Cyclone 650W Motor/External Controller Dissection

[rhitee05]

I got the parts of my Cyclone kit in recently and wanted to post some pictures for all to see. I know some pictures have been posted on here before, but not very recently so things may have changed. This is the 650W motor and external controller direct from Cyclone-Taiwan.

Full set of controller pics here:
http://www.flickr.com/photos/rhitee05/sets/72157622294614442/

The specs on the outside of the controller read 24V/30A, which would be 720W. Based on the component specs, it's probably the same controller used for the 900W/1200W kits, although they probably uprate a few things for 48V operation. Some detail on the innards:
- Capacitors are 50V rated. Standard can-type electrolytic 3300uF for the main input, and a few more 10uF scattered around. There's also a small-ish AC-type capacitor that I didn't get a good shot of.
- MOSFETs are IRFP064V. They seem to be the genuine article, based on comparison of the markings to IR's datasheet. These are 60V FETs in a TO-247 package rated for >100A, Rdson<5.5 milliohm. They seem quite decent actually. Not quite as good as the benchmark 4110's, but fairly close except for the voltage rating (not sure you'd want to push >60V through this controller anyhow), and the TO-247 package actually has better thermal characteristics.
- The FETs are sunk to the extruded case through a thermally-conductive, electrically-isolating pad. It's pretty thin, so the thermal performance should be decent. The extrusion of the case is decently massive, and will probably dissipate fairly well given a little airflow. There's also a small temp sensor attached to the case near the FETs.
- Shunt resistors are 3x 51 milliohm in parallel, so net shunt of 16.7 milliohms. They look like maybe 5W resistors, and they're pressed against the case with some thermal grease.
- External dimensions are 5.25"L x 5.25"W x 2.25"H (approximate). It's 6" wide including the mounting ears.
- Getting mine open was a pain in the a$$. They used steel screws into very crudely tapped holes in the aluminum extrusion. I was able to get two screws out easily enough, but number 3 required a little finesse with hammer and vise-grips. That was enough to let me rotate the panel around, but #4 isn't coming out without a saw. I'll probably pick up some stainless steel screws to replace these at some point.

Motor pictures here:
http://www.flickr.com/photos/rhitee05/sets/72157622184670601/

Not very many pictures here yet. I'm going to tear the motor open soon, once I get some degreaser, new synthetic grease, etc. to clean it up and do it right.
- The gearbox is labeled 9.33:1 ratio, which is slightly different than the 9.55:1 quoted online.
- Motor case is ~3.5" diameter, overall dimensions 3.5"W x 3.5"H x 6.5"L (including shaft and gear, about 4 5/8" long to mounting face).
- There's a small weep hole on the bottom, opposite the wire entry point.
- Interestingly, the motor rotates CW, not CCW as I expected. The standard Cyclone configuration (motor behind BB driving chain) requires CCW rotation. I assume the 650W and up kits must be intended solely for the 3-chainwheel-style kits, where it uses a separate chain to drive the BB, which needs CW rotation. My understanding is this is the same motor used in the 350W/500W kits (without internal controller), so they must swap the direction of rotation when they take out the controller.

I'm happy to e-mail full-res pictures to anyone who's interested, ~3 MB each. I'll continue to post more pictures as I tear the motor down and get my build started.

 

[kfong]

Great start, I’ve only tore apart my controller. Looks the same. I’ve been using it for a year now and it's still running good. The Freewheel needs replacing, but still seems to function. It's the only complaint I have of the setup. At first I didn't like the fact that the cranks were extended on the left, but I really don't notice it, even with all the trail riding I've been doing. I'm quite happy with it to date. I haven't taken the motor apart, will do it this winter and see how worn the gears are. It's been over a 1000 miles, probably more like 2000 now. I picked up a Tidal Force bike from Oatnet. I plan to move the motor higher up, so it will be better suited for trail riding. My current frame is a Diamondback, solid frame. It’s been holding up well in the trails, and for some reason I really like riding it on the trails. Hopefully that doesn’t change with the new setup.

 

[JEB]

I installed a 500 watt Cyclone 2 chain kit, the crank freewheel did not look like it had any grease in it, only oil.

I made a plunger cone adapter so I could force grease into the unit without takeing it apart, used synthetic (sp) bearing grease.

Looking at the excess grease that was forced throught the freewheel, there was just a thin film of oil on the grease,
the unit works fine with the new grease installed.

A friend that used his freewheels just with the oil in them, replaced them often. But he was/is powering his bike with a 2.5hp Honda ICE.

 

[kfong]

Yeah, mine is nearing the end of its life. It creaks loudly when I pedal now. What I like to know is if these things fail completely on you or do they just start to slip. I don’t want to be stuck pushing the bike, but would like to get the most out of them before they break.

 

[kentlim26]

The 24v controller is the same size for 36v or 48, i have all of them, i have blow the 24v. The 24v controller cannot really work well with the 36v Sla battery for long. From a Start and move slow... i heard pok pok sound... inside the cap has BLOWN up, my replacement came for 36v.. i am now using 36v controller. this one is great controller and has no problem for 36v battery or 48v battery sLA or lifepo4. The sticker have 3 types.. 24v 36v and 48v. the motor you are using can also fit for 36v controller or 48v controller. If you like something famous brand controller can match with the cyclone motor is kelly. the 36v controller i had been using for a long time- still going strong.

for 36v controller from headlines i run 48v 20hr lifepo4 battery. it gives me Super impressed range 74km, i took several days to ride and until the battery bms cut off. i don't recommend my battery for buying, it is taking big risk. some people who got my battery got lesser amper, some of them fails their bms. Until today.. my friend (singapore ) bms not yet solved. He is getting a 3rd BMS. now i will run my range at least 65km then i will go for a battery charged. i am also not sure about the future of vpower battery will it live that long ? i am saving up for better lifepo4 either from shangdong hi power or other brands.

to buy a kelly controller or the future product cyclone is coming out twin motor x 900watts? I really like to know if these 2 motor of 900watts can really clock that fast ? top speed said to do 95km/h, visit my http://www.flickr.com/photos/kentlim26 about how it can do 95km/h - crazy fast but i will like to know anyway about these new products.

kentlim26

 

[rhitee05]

I need to spend some more time with the controller open to reverse-engineer the circuit a bit more before I try it at higher voltage. I'd be quite happy running at 10s LiPo, so 42V hot off the charger. The caps and FETs should be good to that voltage, although it wouldn't hurt anything to upgrade a bit. I need to see if the regulator circuit for the control side is good to that voltage, might need a modification or two.

 

[boostjuice]

Rather than the hassle of reverse engineering and desoldering components You could just order one of the 48V controllers by itself from cyclone. They now have them listed for sale singularly on their website and the price is pretty cheap.

The 36V 900W/48V 1200W controllers are the same and have 75V rated Fets (IRFP2907) and 100V rated capacitors so max voltage = 75V (Checked myself)

The 24V 650W/36V 720W controllers are the same and as rhitee05 has discovered have 60V rated Fets (IRFP064V) and 50V rated capacitors so max voltage = 50V

http://www.cyclone-tw.com/order.htm

"650W 24V controller with throttle" = 115USD

"720W 36V controller with throttle" = 115USD

"900W 36V controller with throttle" = 140USD

"1200W 48V controller with throttle" = 140USD

PS: of interest to current cyclone owners, Cyclone is now selling sealed cartridge bearing Bottom Brackets with the 153mm asymmetric extended spindle to match the planetary motor setups. They don't however seem to sell them for the larger 1000W ungeared motors that need an even longer spindle again.

oh and they are now selling 10C continuous claimed batteries/cells for about a third cheaper than their PSI cells.

The end caps look suspiciously like headway cells, but the stepped tubing seems to indicate that they are something else. It would be interesting for someone to take the plunge on these and do a review.

 

[rhitee05]

Rather than the hassle of reverse engineering and desoldering components You could just order one of the 48V controllers by itself from cyclone.

Where's the fun in that?

For not too many bucks, I can drop in the TO-247 version of the 4110 FETs (IRFP4110) and drop in new caps. I should be able to do that for a few bucks less than the difference between the 650W ans 1200W controllers.

If I was feeling particularly evil, I could also adjust the sensing network to up the amp rating. As you pointed out, its a pretty cheap controller if I blow it up...

 

[gunthn]

Rhitee,

I did something particularly evil to my controller the other day..

I rode up to the other side of the city to get the wife's car which was about to receive a ticket as she was at work. I got there and no ticket so I put the bike on the rack and caught the throttle wire yanking it somewhat further out of the controller. I didnt bother to check and went home. Later I went to pick up kids with bike and no go. I pedalled and later opened up the unit was ripped right off the board. As it is soldered, I got out my giant old weller and went to work. Now I am just not that good at this.. While attempting to remove the old wires from the holes in the board I ripped whole traces back. Dejected I almost gave up but upon further more powerful examination of super glasses I noticed that there were unused holes further down. Solder away I did. And finally she lives#@#! Well the temp sender came out and is causing a flashing red but I think that should be it.

Beware the control wires. They are sensitive and difficult to repair.

 

[rhitee05]

Yes, there is absolutely no strain relief on any of the wires. They seem to be mostly relying on the thick 10ga power wires to hold the rest in place, which works okay so long as it doesn't get a really good yank... Glad you were able to get it mostly back into working order. The two sets of control wires in my controller (throttle/ebrake and halls/motor temp) are actually connected to the board with snap-in connectors, which sounds different than your unit. Hopefully, the connector would pop loose before a wire rips out of it. That would be a REAL pain to fix.

 

[gunthn]

Those look like snap connectors on the inside, but they are soldered directly to the board at least they are on mine.

 

[rhitee05]

Interesting. I haven't actually tried disconnecting mine yet...

 

[zeeltom]

I need type IC U1 in this controller. Sorry my english is bad
This U1 is death and im not read who is this U1.
U1 make BOOM ... and two transistors IRFP064V.

Please help

 

[boostjuice]

I need type IC U1 in this controller. Sorry my english is bad
This U1 is death and im not read who is this U1.
U1 make BOOM ... and two transistors IRFP064V.

Please help

It sounds like your controller has had it's motor control IC part number sandpapered off. A common practice for manufacturers trying to hide/protect their circuit design.

U1 is either HITACHI ECN3030 or ECN3031F

I have two Cyclone motor controllers. In my case, my "48V/1200W" has an ECN3030 while my "24V/36V/650W/720W" controller has a ECN3031F.

The only difference between the two is the voltage reference that the current limiting triggers at.
If you told me what the values of your shunt resistors were then i could make an educated guess as to what version you have.

Here's the datasheet:
http://www.hitachi.co.jp/products/power/pse/images/pdf/ic/3030sp.pdf

 

[zeeltom]

Thanks very match, im looking for this

 

[kadali]

I need type IC U1 in this controller. Sorry my english is bad
This U1 is death and im not read who is this U1.
U1 make BOOM ... and two transistors IRFP064V.

Please help

It sounds like your controller has had it's motor control IC part number sandpapered off. A common practice for manufacturers trying to hide/protect their circuit design.

U1 is either HITACHI ECN3030 or ECN3031F

I have two Cyclone motor controllers. In my case, my "48V/1200W" has an ECN3030 while my "24V/36V/650W/720W" controller has a ECN3031F.

The only difference between the two is the voltage reference that the current limiting triggers at.
If you told me what the values of your shunt resistors were then i could make an educated guess as to what version you have.

Here's the datasheet:
http://www.hitachi.co.jp/products/power/pse/images/pdf/ic/3030sp.pdf

hi my friend,
i have 36V 35A controller and i have same problem do you know mine? thank you so much.

 

[boostjuice]

Kadali sent me a PM requesting information on what BLDC controller chip he has in his "36V/35A controller". After being sent photographs of his shunt resistors, I replied with the following.

Hi ,

Your 3 resistors are 0.05, 0.05 & 0.23 ohms which equates to a ~0.0233 ohm parallel equivalent,

From the BLDC controller chip Datasheet we see that;

Overcurrent Protection Capabilities
ECN3030F: Vref = 0.50V
ECN3031F: Vref= 1.23V

Based on this information, I am positive you currently have the ECN3031F motor controller chip installed.

The ECN3031F has a shunt threshold voltage reference of 1.23V which means it will limit peak current spikes to 52.8A (36V x 52.8A = ~1900W)

If the ECN3030F was installed instead, it has a shunt threshold voltage reference of 0.5V and so would limit peak current spikes to 21.46A (36V x 21.46A = ~772W)

You might think that an ECN3030F is more likely installed presently, because it has a closer peak power rating to that advertised by cyclone for the '36V/35A' controller ("960W"). The reality is however that the peak current limit is always much higher than continuous current demand. The controller is rated at 36V/35A continuous (36V x 35A = 1260W continuous) and this power level is above even the peak power level that would be allowed if the ECN3030F was installed (~772W). So it must be the ECN3031F that is installed.

Unfortunately, even if you replace this chip because it has obvious signs of damage, there is no guarantee that this alone will make the controller work again. When your controller blew up, it could have damaged the MOSFETS and other components related to the power electronics. It would be a good idea to at least test the MOSFETS for damage before going to the trouble of replacing the motor controller chip.

Here is a starting point for testing MOSFETS; http://www.4qdtec.com/mostest.html

If you have no electronics skills, it is probably best you find someone who does to do such tests. You will have to unsolder all 6 MOSFETs to test them in isolation from each other and the circuit, and you must take care to not damage them through ESD (Electro Static Discharge) by using an Anti static wrist strap at all times. Also correctly soldering the tiny pins of the ECN3031F would prove extremely challenging using a hand held soldering iron. You would be given a far greater chance of successful soldering using the hot air or oven techniques.

Also, finding a supplier of ECN3031F who will sell them in small quantities will be very challenging or maybe even impossible. This is because they are a very uncommon chip that is usually bought only in bulk by OEM manufacturers.

I did a search using http://octopart.com/partsearch/#search/requestData&q=ECN3031F and could find only 1 supplier http://www.xsmicro.com/rfq.php?pn=ECN3031F&mfr=HIT&ref=octoparts You will have to contact them to ask if they will supply you with a very small quantity.

Octopart is a very good search engine for components and the fact that it only brings up one potential supplier for this part shows how rare it is.

Good Luck!

 

[k-harvey]

My 650 (cw motor kit-I stick to cw as it will do single chain kits & other)was no more power than 500 so upgraded to 36 what a difference i recomend it have tried at 48v but the motor wouldn't last long i recon on that . Now trying kelly controller as other would cut out even with the resistor as Paco recommeded.

 

[Velcro01]

G'Day all,
Just for future reference, for those who would like to mod their throttle the yellow wire going to the pcb is the throttle, red is 5V and black is gnd. See my pic for reference in case your controller is different (mine is 960W).

My cap is 80V, not that it matters with 75V fets...

Cheers, Martin.