Cromotor + MethTek Monster 24FET + 20S lipo build

[Lebowski]

we still need to try out my Big Lebowski controller.... maybe after I'm done with a software upgrade
(which adds winding inductance measurement and some additional menu options and stuff)

 

[hjns]

Yes, Yes, Yes....

Hopefully, this weekend I can also start diagnosing the MethTek Monster.

 

[hjns]

MethTek Monster has at least 8 blown FETs. No feet left on half of them, some others only have the middle pin left. Body of one FET has evaporated. Furthermore, the extra caps in phase B that Methods added seem to have exploded. Now cleaning the pcb to see if there are any traces left.... and start measuring for shorts in the other 16 FETs.

Anyone interested in selling me 24 FETs?

 

[hjns]

Double post.

 

[hjns]

Uploaded to ES due to slow Dropbox connection.

 

[hillzofvalp]

Jeez, what are the limits of these controllers? What about a beefed up near-stock one?

 

[neptronix]

That's impressive. You have now passed the rite of manhood, sir

 

[jpgey]

Hmm. Not really a happy day. I may have blown the MethTek Monster due to a failing hall. I tested different hall-phase combos with the MethTek Monster, and suddenly there was this explosion-sound inside, a nice spark outside, and the telling smell of burned circuits/caps. Only after the magic smoke appeared in the MethTek Monster did I use my Lyen tester. And behold, one of the hall sensors is dead.

But I am afraid the MethTek Monster is dead as well....

Hello hjn, thank you for sharing all these Cromotor temp datas !
I dont understand why your MethTek Monster had blew:
You had a first failure because of partially inserted bolts, and burn a Lyen sensorless
After this failure you put a Lyen 18fets and it was working.
Then you try the MethTek Monster and burn it "due to a failing hall" and you share a picture with cutted phase.

Sorry, my English isn't very good and there is something i do not understand

 

[hjns]

Hi JP,

Sorry for not being clear. This is the order of events:

1. Bought a Cromotor
2. Ran it with a Lyen 18FET 4110 controller
3. Ordered MethTek Monster 24FET controller
4. Tried to prepare my bike so that I could do a nice comparison between the two controllers.
5. Removed the "current-throttle" via the CAv2 and connected the throttle directly to the Lyen controller. The so-called "current-throttle" is a conventional throttle connected to the CAv2 and not to the controller. I had installed this to be able to determine the amount of current instead of velocity with my throttle. This works much better than a direct throttle (in my experience at least).
6. Then I started riding with the throttle directly connected to the Lyen controller, and then the brake disc came loose.
7. I did not pay enough attention to the wires, and just assumed that a hall had blown. In reality, it was a crunched hall wire with intermittent contact, but I did not know at that time.
8. When testing the cromotor with the Lyen controller, it worked.
9. Assuming I had a working motor, I disconnected the Lyen controller, and connected the MethTek Monster.
10. Then I blew the MethTek Monster.
11. Only after that did I inspect the wires. The rest is history.

 

[hillzofvalp]

I do pray you recover from this incident. In general I abhor rear discs on bicycles especially with giant hub motors. I believe they add a lot of chances for problems to crop up.

 

[jpgey]

Hi JP,
Sorry for not being clear. This is the order of events:

Thank you !

 

[Alan B]

Really sad to see this. Hope it is just the FETs.

Modding these controllers to bypass the limits has its risks.

 

[methods]

Modding these controllers to bypass the limits has its risks.

I take issue with the phrase "bypass the limits". That smacks of some hack who jams aluminum foil into a fuse slot or pulls the bulbs in his check engine light. We are not "bypassing" any protection system... the controller is programmable and we are simply programming it - for higher performance. From the factory they usually program it to around 50A - this is because their technology is inferior. We use better fets, better caps, heaver traces, thicker wires... we assemble a controller that is fit for higher power - so we program it accordingly. No bypass is done. This *is* a higher power controller and it is programmed accordingly.

Now - these blowups that happen on the bench.... That could easily be programmed out. In the future we will ship controllers programmed to have a very low phase current and very short over-current window. This will allow people to test without concern. We do all our bench testing with the controllers programmed to 300A and we have never had a blowup - but then we are **super dandy** on the throttle. I barely give it anything when stepping through the phase combinations - just enough that I can hear the PWM and I let off. I also run an ammeter (a CA) while testing so I can instantly see over-current situations.

As for the pictures above... that is some pretty bad charring and it looks in part to be the result of an over-sized fuse or no fuse. What size fuse was in the system? Once those fets shorted the fuse should have blown saving the circuit board. Maybe it was a slow-blow fuse, that much damage could occur in the time it takes for one of those to drop the solder blob. Sadly that board looks like a loss. We could get it working again but it may not be worth the time. It is very hard to work on the circuit board down between the fets.

I have a prototype here that had a few blowups due to an errant solder blob on the phase that touched the phase power to the gate drive circuitry. It just needs a few more parts swapped out and it should be working fine. We will see what we can do about getting that one sorted out and sent over to you so that you can continue your testing.

-methods

 

[hillzofvalp]

I want to know the exact parts you recommend I add to Lyens version of this controller so i can make it "higher performance." Besides beefing up shit. which caps, etc? can you sell me kit? If that's off topic then pm me.

 

[hjns]

HoA, everything related to MethTek's Monster is on topic, as far as I am concerned.

It's my own stupid fault. I held the throttle for a second or two. But obviously long enough to do some serious damage. I think that I made several mistakes. One mistake was to have the throttle connected directly to the controller, leaving me with a lot of dead space before the throttle would provide a signal to the controller. I should have used my current throttle via the CAv2, and limited current to 2A only and should have used a 5A fuse in line, instead of the 2x 40A fuses in parallel. Also, I did not check the rest of my electrical setup. Obviously, a broken hall wire didn't really help.

I take full responsibility for this terrible waste of a good controller. There are several things to learn from my failure.

• Trying out different hall/phase combos IS testing. In my mind, the "real" testing had not yet begun. I only wanted to find the right combo. However, finding the right combo carries serious risks. Therefore, a low amp fuse is key in saving your controller and motor. Do it when trying out different combos!
• Trying out high power controllers carries a lot of risks. One has to be sure of an otherwise perfect system. I did not check this with a lower power setup. I just "did" it and started trying out different combos "to make it work", which is the wrong attitude.

At the moment I would be very happy with another MethTek controller to test and try out. Thanks for that offer, Methods. Too bad that the pcb seems a goner, otherwise I would have tried replacing the FETs myself.

 

[hillzofvalp]

Are the caps good? Wanna send me those?

 

[hjns]

Hi HoV, let me take a look at the CAPs. I may need them for the other leftover that Methods just promised.....

Anyway, I had a nice ride this afternoon. One hour, partially snow, some very steep ascends in the 30% range. I went very slow, max 30kmh, due to mud, snow, leaves, etc. It was great. Lowered the pressure in my tires to 2.5bar. Lots of grip, but sliding in the snow anyway... Highest currents around 50-60A when accelerating up the steep hills from 10 to 20kmh max. Under these circumstances 4-5kW is enough...

Next time I will bring the GoPro.

 

[hjns]

OK, I replaced the CAv2 with CAv3 with B21 flashed. It works great!

The details are described in my post in the CA thread here. Main bits are that it works, and I need to fine tune the ride now. Ramp up and Wgain are the most important parameters. Will be fun.

 

[hillzofvalp]

Want.

Is 150A/250A a normal setting for the 24FET if it is beefed up (no high performance caps)? What were you planning on running?

 

[hjns]

Hi HoV,

I had planned for 150A/225A aka 1:1.5 ratio for battery/phase current. But I don't think my MethTek Monster is repairable. I will have to wait and see if Methods can provide me with the leftover....

 

[hillzofvalp]

So are the upgraded cAps in his controllers just for longevity? I could get that much power out of a beefed up "normal" 24 fet?

 

[Lebowski]

hjns, do yo ustill ride your bike with all this snow around ? How is controlability ?

 

[hjns]

Hi Bas,

Yep. At this moment, the roads around Basel are still clear. No ice, no snow on the roads, so I am still running my Schwalbe nobbies.

When the black ice starts appearing, I will switch to Schwalbe Ice Spikes and take it slowly.

 

[Lebowski]

Hi Bas,

Yep. At this moment, the roads around Basel are still clear. No ice, no snow on the roads, so I am still running my Schwalbe nobbies.

When the black ice starts appearing, I will switch to Schwalbe Ice Spikes and take it slowly.

Big mistake, Schwalbe snow tires. the spikes in these are not hardened so they last about 1500 km.
Next time get the Nokian brand, the hardened spikes on those last much longer.

Here there was lots of half frozen compacted snow on the roads, not very well cleaned :? I used
my motor to go up the hills but on the flat, the 20kmh without motor was fast enough :|

 

[hjns]

Well, you may be right. However, I bought those monsters last winter, and both tires still have all the spikes after all-in-all 3 weeks of usage, which was approximately 250km. So, I can still use them for another 1250km....