Page 4 of 5

Open Source DC controller

Posted: Thu, 18 Mar 2010, 18:35
by mcudogs
Power Board
Controller

Differences between ver c & d are

Power Board
- Etched not milled

Controller Board
- Extra input for rev sensor
- Spare analog / opto isolated input
- Extra digital output for driving a relay or lamp
- pack voltage sensing
- connector for MOSFET gates instead of direct connection via resistors

Cheers

Open Source DC controller

Posted: Thu, 18 Mar 2010, 19:31
by BiGH
THANKYOU!! :)

Open Source DC controller

Posted: Fri, 19 Mar 2010, 00:13
by BiGH
do we have any photos of completed Rev 2d? boards?

Open Source DC controller

Posted: Fri, 19 Mar 2010, 01:26
by mcudogs
Link to assembly instructions

Hope this helps.

Don

Open Source DC controller

Posted: Fri, 19 Mar 2010, 14:50
by Taffy
That is looking good Don. Be interested to hear what its like when complete and running in the car.

Open Source DC controller

Posted: Fri, 19 Mar 2010, 15:17
by BiGH
to the more electrically inclined. Why do we use 16 x 820uF caps instead of less Larger capacity ones? is it a price thing? or is to do with the caps properties?

Open Source DC controller

Posted: Fri, 19 Mar 2010, 18:37
by mcudogs
It's the cap properties, ESR, ripple current and temperature.

Open Source DC controller

Posted: Fri, 19 Mar 2010, 23:00
by jstan
Anyone interested in building a controller, I have just received 100 mosfets that I had to wait almost three months for to get a good price, I still have 70 to spare. You require 10 per controller. They cost me $7.00 landed.
John

Open Source DC controller

Posted: Sat, 20 Mar 2010, 01:23
by BiGH
JohnStan - thanks but i'm looking at using IGBTs

Mcudogs - I'm wanting a higher voltage caps than 200v, did you want to suggest some parts that had say a 400v limit?

Open Source DC controller

Posted: Sat, 20 Mar 2010, 01:51
by mcudogs
Try the digikey P11666-ND. They are only 560uF so you may need to use more of them.


Open Source DC controller

Posted: Sat, 20 Mar 2010, 01:59
by mcudogs
Taffy wrote: That is looking good Don. Be interested to hear what its like when complete and running in the car.


Taffy, it's now complete and installed in the car. I have powered it up and run the motor with wheels off the ground. It was very smooth. I am just finishing the BMS wiring and I should have it on the road next week.

Some more pics here Cougar Pics

Open Source DC controller

Posted: Mon, 22 Mar 2010, 02:01
by polo-ev
For those that are interested or missed out last time.
We now have more complete kits & kit components if you want to build one of these controllers.
See the link here for all the info.
http://www.hometheatre.net.au/index.php ... 9468e7cfd5

Open Source DC controller

Posted: Wed, 12 May 2010, 20:29
by BiGH
mcudogs wrote: Try the digikey P11666-ND. They are only 560uF so you may need to use more of them.


Hey Everyone,
I don't have the logic side done as yet, i'm waiting on the next incarnation that will include CAN, but i've got a need to order some parts for my ebike from Digikey, and I figgured I may as well order the powerside capacitors at the same time.

adamj12b has said he's using Panasonic TS-ED caps. 6 for each IGBT, 560uF each 400V.
MCUDogs has linked me to the same caps.

How does one select the right caps?
I was wondering why these properties? How can I calculate alternatives?


The Soliton1 advertises that it has: "Cutting edge capacitor: State of the art ultra low loss 600V film capacitor. No electrolytics to dry out and fail!"

What kind of film capacitor would they be talking about - I checked wikipedia, and there is a plethora of film caps.

On the Spec sheet for the Panasonic TS-ED it states 105 degrees 3000 hours. I'd like to get a capacitor that won't die after such a short time.

on a side note, I'm going to look at a donor car tonight... oh the excitement (1992 Toyota MR2)!

Thanks,
Haydon
    

Open Source DC controller

Posted: Wed, 12 May 2010, 22:21
by bga
Ones like the these from Epcos
of the chinese ones I'm using from Eaco

Because these are parallel plate polypopylene, there is no electrolyte as in normal electros. This allows film capacitors to have very long lives at elevated temperature.
Another useful property is extremely low ESR, in the 3 milliohm range for 560uF. In an electro, the capacitor dielectric is aluminium oxide in an anodised layer with a wet electolyte making contact between then plates. The higer ESR (237mR for the TS-ED) is caused by the electrolyte conductivity, in part.
The RMS current handling is much better with film capacitors, 50 amps vs 3 amps for the TS-ED electros because of the low ESR and losses.

A lot more electos are needed to obtain the same RMS current capacity as with big power film caps, but the electros are more compact per uF than are film caps.
The film capactors are readily available up to 1000V eliminating the series-parallel eletro networks that would otherwise be needed in higher voltage systems.

There is a break even point near 300 volts where film caps are clear winners. At lower voltages, they are more bulky due to the thicker plastic dielectric (they are almost always rated for 600V or more) and so cant't readily complete with aluminium electros.

When I did the evaluation for my 600V IGBT drive, I concluded that I would need so many electros that the capacitance would be silly and the precharge time and current would be unreasonable.
With high per unit current capacity and low ESR, a much smaller capacitance film capacitor bank would work equivalently with a much shorter precharge time.

[edit added:]
The purpose of the capacitors is to buffer energy from the battery, shielding it from the high current spikes when the motor speed is low and the current is high.
In this situation (like planting it at the lights) the capacitors are seeing full motor current (700 Amps?) every time the transistors turn on, which is about 1/10 of the time or less so the average battery current is much lower, probably 50 or 70 amps.

Without sufficient capacitor, the leads to the battery and the battery itself would be subjected to these large low duty cycle spikes, leading to increased battery and wiring losses.

The rest of the time (90%) the flyback diodes in the controller are conducting the motor current, causing them to get hot or die sometimes.


Open Source DC controller

Posted: Wed, 12 May 2010, 22:52
by Johny
bga, where did you buy your Eaco capacitors from - which distributor?

Open Source DC controller

Posted: Thu, 13 May 2010, 04:23
by Thalass
How much does it cost to build the Cougar? And how well does it handle a lower voltage and current motor? I imagine it would just be a matter of changing various settings.

I'm thinking of using an Altrax 7234, with a Mars ME0709 motor for my bike, and that motor is 15kW peak (which is around 200a @ 72v), but if this controller can be built to match that sort of power requirement for a similar price ($735 for the altrax from EV Works) I'd probably prefer it. Being a fan of Open Source and all.

Open Source DC controller

Posted: Tue, 18 May 2010, 17:14
by hipo_ev
I read that in the current design, you can swap out the mosfets for IGBT to support higher voltage and amps.

I would like the controller to support ~240v and 1000a, so is it as simple as purchasing IGBTs that are rated high enough to support that voltage and amps - or do other components in the opensource controller need to be replaced/upgraded, as well, to suit the higher voltage?

I have very limited understanding of electronics, so what exactly would need to be done? Does anyone have a revised design that includes support for IGBTs?

Basically I was looking at using something like THIS. With each of those IGBTs rated at 1200v and 400a, I was hoping that using 4 of them would give the controller enough headroom to support 1000a and ~240v. But what else needs to be done to make that a reality?

Thanks.

Open Source DC controller

Posted: Tue, 18 May 2010, 17:27
by Johny
I don't think you have to do anything to change to IGBTs but those who have done it should comment here.
My post is to warn you that the devices that you have indicated have very high drive requirements and not very good saturation characteristics.
Higher voltage IGBTs appear to have higher saturation values so you may be better to go for 600V devices (brand to brand vary a lot as well). These ones are 2.5 volts at 200 Amps which would have you dissipating 2.5kW at 1000 Amps.
Gate drive wise, 75000pf Gate capacitance is very high and I doubt the standard Cougar could drive it properly.

Open Source DC controller

Posted: Tue, 18 May 2010, 17:42
by hipo_ev
Thanks for the reply Johny. The IGBTs I listed were just some that I found for sale on Ebay. I have not purchased them so am very flexible as to what I will use.

As mentioned I have very limited (high school only) knowledge of electronics - basically I was just searching ebay for an IGBT that was rated higher then 240v and was cheap enough so I could buy enough required to support the 1000a I am wanting.

Would something like THIS be more suitable?


Open Source DC controller

Posted: Tue, 18 May 2010, 17:48
by BiGH

I think your probably aiming for a similar level of performance as I am. i'm looking at doing a toyota MR2 with around 288v (or whatever the max of the motor is) with up to 1000a). Assuming some voltage sag thats what 250 kw? more than enough power for a light small car with limited batteries. I'm thinking of having two modes - one for HiPo driving and one for daily, limiting the amp draw to preserve the batteries.

The best place to check this out is the ecomodder thread. Ive been following jack bauer's progress very closely there. Jack was able to drive his IGBTs with a 12v level (which is what the standard cougar has). and with a few minor modifications requried to reduce ringing when the igbt opens... The specs however call for a 15v drive, which requires a different dc to dc converter, which I belive Jack is experimenting with at the moment. Jack is very open to what he's doing though, so once he's happy with it i'm sure he'll publish the final values.

NB: If you're looking at running higher voltages, you also need to get higher voltage rated caps. I kicked up a bit of a storm when I asked on the ECOmodder forum. There are some very cool capacitors out there (with extremely rated long lives. the standard electrolytics only have a rated lifespan of 3000 hours).


I would try and go for higher amp rated modules than higher voltage. You're never going to run over 500v in a DC car unless its a twin motored drag car. I don't konw of many motors that will live with that level of voltage without frying. (this is one of the reasons the white zombie drag car went to two motors)

I recently bought 8x Powerex CM400DU-12F 600v 400a from ebay (i only intend to use 4 on the controller but the 8 price was much better than the 4 price. ) Just means i've gotta sell off the other 4.
Haydon.

Open Source DC controller

Posted: Tue, 18 May 2010, 17:55
by Johny
One of the things you may have to be careful about using these big packages is that the Bus bars are going to be a lot different. With limited knowledge you are going to be safer getting advice from the guy(s) Haydon has mentioned. I kind of got the feeling that they were using similar packages to the Mosfets. This would limit the mechanical changes.
Image

Open Source DC controller

Posted: Tue, 18 May 2010, 18:03
by hipo_ev
Hi Haydon,

Yeah, sounds like we are aiming for similar performance.

I've mentioned it in my other thread, but basically I'm looking to build a car suitable for track work so it can be used in time attack events.

The car will need to be able to perform 3 laps around a race for each round of the time attack. 1 warm up lap, one hot lap (max throttle) and one warm down lap. The hot lap should take no more then 3 minutes, but that still means upto 3 minutes of max throttle, and I'd assume with max amps drawn out of every corner. I doubt it will be max amps 100% of the time, but it would still be very demanding.

Have you already built your controller with those IGBTs? Do you have a schematic, and a parts list for the higher rated caps etc?

Thanks for the replies guys.

Open Source DC controller

Posted: Tue, 18 May 2010, 18:16
by BiGH
hipo_ev wrote: Hi Haydon,

Yeah, sounds like we are aiming for similar performance.

I've mentioned it in my other thread, but basically I'm looking to build a car suitable for track work so it can be used in time attack events.

The car will need to be able to perform 3 laps around a race for each round of the time attack. 1 warm up lap, one hot lap (max throttle) and one warm down lap. The hot lap should take no more then 3 minutes, but that still means upto 3 minutes of max throttle, and I'd assume with max amps drawn out of every corner. I doubt it will be max amps 100% of the time, but it would still be very demanding.

Have you already built your controller with those IGBTs? Do you have a schematic, and a parts list for the higher rated caps etc?

Thanks for the replies guys.

Yeah sounds like it, im looking to convert an MR2, but don't have the donor car yet. I haven't built a controller yet, i'm still in research and part acquision phase. I got the IGBTs while the ebay item was on sale. I need to buy capacitors, bus bars and the cougar controller. I'm currently debating the best capacitors to use http://ecomodder.com/forum/showthread.p ... post174180.

I know that MPH is working on a new version that has a CANbus interface which is very cool. I'm holding off purchasing the kit until the results for that come out.
Johny wrote: One of the things you may have to be careful about using these big packages is that the Bus bars are going to be a lot different.


Yep thats right - basically the entire powerside has to be redesigned. HOWEVER this is a good thing, as the busbars simply need to be drilled and mounted to the IGBTs at the right spacing. This also means there is no need for an etched or milled thick PCB. The USA guys are mounting big caps directly to the bus bars.

Your going to need some active cooling on the motor and the controller (most likely water cooling for controller). IGBTs get hot and can undergo thermal runaway.

For an awesome cooling solution check this out: http://endless-sphere.com/forums/viewto ... 9&start=75 I think we would both want a setup like that. When people talk about saturation do they mean electrical or heat? because if its heat this would be awesome.

Open Source DC controller

Posted: Tue, 18 May 2010, 18:20
by BiGH
Oh, i'm also trying to find some contactors that are rated to that level (300v 1000a). I want to run multiple contactors through the battery pack to split the 288v LETHAL DC into smaller safer voltages. so that when the car is off its split into 72v packs. which might also help with charging

Open Source DC controller

Posted: Tue, 18 May 2010, 19:06
by hipo_ev
Sorry for the off topic

The two highest rated contactors I know are:

Tyco Kilovac EV500-4A
Gigavac GX16

They can both break ~3000A, and safely break ~2000a at 240v, but even those are not rated at 1000a. The Tyco is rated at 750a and the Gigavac is rated at 600a.

Both should, however, be perfectly fine if you do not exceed the rated amps. atm I'm looking at running 3 Gigavac GX16s. One on the + from the pack, one on the - from the pack, and one in the middle of the pack to split the pack into 2.

But my main point of safety will not be the contactors, but rather a manually operated (switch/lever operated by cable) circuit breaker, which I am still researching - does anyone know of a single pole circuit breaker rated at a min of 2000a and 300v that is relatively inexpensive :p


Anyhoo, I'll go check out the ecomodder forums and see what others are doing to add IGBTs to the DIY controller.

Thanks.