Industrial motor drive power and control circuit supplies? 1

[abfer]

I'm trying to design a 3 phase ac ind. motor driver. I planned to obtain dc bus voltage from 380VAc to feed the motor via power semiconductors and obtain the power needed for control circuits from 220VAC via a dc-dc converter. For this purpose, 3 phase and 1 neutral line have to be connected to the drive. But i have some hesitations and questions since i have not got much experience with the commercial drives:
1- What strategy commercial drives usually follow? Seperate input for control circuits? e.g. 24VDC. or obtain from dc bus voltage which is high (500VDC or more) and makes difficult to obtain dc voltage for control circuits than obtaining from 220VAC.
2- Should i care anything while obtaining DC Bus voltage from 380VAC. When we rectify it we get max ~537V. But capacitors aren't more than 450V. Must i use serial connected capacitors? If so,then capacitance will decrease, will it suffice?

Also at a 10KHZ carrier frequency with igbts have reverse connected diodes, should i use snubber circuits on power devices with the motors lower than 15HP.

 

[Skogsgurra]

You are about to do what Columbus did when he started out on a voyage without knowing much about the itinerary and the goal. The kind of questions that you ask indicate that you are far from ready for such a voyage. But it can be a rewarding experience - if you have financial backing and the endurance to work through it all.

But you should know that competition is fierce and that very competent drives can be bought for less than what you would pay for the components. These drives are the result of many years of experience, optimization and "productification".

And, yes. Do, by all means, derive auxiliary power from the DC link. You will not be able to brake to stillstand on power outage if you connect the power supply to AC. For obvious reasons.

Good luck, anyhow.

 

[abfer]

Yes you are right. I can't compete to huge industrial companies but my first goal is to learn and design a good enough driver. It won't be a perfect drive at the beginning from the perspective of functions, cost etc. May be then i further develop it and add some application based specialized functions which makes it attractive for local usages. Anyway my first purpose is learning but i also target it to be as close as possible to a commercial drive. So if we obtain the auxiliary power from the dc bus then there is no need to neutral line but ground is needed, i think, right?

 

[jraef]

10 years ago, I would venture to say that roughly 1/2 of the 100+ VFD designs that I encountered derived auxiliary power from the AC line and the other half derived it from the DC bus. Now I would be hard pressed to find one that still had an AC source. The globalization of product sales weeded out those who were not flexible enough to be powered up anywhere in the world. Using the DC bus as a source ensures only having one input voltage point to be concerned about.

That said, I agree with Skogsgurra as to making sure you really want to embark on this journey. The largest corporate giants in the world have invested mightliy in brining costs down to levels that allow no reasonable margin for new innovators, unless you bring something that has NEVER been done before and can convince a large enough market that they can't live without it. In that endeavor, using an Engineering Tips group like this to solve your problems is probably not going to serve you well, and indicates that you may be biting off more than you can chew.

I'm not trying to be negative, I'm just offering unsolicited advice based upon 1 paragraph I have seen you write, so take that for what it is worth and good luck.

"Venditori de oleum-vipera non vigere excordis populi"

 

[abfer]

Ok i didn't emphasize on commercial or profit based something. I'll decide that part later. It's not certain for now, i only mentioned about some possibilities. Thanks for the comments and forget about the commercial side of this please

 

[Skogsgurra]

As a learning experience, I can recommend building your own VFD. You will learn about drive systems, motors, fast transients, power electronics, cooling, EMI,DSP, lots of frequency domain thinking and quite a lot of analog techniques. And safety. And HMI. If you add communication - which is a must today - you will also learn about that. And more.

So I could recommend such an activity to anyone that feels that she/he needs a deep understanding of one of the more complex and varied electronic systems that exist today. But, there is an even better way: attend classes in basic electricity and electronics. That will eventually put you in a position where you will be able to resolve issues like the one with capacitor voltage and capacity (in your OP) with great confidence.

Learning by trial and error is a painful proposition when it comes to power electronics - and dangerous as well. Exploding electroltyitic capacitors have killed one guy that I once knew. An I can thank my eye-glasses for not being blind when an IGBT module exloded right into my face. Epoxy shrapnels are not to laugh at.

So, I think that you should consider the (class) alternative. It will give you a good education and also proof that you have the skills needed. Then get a job with a good company and work your way up. There are very few short-cuts. And they seldom work.

Gunnar Englund

 

[sreid]

Most if not all drives derive internal power from the DC bus using a small multiple output switrching power supply. Besides 24 VDC, 15 VDC for gate drives and 5V or 3.3V to power digital electronics (and often others).

The drive will also have to withstand higher voltages (say 1000 VDC due to regeneration when inertial loads are decelerated. 450VDC capacitors are used in series with parallel balancing resistors. 450 V capacitors have higher surge ratings (550 VDC?).

 

[sreid]

To save yourself some grief, take a look at Powerex Intellimod power modules. Integrated gate drivers and short circuit protection.

 

[ScottyUK]

In addition to sreid's link, have a look at International Rectifier's and Semikron's websites. Both sites contain a large amount of application data.


----------------------------------

If we learn from our mistakes,
I'm getting a great education!

 

[itsmoked]

Do heed skogs and others cautions. You will be working in a very lethal environment.

As my professor was fond of saying, "In logic you can make
all the mistakes you want, in power you can't make any".

This site has a huge amount of data on "how to" AC drives.

http://www.microchip.com/stellent/idcplg?IdcService=SS_GET_PAGE&nodeId=1490&filterID=497

 

[mc5w]

Also, designing a RELIABLE rectifier that runs directly off of 415 volts or 480 volts is very tricky. A directly connected rectifier tends to convert voltage spikes into useable power hence you must have these things:

1. Diodes with a peak reverse voltage of at least 2,000 volts.

2. A switched or unswitched load ( bleeder resistor or dynamic braking resistor ) that will get rid of excess voltage even when the motor is not running. This is how electronic fluorescent ballasts burn out when the bulbs fail or are removed.

3. Resistance-capacitance snubbers across each diode. If you do not have these then dV/dT when a diode turns off can cause the diode to fail to turn off.

Please see

http://home

dot earthlink dot net/~mc5w/badnewsballasts.txt for more details.

4. Choke input power supply otherwise you will have more power quality problems and very high peak current in the diodes.

5. A 3-phase inductance or Universal Harmonic Filter in the 3-phase supply circuit. Mirus International makes a Unversal Harmonic Filter for VFDs including 1 that does single phase to 3-phase conversion. Both devices produce current waveform distortion equivalent to an 18-pulse rectifier.

Actually, you can learn just as much by taking apart old drives, especially the ones that use a combination of analog and digital circuits such as Emerson's Prism drives.

During the summer of 1994 I had to adjust/recalibrate 5 Prism drives to get 5 steel coil straighteners to work right. Once I got them to give me 60 Hertz at 6 volts DC signal input I programmed the sonar control to produce a minimum signal of 1.5 volts when calling for the motor to run so that the motor would not stall out. The motors were dependent on a shaft driven fan for cooling. Essentially, the sonar senses the depth of the steel coil loop between the straightener and the punch press. The Coiltek sonars from the Carey Corporation were programmed to operate in start-stop mode when steel demand was low and operated the VFDs as 15 Hertz soft start at low demand. At higher demands the sonar could operate the motor at up to 90 Hertz.

One problem was that somebody else had tinkered with the Prism drives thinking that they could improve performance by maladjusting the things. Prism drives are a bit of a pain to adjust and calibrate to work a dertain way but you will learn more because the only "black box" is the PROM that is used to create a sinusoidal reference waveform using a digital to analog converter.

I have also had to adjust and recalibrate an emerson Spectrum drive which is a 4 quadrant 3-phase in DC out drive. It can take a day and a half to zero offset the differential amplifiers. This is because the 10 turn pots only have 1/4 of a useful turn because they are connected +15 volts to -15 volts instead of + and - 1.5 volts. If you have time you could unsolder the 10 turn pots and install some fixed resistors to the ends to make them more useful. However, the Spectrum driveis very easy to understand because there are no black boxes.

 

[itsmoked]

mc5w I thought they used beta sources to check the steel thickness...

 

[Skogsgurra]

Yes, they do. Or even gamma. For thickness. But this, I think, is about checking the hang between coil and press.

How did we get into mechanical problems? It started electric...

How silly - that's the way it normally goes.

 

[itsmoked]

Roger that!

 

[abfer]

Thanks for your valuable comments.

"As my professor was fond of saying, "In logic you can make
all the mistakes you want, in power you can't make any"."

What a coincidence. I also thought about this several days ago. Because first in school i direct my way to logic based things which has no danger. I changed my way to power electronics later. Now i can't be sure that i am doing everything right and checking again. Because it's not funny to make a capacitor explode or a power switch get burnt also may be dangerous. So i check my connections at every step.
About the classes; I already attended classes. But classes don't give every thing in practical especially basic ones. Even it gives you the knowledge you may not perfectly understand it and may not remember when you need it. To really understand a subject you have to need that subject and work on it. For example first i thougt that i can use a 600V capacitor. Than i couldn't find it. I thought that i can connect two in serial. Then i couldn't be sure that if it has drawbacks. I know that electyrolitic capacitors have big tolerances but i didn't first think that loading may not be equal even i read several subjects similar to this before but i forgot at that moment. This is why i'm asking here; doing as possible as low mistake. You can't learn these details even the instructor tolds you about these subjects. Because he/she tolds a lot of thing, a lot of details, a lot of experiences. Simulations also don't give you this information. Because there is no voltage limit at capacitors or unbalance etc. You have to experience it by yourself. So real world application is very important in engineering, i think. One good way, as you've said is, working in a good company but i have no opportunity at this time.

The other subject. Dc-dc converter. I first did that. It's working from the utility grid which is 230VAC. I made two mistakes there. First is not a real mistake but i see that it has no application nowadays. I must design a dc-dc converter which is fed from a high voltage dc bus. The second it must have isolated outputs for gate drivers and control circuits. I forgot to seperate it. Anyway it was my first dc-dc converter. The next time i know my requirements better and i have more experience than before. So i can do a better one. I didn't lost anything. Also it has commercial side but i didn't use that side. Because i must really make a good design, and long enough test etc. But now, i'm only learning and experienced people like the ones here, in design or applications not important, can give me very valuable knowledge. So thanks to all again.

 

[itsmoked]

abfer; I went down the Power Path at school but have ended up doing embedded systems (computer designs), so I can appreciate your situation.

Do check my link above. You will need a micro to do all this, and a fast one at that. Microchip has micros designed for this application. Also they have very inexpensive development tools. You could probably get everything for a few hundred dollars, (maybe less). Since you have knowlege in logic this should work well with your education.

I salute your attitude. Just be careful.

 

[abfer]

Thank you itsmoked. I'm familiar with microchip products. I made a low voltage dc motor drive with one of PICs. It has application based specific properties and used commercially. I combined a dc motor drive and application based automation options. You can't do this job with a commercially available drive. Because it has generic properties and you need a seperate control system which will say the drive what to do. So the system cost would be high. So my drive may not have advaced properties as a professional drive. But it has a lot of advantages over it when we think the spesific application. Also it's working good and it is reliable and system cost is much more cheaper than a commercial one. If i had chance, i'd made it more advanced. This may be an answer for commercial doubts of peoples here. So i have some little commercial experience also The world is not small as we think may be and there is space for everyone small or big

 

[Skogsgurra]

You impress me, abfer!

I think that I misjudged you. I couldn't have said the "To really understand a subject you have to need that subject and work on it" part better. Keep on to your vision. It is worth a lot.

 

[abfer]

Thanks skogsgurra, your comments and knowledge will guide me as the other friends'.