing have been done. Ordered the dynamometer for formal testing which will happen in next month. I'll keep you guys updated.

hanermo,

It's nice to know that you are interested in marketing of my products. Production is expected to start at least six months from now. I'll keep you guys updated on this.