Home-Made Simple IGBT based DC motor controller

classic Classic list List threaded Threaded
16 messages Options
Reply | Threaded
Open this post in threaded view
|

Home-Made Simple IGBT based DC motor controller

MPaulHolmes
I am getting 8 IGBT's (1200v, 600a, 4100watts each).  Does anyone see potential problems from replacing the power mosfet in the following simple 555 based PWM controller with the 8 IGBT's in parallel, which should be able to handle 32.8kWatts continuous?

http://www.dprg.org/tutorials/2005-11a/index.html

Thanks!!!  And if there are problems, which I'm sure there are, are there any suggestions for improvement?
Reply | Threaded
Open this post in threaded view
|

Re: Home-Made Simple IGBT based DC motor controller

Ian Hooper-3
Hi Paul,

The first problem I see is that a 555's internal transistor won't be  
able to supply enough surge current to switch large devices. You'd  
probably need to add an IGBT driver between the 555 and your IGBTs  
(I'm a fan of the IXDD414PI) plus gate resistors.

Suggestions for improvement.. Use a microprocessor? Seriously though,  
they're really very easy to use. Here's one I did a while back you're  
welcome to copy: http://zeva.com.au/speedy/

Ian Hooper
--
Zero Emission Vehicles Australia
http://www.zeva.com.au

On 29/03/2008, at 3:01 PM, paul holmes wrote:

>
> I am getting 8 IGBT's (1200v, 600a, 4100watts each).  Does anyone see
> potential problems from replacing the power mosfet in the following  
> simple
> 555 based PWM controller with the 8 IGBT's in parallel, which should  
> be able
> to handle 32.8kWatts continuous?
>
> http://www.dprg.org/tutorials/2005-11a/index.html
> http://www.dprg.org/tutorials/2005-11a/index.html
>
> Thanks!!!  And if there are problems, which I'm sure there are, are  
> there
> any suggestions for improvement?
> --
> View this message in context: http://www.nabble.com/Home-Made-Simple-IGBT-based-DC-motor-controller-tp16367161p16367161.html
> Sent from the Electric Vehicle Discussion List mailing list archive  
> at Nabble.com.
>
> _______________________________________________
> For subscription options, see
> http://lists.sjsu.edu/mailman/listinfo/ev
>

_______________________________________________
For subscription options, see
http://lists.sjsu.edu/mailman/listinfo/ev
Reply | Threaded
Open this post in threaded view
|

Re: Home-Made Simple IGBT based DC motor controller

Jack Murray
In reply to this post by MPaulHolmes
You need to do more research on motor controllers.
Look at the application notes for the IGBT modules.
Jack
--- paul holmes <[hidden email]> wrote:

>
> I am getting 8 IGBT's (1200v, 600a, 4100watts each).
>  Does anyone see
> potential problems from replacing the power mosfet
> in the following simple
> 555 based PWM controller with the 8 IGBT's in
> parallel, which should be able
> to handle 32.8kWatts continuous?
>
> http://www.dprg.org/tutorials/2005-11a/index.html
> http://www.dprg.org/tutorials/2005-11a/index.html 
>
> Thanks!!!  And if there are problems, which I'm sure
> there are, are there
> any suggestions for improvement?
> --
> View this message in context:
>
http://www.nabble.com/Home-Made-Simple-IGBT-based-DC-motor-controller-tp16367161p16367161.html
> Sent from the Electric Vehicle Discussion List
> mailing list archive at Nabble.com.
>
> _______________________________________________
> For subscription options, see
> http://lists.sjsu.edu/mailman/listinfo/ev
>

_______________________________________________
For subscription options, see
http://lists.sjsu.edu/mailman/listinfo/ev
Reply | Threaded
Open this post in threaded view
|

Re: Home-Made Simple IGBT based DC motor controller

Josh Wyatt-2
In reply to this post by MPaulHolmes
paul holmes wrote:

> I am getting 8 IGBT's (1200v, 600a, 4100watts each).  Does anyone see
> potential problems from replacing the power mosfet in the following simple
> 555 based PWM controller with the 8 IGBT's in parallel, which should be able
> to handle 32.8kWatts continuous?
>
> http://www.dprg.org/tutorials/2005-11a/index.html
> http://www.dprg.org/tutorials/2005-11a/index.html 
>
> Thanks!!!  And if there are problems, which I'm sure there are, are there
> any suggestions for improvement?

Irony is bizarre.  I just built this circuit for some series-wound regen experimentation.  I'm using this circuit to drive a single IRFP450 (14A, 500V) MOSFET controlling a small 1/2 HP DC motor I have (running at around 12khz).  I can tell you that I had to add a 10K resistor between the gate and pin 7 of the 555.

This is just for testing, mind you.  For a more production-ready PWM controller, you should consider starting with either a dedicated PWM controller (SG3524 perhaps), a micro as someone else suggested, or one of the quad OpAmp approaches out there (google for "lm324 pwm").  Reason being, the 555 cannot achieve the full 0-100% duty cycle range that is needed.

Lastly, you should consider MOSFETs instead of IGBTs.  They're cheaper, more readily available, and the tech is a bit more mature.  You may need more of them than IGBTs, though.  There's a pretty good "How do I choose?" whitepaper published by International Rectifier, google for "choose wisely whitepaper" to find the PDF.  Although, that article seems to be geared more toward single devices (versus paralleled, etc).  Basically, you will see that MOSFETs should be pretty good for applications below about 250V, and higher frequency switching, whereas IGBTs are preferred for lower frequency operation (less than 20kHz) and higher voltage.  There's some good discussion in that paper also about tail current and other factors that can potentially end up being a big efficiency/heat problem with any kind of bipolar device (BJT, IGBT, etc).

Someone else already posted this, but with any kind of parallel power handling, you will need to incorporate some kind of driver between the controller logic (555, micro, OpAmp) and the power devices, to achieve full turn-on and other signal cleanup tasks.

Thanks,
josh

PS: Cross-posting to evtech list due to subject matter.

_______________________________________________
For subscription options, see
http://lists.sjsu.edu/mailman/listinfo/ev
Reply | Threaded
Open this post in threaded view
|

Re: Home-Made Simple IGBT based DC motor controller

MPaulHolmes
That is bizarre!  Here's a couple more questions.  First, where is the EV tech list?  That sounds like a nice place to post this sort of stuff.  Second, Since it's an isolated gate device, and the leakage current from gate to emitter is about 0.5 microAmps (when the device is conducting from collector to emitter, as far as I can tell from the IGBT documentation), why can't the 555 deliver 4 microAmps (8*0.5)?  Can't it put out 200milliAmps?  I could see that signal cleanup would be important, though.  I hadn't thought about that before.

-Paul
Reply | Threaded
Open this post in threaded view
|

Re: Home-Made Simple IGBT based DC motor controller

Lee Hart
In reply to this post by MPaulHolmes
From: paul holmes
> I am getting 8 IGBT's (1200v, 600a, 4100watts each).

Ah, the fun is about to begin! :-) I suggest derating them about 3:1 i.e. if it's a 1200v 600a rated part, don't use it over 400v 200a nominal. This allows for voltage transients and real-world heatsinking. Those published ratings are absolute maximums that can't be achieved in real life.

> Does anyone see potential problems from replacing the power MOSFET in
> the following simple 555 based PWM controller with the 8 IGBT's in
> parallel, which should be able to handle 32.8kWatts continuous?

This is an excessively simple circuit. It has left out so many parts that it would barely function. Trying to use it for a high power controller would be a disaster.

> And if there are problems, which I'm sure there are, are there
> any suggestions for improvement?

 - Don't parallel IGBTs. It's hard to do right for a first-time builder.
 - Add a freewheel diode across the motor (can be one of your IGBT
   modules used as a diode).
 - Forget the 555. Use a switchmode power supply chip (of which there
   are hundreds). Pick one with current limiting, internal regulator,
   undervoltage protection, higher output drive current, etc.
 - Add a gate driver circuit. You need at least 10x more gate drive
   current than a 555 timer can supply.
 - Add a shunt to sense current, an amplifier, and use the resulting
   signal to control the PWM chip.
 - Sense IGBT temperature, and use it to limit or shut down the
   controller.
 - Add open or shorted speed control pot failure shutdown.
 - Add a robust power supply for the logic.
 - Add a large bank of filter capacitors across the battery input.

This should get you started. Good luck!


--
Those who say it cannot be done should not interrupt the one who is
doing it.    --    Chinese proverb
--
Lee A. Hart, 814 8th Ave N, Sartell MN 56377, leeahart-at-earthlink.net

_______________________________________________
For subscription options, see
http://lists.sjsu.edu/mailman/listinfo/ev
Reply | Threaded
Open this post in threaded view
|

Re: Home-Made Simple IGBT based DC motor controller

kenscircus
In reply to this post by MPaulHolmes
Hi Paul,
 
The 0.5 microAmps is the static DC leakage.  In operation the AC
characteristics will dictate the drive current requirements because the IGBT or MOSFET
gate is a capacitive load.  Additionally, the drive circuit must be able to sink
the miller currents generated by the gate to drain capacitance in a controlled
manner.
This question is directly answered in an International Rectifier application
note # AN-944.  It is a PDF file that can be downloaded here:
_http://www.irf.com/technical-info/appnotes/an-944.pdf_
(http://www.irf.com/technical-info/appnotes/an-944.pdf) .
 
Hope that helps!
 
Ken
 
 
 
In a message dated 3/29/2008 3:46:26 PM Central Daylight Time,
[hidden email] writes:

That is bizarre!  Here's a couple more questions.  First, where is the EV
tech list?  That sounds like a nice place to post this sort of stuff.
Second, Since it's an isolated gate device, and the leakage current from
gate to emitter is about 0.5 microAmps (when the device is conducting from
collector to emitter, as far as I can tell from the IGBT documentation), why
can't the 555 deliver 4 microAmps (8*0.5)?  Can't it put out 200milliAmps?
I could see that signal cleanup would be important, though.  I hadn't
thought about that before.

-Paul



**************Create a Home Theater Like the Pros. Watch the video on AOL
Home.      
(http://home.aol.com/diy/home-improvement-eric-stromer?video=15&ncid=aolhom00030000000001)

_______________________________________________
For subscription options, see
http://lists.sjsu.edu/mailman/listinfo/ev
Reply | Threaded
Open this post in threaded view
|

Re: Home-Made Simple IGBT based DC motor controller

MPaulHolmes
In reply to this post by Lee Hart
Lee Hart wrote
 - Don't parallel IGBTs. It's hard to do right for a first-time builder.
Uh Oh.  That's not good.  How on earth do I get the 40 peak HP that I'm wanting for my EV conversion?  (other than starting over with a bunch of MOSFETS)
Each IGBT has a power rating of only 4.1kW. My EV is already going to be really wimpy.  

That to do list will keep me busy for a while!!!  Thanks Lee!

-Paul
Reply | Threaded
Open this post in threaded view
|

Re: Home-Made Simple IGBT based DC motor controller

Peter VanDerWal
>
> Uh Oh.  That's not good.  How on earth do I get the 40 peak HP that I'm
> wanting for my EV conversion?  (other than starting over with a bunch of
> MOSFETS)

I'm confused.  Are you concerned about cost?
If so why on earth are you building you own controller, that almost always
ends up costing more (sometimes far more) than buying a commercially built
controller.
Even if you don't let the smoke out of a few sets of power silicon, buying
all of the necessary components to build a controller will cost as much as
and assembled one.



_______________________________________________
For subscription options, see
http://lists.sjsu.edu/mailman/listinfo/ev
Reply | Threaded
Open this post in threaded view
|

Re: Home-Made Simple IGBT based DC motor controller

Rush Dougherty
In reply to this post by MPaulHolmes
> First, where is the EV tech list?  

Evtech mailing list
[hidden email]
http://mailman.evtech.org/mailman/listinfo/evtech



Rush
Tucson, AZ
2000 Insight, 66.7lmpg, #4965
www.ironandwood.org
www.Airphibian.com
www.TEVA2.com



_______________________________________________
For subscription options, see
http://lists.sjsu.edu/mailman/listinfo/ev
Reply | Threaded
Open this post in threaded view
|

Re: Home-Made Simple IGBT based DC motor controller

Ryan Stotts
Here's a 50 volt, 110amp(draw), 5,500 watt controller to play with:

http://www.castlecreations.com/products/phoenix_hv_series.html

25v, 180a, 4500w

http://www.castlecreations.com/products/phoenix-180.html

Brushless though.  What happens when it's hooked up to a brushed motor?

Why can't it be scaled to 600 or 1200 volts and ludicrous amps?

_______________________________________________
For subscription options, see
http://lists.sjsu.edu/mailman/listinfo/ev
Reply | Threaded
Open this post in threaded view
|

Re: Home-Made Simple IGBT based DC motor controller

Steve West-6
In reply to this post by MPaulHolmes
> Each IGBT has a power rating of only 4.1kW. My EV is already going to be
> really wimpy.  

Perhaps you are confusing the power dissipation rating of the IGBT with how
much power it can *deliver*. The beauty of PWM is the the transistor spends
most of its time fully on (relatively low power dissipation) or fully off
(no power dissipation?). The transitions between these states are what you
want to keep to a minimum by sizing the drive circuitry appropriately. If
you can achieve 90% efficiency then in theory a single IGBT will deliver
almost 37kW (50HP) to your load.

Steve


_______________________________________________
For subscription options, see
http://lists.sjsu.edu/mailman/listinfo/ev
Reply | Threaded
Open this post in threaded view
|

Re: Home-Made Simple IGBT based DC motor controller

MPaulHolmes
In reply to this post by Peter VanDerWal

Peter VanDerWal wrote
I'm confused.  Are you concerned about cost?
If so why on earth are you building you own controller, that almost always
ends up costing more (sometimes far more) than buying a commercially built
controller.
According to another post on here, the 4.1kW is the amount of power the device can dissipate.  So at 90% efficiency, it can dump like 36 kW to the motor.  So, if I only have to use 1 IGBT, which cost me $12 (retail for over $400), then I'm doing OK costwise.  It's more elegant to have 1 component do the job of several weenie mosfets.  
Reply | Threaded
Open this post in threaded view
|

Re: Home-Made Simple IGBT based DC motor controller

Peter VanDerWal
>
>
>
> Peter VanDerWal wrote:
>>
>> I'm confused.  Are you concerned about cost?
>> If so why on earth are you building you own controller, that almost
>> always
>> ends up costing more (sometimes far more) than buying a commercially
>> built
>> controller.
>>
>
> According to another post on here, the 4.1kW is the amount of power the
> device can dissipate.  So at 90% efficiency, it can dump like 36 kW to the
> motor.  So, if I only have to use 1 IGBT, which cost me $12 (retail for
> over
> $400), then I'm doing OK costwise.  It's more elegant to have 1 component
> do the job of several weenie mosfets.

Gee why didn't I see that?  I don't know why all those other folks are
using such inelegant solutions to such a simple problem.  It's probably
all those engineers letting their college learning get in the way.

Go for it dude, let us know how it works out.



_______________________________________________
For subscription options, see
http://lists.sjsu.edu/mailman/listinfo/ev
Reply | Threaded
Open this post in threaded view
|

Re: Home-Made Simple IGBT based DC motor controller

Ian Hooper-3
In reply to this post by MPaulHolmes
On 30/03/2008, at 4:41 AM, paul holmes wrote:

> Second, Since it's an isolated gate device, and the leakage current  
> from
> gate to emitter is about 0.5 microAmps (when the device is  
> conducting from
> collector to emitter, as far as I can tell from the IGBT  
> documentation), why
> can't the 555 deliver 4 microAmps (8*0.5)?  Can't it put out  
> 200milliAmps?
> I could see that signal cleanup would be important, though.  I hadn't
> thought about that before.

I didn't see this one answered yet.. Basically the gate in IGBTs and  
MOSFETs is a little like a capacitor which you have to charge to turn  
the device on and discharge to turn it off.

So to switch these devices quickly (as is necessary in motor control)  
you have to be able to charge and discharge the gate very quickly - at  
a rate of several amps for larger devices, but only for a fraction of  
a millisecond. It's quite a different sort of load to what a 555 is  
designed to output, but exactly what those $2 driver chips were  
designed for..

On 30/03/2008, at 3:19 AM, Peter VanDerWal wrote:

>> I'm confused.  Are you concerned about cost?
> If so why on earth are you building you own controller, that almost  
> always
> ends up costing more (sometimes far more) than buying a commercially  
> built
> controller.
> Even if you don't let the smoke out of a few sets of power silicon,  
> buying
> all of the necessary components to build a controller will cost as  
> much as
> and assembled one.


I have to disagree with this one.. I've build several motor  
controllers now and it's always significantly cheaper in parts than  
buying a controller. OTOH if you count the value of your own time  
(learning and building) you definitely can't do it as cheaply as  
buying one! But learning and building is fun, so it doesn't count ;)

-Ian

_______________________________________________
For subscription options, see
http://lists.sjsu.edu/mailman/listinfo/ev
Reply | Threaded
Open this post in threaded view
|

Re: Home-Made Simple IGBT based DC motor controller

Dan Frederiksen
In reply to this post by MPaulHolmes
Paul Holmes wrote:
>I am getting 8 IGBT's (1200v, 600a, 4100watts each).

as some (but way too few) have somewhat pointed out, the 4100w is not its power but some peak power loss in it. it's actual peak power is 1200V and 600A which is 720kW. each :)  that is however some peak rating which doesn't apply to comfortable sloppy continuous power but still quite a bit more than you expected :)
let's say you want to run the system at little over 100V then one or two of those would suffice. by using more than 2 you can however avoid the need to have current limiting in the controller (which would otherwise fry the transistors because a stopped motor given full throttle from the batteries is essentially a short circuit so current will rise very rapidly if the batteries can provide)
magnetic devices are not as straight forward as resistive loads.

as others also have pointed out the 555 can't drive those IGBTs so for that you give the 555 signal to a socalled driver (or several in parallel) that are better able to switch the gates of the big IGBTs

some have said FETs are cheaper, but that's not true. FETs are weaker and more expensive and a bit better at very high frequency (which you don't really need). FETs can have very low power loss at low voltage if you put a lot of them in parallel because they act a bit like resistors (which gets lower in parallel) whereas IGBTs are more like diodes in that they have a fixed voltage drop (you feed it 100V only 98V get out almost irrespective of how many you put in parallel). but unless you want to make a rather expensive controller with a lot of FETs that runs very very cool it's not really an issue. (from what I understand)

I have been tinkering with a controller design (similar to Ian Hooper's but of course better :) which has a lot of the concerns (but maybe not all) taken care of
www.zev.dk/misc/diagram5.gif
it might help you. it's based on a microcontroller (programmed PWM) instead of your 555 but conceptually the same.
you might need a bigger gate resistor (R2) and/or you might need a bigger driver