Upload Your Circuit Schematics & Requests Here...

[ngocthach1130]

Yeah i just bought the 36v version of the yi yun 42-x. Trying to figure out if it contain any brake sensor pin for emergency power cut off. Don't think it'll help much though.

A few page back i was playing with the Yi Yun 40-x model. After blowing up the capacitor the first time i blew a fuse without knowing. After replacing the cap and fiddling for a week before i notice the fuse is blown. My impatience personality and stupidity set in and light bulb went on. Shorted the fuse holder with a short piece of solid 10 guage wire and proceed to connect everything. Started small from 36v and worked my way up to 84V. Everythign worked fine with wheel off the ground.

Dragged the scooter into my living room to test ride it. Bad...Bad...Bad idea. Started off with only 48V. With a slight throttle, the scooter went nuts and tried to kill me. Went full throttle and the stupid brake sensor no longer worked. I held on for dear life as the scooter swing side to side throwing everything over. Managed to wrestle it to the ground with wheel off ground and pulled the rear brake. I guess with full throttle and brake applied the motor was probably pulling major amp so controller started smoking and blew a FET. I think it took out a gate resistor while at it. I had reconsidered dragging it down to the street to test drive but it was about 1 in the morning. Don't want to dive into anything nor ride straight into traffic.

I'm trying to figure out if the throttle malfunction or i had a short somewhere. Might have put too much solder somewhere and bridge the leg of the fet. This might be the culprit since one leg on the fet is obliterated. Completely vaporize and the housing of the Fet cracked.

I just bought the yi yun 42-x too. One thing i notice is the solder bus to the capacitor on the Yi-Yun 40-x model is a lot flimsier than on the 42-x. At the same time, it only use 3 mosfet unlike the 42-x series.

 

[fechter]

Once a FET shorts internally, you will have full battery juice to the motor and the brake switch or throttle will not stop it. You need a battery disconnect switch or main relay to kill it in an emergency.

The current under load might have been too much for the FETs.

 

[TylerDurden]

Dragged the scooter into my living room to test ride it. Bad...Bad...Bad idea. Started off with only 48V. With a slight throttle, the scooter went nuts and tried to kill me.

This is why all live testing should be videotaped.

Same goes for flushing wild animals from a home...

There is prize money in it for somebody.

:lol:

 

[ngocthach1130]

yeah, pretty stupid to by pass the fuse and a bad idea to fill the shunt with that much solder. I thought i put too much but, you know the old saying. More is alway better.

 

[TylerDurden]

When I was a kid, my pal and I would launch model-rocket engines by jamming a "strike-anywhere" wooden kitchen match in the outlet and lighting the wood end... (we were to cheap to buy proper fuse)

If you ever get the little tip of a kitchen match stuck in a rocket engine, do not take it inside and try to pick it out with a paperclip.

:lol:

 

[ngocthach1130]

Lol we're getting off topic but I remember playing with magnifying glass and melting holes into plastic cup. Forgot and left it on the counter. I got yanked out of bed at like midnight when my mom flooded the floor when she accidentally used the holes filled cup.

When i was small we weren't allow to play with firecracker in the house. I was never the conformist so i decided to get some and lit some off. I used some plastic container to cover the huge wad of fire cracker hoping to muffle the sound. To my surprise the bang was louder and sending plastic shrapnel all over the place.

Alway loved taking apart my motorized toys or charger. Got electrocuted a few time. Guess i never got over the buzz from the shock. Still love to play with stuff like that till now.

 

[rf]

[...]
Always loved taking apart my motorized toys or chargers. Got electrocuted a few times.
[...]

That's shocking!

How did you manage to come back to life each time?

 

[ngocthach1130]

Here's some update to my quest on modifying the YiYun controllers. I'll provide some picture and you'll see what some difference between the YiYun 40 or the 42 series.
On the 40 series controller i opted to use smaller diameter capacitor but they are longer so i still couldn't fit them in the case. I chose to just saw the top off. After numerous exploding caps or wide open throttle incident i haven't test ride the scooter with these controller yet. I've only tested them with wheels off ground and no loads.

I found the 42 series controller had another hole for an additional positive lead to the battery so i added it. At the same time i relocated the wire so they do not bunch up together.. Beefed up the bus for both controller significantly. On the first controller i used rosin core lead free solder and since i was beefing up the bus i added a lot. There's a sticky mess of resin still left on the first board. I bought a huge roll of solid core silver solder. Worked great. the wire bus to the caps on the 40 series is so puny. I wonder if these will pop at high load or higher voltage. So far i only tested them at 84v.

The smaller controller is the YiYun 40 series. It use the fairchild FDP2532 Mosfet. They said it's rated for 150v and 79A. Dunno about that. Last time i vaporized one leg of the Fet and it went nuts on me. They also have thinner, longer, Nippon Chemi-Con 200v electrolytic caps.

The bigger controller in the YiYun 42 series. After that incident with the shorted mosfet i bought IRFB4110 and that's what the bigger controller have. It has more space so i switched the big blue caps from the smaller controller to this one.

Both controller have MBR20200CTG Schottky diode.

Maybe when the weather clear up or a slightly warmer day, i'll drag the scooter down to the street and test out both controller at 84v. I'll find a low traffic time incase anything goes wrong. If i find someone who's willing to record it we'll have something to watch. Even if it's a disaster and blew up then we'll have more to talk about.

Thanks guys, Fechter, Rlt for your great write up and thorough walkthrough on the modification

 

[Beagle123]

Thanks for the pics ngo,

Does anyone know what adding capacitance to the controller does? Its really easy to solder in new capacitors, or add caps. I could have done it on mine, but I didn't know what effect it would have.

I second the thanks to fetcher and RLT they really helped me when my controller blew.

 

[fechter]

At full throttle when the duty cycle is 100%, they don't do anything. At partial throttle or under current limiting, they have to absorb the energy stored in the motor windings during the on part of the duty cycle. This current pulse is pretty big, and the capacitors have to handle quite a few amps. If the capacitors are too small, the current pulse will cause a voltage spike at the same time. Spike too big = blown FETs. Undersized capacitors will also get hot under load. Heat destroys caps, so that's bad.

I've had at least two controllers that failed when the caps exploded. Once the cap was gone, the FETs immediately failed.

Other than a humongous spark when you hook up the battery, there's no downside to having excessively large capacitors. A pre-charging circuit will eliminate the spark.

 

[ngocthach1130]

From the tncscooters website, it look like the yiyun 48 series have built in reverse. These are brushed dc motor so i don't see why we can't make a simple switch that would reverse the output at the controller to reverse polarity going to the motor.

beagle, if you get tired of that whine from the yiyun 42 controller, you could try out the smaller controller. For some reason there's no high pitch noise like the bigger controller.

 

[fechter]

The reverse would just be swapping the motor wires through a relay.
This would not be good for braking, and would likely cause destruction if the switch was activated while the motor was running. In reverse, you would be draining the battery to slow the motor, and the voltage across the controller would be the pack voltage plus the emf of the motor. At full speed, hitting reverse would double the voltage across the controller....poof!

It would be better to short the motor through a resistor for braking.

 

[ngocthach1130]

Actually i thought you actually needed the reverse for reversing the bike or your EV. One reason i said this is because on both of the modified controller, i notice the resistance to push the scooter back is much greater than the stock controller or the unmodified controller. I figure at some point using your leg to back up your scooter or bike might be very hard so you would need reverse. This might be because of the higher rated schottky or cap. I figure it wouldn't be good to flip the controller into reverse while moving. Most EV don't even recommend this.

 

[RLT]

Picked up a couple of Wilderness Energy BRUSHED (BD36) Hub motors for my 2WD project.

Of course the first thing I did was open up the controller to see what was what.

Here is the top:

Nothing too interesting: A 1000uF / 63V Cap, a 470uF 63V Cap and a 220uf 50V Cap.... The PWM controller chip, the shunts and a few other little things.

Here is the bottom:

On the other side are the same two 63 v Caps from the front and (7) TO-220 pack components:
(4) P75NF75 (75V /80A) MOSFETS ( http://www.st.com/stonline/products/literature/ds/8451/stp75nf75.pdf )
(1) 20100 Frewheel Diode
(1) 67F080 Thermal Switch (80°C / 175°F)
(1) 13005A NPN Power transistor

I have no Idea what the 13005A is for (FECHTER!!! ) .
There are also a couple of small surface mount ICs... I didn't think to get out a magnifying glass to see what they are.
Maybe one of them is a regulator for the input to the PWM control?

The case is about half the size of the YiYun 42 series cases No waste space at all, and a somewhat tighter PC board layout. The case is mostly plastic, and while the fins are pretty deep on the heat sink, I think that the YiYuns will be better at dissipating heat... Which may be why the YiYuns don't have a thermal switch and the WE does.

The throttle is basically the same Hall Effect thumb throttle as the YiYun's use, but it has four conductors instead of three.... I ASSUME that the extra conductor is for the 'fuel gauge' LEDs.

Two thing I like about the WE controller is that it has a handy little power switch on the case, and the MOSFETs are good for 15V and 15A more than the 3435s in the YiYun ....
Two things that worry me about it is that there is only the one freewheel diode (compared to the three 20100s in the YiYuns) and that the case is nowhere near water resistant.... I wouldn't even call it splash resistant unless you seal the edge around the heat sink with silicone or something.

Not sure of how easy it would be to run these at much higher voltage since there doesn't appear to be a power resistor / regulator circuit to handle the input to the 'brain'... And I don't know if there is a low voltage cutoff or not either.

 

[fechter]

The 13005A looks to be the voltage regulator, in conjunction with the zener diode parked next door. With power applied to the board, you could measure the emitter to see what the regulated voltage is. Usually 12-15v. This takes the place of the big resistor in the other models.

If you see what the voltage rating of the 13005A is, that would be an upward limit on the input voltage. The nice part about that setup is you can change the input voltage over a wide range without changing anything.

Only one freewheel diode? That's kinda pushing it. That might be OK at a lower current limit(any idea what the current limit is?). It wouldn't be too hard to add another one.

Looks like everything should handle 48v OK.

 

[Dr. Shock]

Looks like it is rated for 400volts.

http://www.st.com/stonline/products/literature/ds/5262.pdf

 

[RLT]

The 13005A looks to be the voltage regulator, in conjunction with the zener diode parked next door.

Interesting... I didn't realize that power transistors could work as voltage regulators.

With power applied to the board, you could measure the emitter to see what the regulated voltage is.

I'll try that when I get a chance.

If you see what the voltage rating of the 13005A is, that would be an upward limit on the input voltage. The nice part about that setup is you can change the input voltage over a wide range without changing anything.

Holy cow! As Dr. Shock reported, it is rated at 400V. SO... This controller in its stock configuration should handle 60V ? . Limited by the 63V Caps, otherwise limited to maybe 72V by the FETS.

Only one freewheel diode? That's kinda pushing it. That might be OK at a lower current limit(any idea what the current limit is?).

I haven't been able to find any solid documentation on it. At batteryspace.com where I bought them, they say that it is configured for 35 amps. Somewhere else I saw a 20 amp limit mentioned... Which kind of makes sense with the 36V - 600 watt rating of the motor... Plus, when I held the wheel tightly, then applied power, when the Watts Up meter hit about 20 Amps, the power shut off. A little solder carefully applied to the shunts ought to bump that ?

It wouldn't be too hard to add another one.

Can you just basically stack them in parallel?
If I use these controllers in my current project, That and bumping the current limit a bit is the only upgrade I think I'm going to need.

Looks like everything should handle 48v OK.

I tried 48 with no problems last night.
I might try 60V just to see what happens, but I currently have no intention to actually run this at 48V or higher. I'll probably stick to 36V, but am considering 40V or 44V of LiIons.
This project is going to be a 2WD (full time rear drive, switch in the front just when needed) intended primarily for doing the high Colorado 4WD trails, so torque, range and reliability are more important than speed.

 

[Dr. Shock]

Yes, how does the 13005 funtion as a voltage regulator? I'm trying to upvolt a couple of brushed controllers to 72v, so it would be interesting to use a 400v FET to do it.

 

[rkosiorek]

the basic idea is called an "amplified zener" a complete explanation of it is on this page:

http://www3.telus.net/chemelec/Projects/Amplified-Zener/Amplified-Zener.htm

the circuit used on the Hongye controller is almost identical to what they have in the article.

the biggest reason for this controller to fail is the single kickback diode is insufficient to handle the back Emf in larger moters and when it fails the FETs will fry next. really should add a couple or three more of the 20100 freewheel or kickback diodes.

rick

 

[Dr. Shock]

"the basic idea is called an "amplified zener" a complete explanation of it is on this page:"

Thanks, Rick. Very interesting.

 

[fechter]

The amplified zener trick is the same setup used in the older Crystalyte controllers. The heat dissipation is the limiting factor, but most controllers only use around 100ma or less to run the logic. The newer controllers have gone to a switching regulator.

 

[lawsonuw]

On the BFx I went with a 1:1 ratio with a slight overlap due to the fear of running the components at their max rating...

But the controller that you're currently evaluating lends credibility to the theory that BackEMF is a fraction of the applied current (which macks sense, if the back EMF was identical to the applied, there would be 0 (zero) force applied to rotation of the armature).

Really the free-wheeling current depends on the inductance of the motor and the switching speed of the controller. In most cases the controller switches fast enough that the motor's inductance keeps the motor current approximately constant. (so a 1:1 ration of on to freewheeling current) Sometimes though a controller is made that switches so slowly that the motor current decays to zero during the off part of a switching cycle. In this case the freewheeling diode will see less current, and the manufacturer can take advantage of the pulse current ratings of a diode to use a cheaper diode. (I don't like such slow switching, it wastes power.)

My 10bits,
Marty

 

[RLT]

Just an update on the Wilderness electric BD36 controller...
Contrary to my earlier report, they apparently are 30+ amps in OEM 'stock' form.
Watts Up said it was drawing 32 amps @ 40.3V before I had to start pedaling up the moderately steep hill to keep the motor from stalling.
So, the 35A rating that Batteryspace lists them is probably correct, rather than the 20A I read elsewhere.

 

[Beagle123]

Hi all:

I'm trying to devise a circuit to test individual cell voltages. I hope to use this 16 channel multiplexer to cycle through each cell and report each voltage to a microprocessor chip:


http://focus.ti.com/lit/ds/symlink/mpc506.pdf

The chip accepts Analog signals from 0-5v. But here's the problem--I can't just send the top battery's voltage to the chip because its at 48v. It would fry the chip. Also, I can't just make a voltage divider because I'd need 16 of them with all different values (and accuracy would suffer) What I'd like to create is a electricl circuit that funtions like a voltmeter. So I can compare the + terminal of the first cell (at 48v) to the - terminal of the same cell (at 44v). So essentially I need to find a component that takes two voltages relative to each other and outputs that voltage in relation to ground (0-5v). I know it can be done because that's the way shunts work. My ampmeter on my bike takes voltages of 48+ volts and does a super accurate comparison of them within .001 volts or so. The result is used by the board that operates at 5v.

Likewise a voltmeter can measure voltage independant of the battery's ground. Hmmm.

Does anyone have any ideas?

 

[tailwind]

Not sure if this is what you want, but it may be a good starting point.

4Cell Battery Checker

http://www.infinity-hobby.com/main/product_info.php?products_id=2158

Greg