Version 2 Crystalyte Controller information

[solarbbq2003]

thanks rick for pointing out reverse location already posted, overlooked it, can add regen braking ( abs ) and and 60/120 select (12/6)to the list, still a few unknown pins left to go

 

[calinb]

i narrowed it down to two FET's that have shorted (shorted across all legs)

So far, I have two blown FETs (V3 and V5 both have all 3 leads shorted together), F1 blown, E2 blown, and the power vias between V4 and D1 are blown open. They carry the current from that small power plane to the solder blob "bus" on the other side of the board. My T's and N's seem okay. Be careful; not all the transistors are alike but you can still match them up with another motor phase elsewhere on the board.

I replaced F1 but, in doing so, lifted a pad (or maybe it was blown off when it popped). Hence, the jumper between F1 and N2. I used a Dremel tool and a Scotchbrite tip to sand off the solder mask on the small power plane. Then I inserted three solid wires into the vias while heating them with a soldering iron to melt them through the solder blob bus on the other side. I folded them over and soldered them to the power plane copper.

I still need to replace E2 and I removed all the FETs because, without a schematic, I wasn't sure if good FETs vs. bad FETs might effect some of my measurement comparisons while looking for other bad components on the board. I might invest in 12 4110s, but I'm worried about throwing good money after bad! Also, I'd like to know if anything else needs to be done to get the board to actually use all 12 FETs. Dave, with ElectricRider.com, thinks the 72 V controller is the only one that uses all 12 FETs. The 48 V versions only have 6 FETs.

I'm also curious if it would be safe to solder up the shunt part way across for more current with the 12 4110s. Has anyone done this on a Ver. 2 yet?

Hope your debug is proceeding well too, monster!

-Cal

 

[The7]

I'm also curious if it would be safe to solder up the shunt part way across for more current with the 12 4110s. Has anyone done this on a Ver. 2 yet?

I would do so if I had V2 ( 6-FET or 12 FET version). Re-inforcing all high current traces with wires. Put wires through the high current "plated through holes".

 

[The7]

Individual gate resistor for FET?

From the picture, it seems that FET 3 and FET 4 use a common gate resistor E2 ( 51 ohm).
So V2 would have common gate resistors for the parallel FETs.
Glad if Fechter could verify this.

These FETs are MOSFETs and they require individual gate resistors to suppress "parasitic oscillation".
Parasitic oscillation will cause irregular firing and damage the FETs themselves.
Common gate resistor for parallel FETs is not good enough.

Besides V2, Shenzhen controller seems also has common gate resistors.

 

[monster]

so what spec is T5? and whats the best way to replace it?

 

[monster]

would it be safer to solder up the shunt and run at twice the current than run at 72V?

 

[The7]

thanks rick for pointing out reverse location already posted, overlooked it, can add regen braking ( abs ) and and 60/120 select (12/6)to the list, still a few unknown pins left to go

Do you mean to use the "reverse" for braking?

Using the "reverse" for braking is NOT the same as "regenerative braking".

There are at least four methods of braking by electrical means.
a) Disconnecting the motor from battery and connecting the motor (as an generator) to an resistor load. All the energy is dissipated as heat in the resistor load.
b) Using the inductive nature of the motor and firing the FETs to "short-circuit" and "open-circuit" the motor in such a way to re-charge the battery. Energy could be recovered and stored in the battery. This is usaully termed as "regenerative braking".
c) Driving downhill, at a certain speed the BEMF is higher than the applied voltage and the motor will also acts naturally as an generator to charge the battery. It has the braking effect ONLY IF it is over-speed. Some people might also call it as "regenerative braking". But I would call it as "generative braking" in order to diferentiate from (b).
d) The "reverse" firing may be also used as braking. It has the greatest effect of braking and could easily cause structural damage. It could bring the e-bike to a dead-stop and then reverse it. All the energy is dissipated as heat in the resistance of the motor. This may be termed as "plugging" .

 

[Knuckles]

d) The "reverse" firing may be also used as braking. It has the greatest effect of braking and could easily cause structural damage. It could bring the e-bike to a dead-stop and then reverse it. All the energy is dissipated as heat in the resistance of the motor. This may be termed as "plugging" .

Oh S##T! There goes my regen idea! So reverse (DPDT idea) is a dead end? That Sucks!
Gosh The7 ... you make science a bummer with all that darn "logical thinking" crap! Gee Wiz!
I am writing a letter to the tooth fairy RIGHT NOW!


It ain't so! But I want to regen brake by wire.
Good Science and Sound Engineering Practice can be a real pain sometimes!

 

[The7]

Knuckles
You could apply these "electrical braking" to your DD but not Bafang.

 

[Knuckles]

I was hoping for that. Will it regen the bats or just generate motor stator heat. I thought it would regen.
I figured the BEMF would exceed the bat voltage if the reverse throttle was engaged while moving in the forward direction.

Please do tell (I am ripping up letter to TF right now) as I want to build it but I am firmly against generating heat. HEAT=WASTE!

And since the Bafang has a freewheel (gearing disengaged while coasting) then it can never be used for regen anyway. YES? Only a DD can regen (act as brake).

OH GREAT TOOTH FAIRY! Please let The7 agree with me! :?
(or is it the Great Pumpkin? I forget. Whom do I pray to again?)
(Isaac Asimov's Ghost maybe?)

 

[The7]

I was hoping for that. Will it regen the bats or just generate motor stator heat. I thought it would regen.

I am afraid that it is your dream. Using "reverse" will only generate heat.

 

[Knuckles]

Gosh you are so negative when speaking truth! Cruel Reality!

Back to the Drawing Board!
(I happy again. Maybe one day The7 will say "good job Bob". maybe .. some day .. maybe ..)

 

[solarbbq2003]

the abs pin on the board is for regen braking ( not related to reversing function),
it uses the inductive method I assume via pwm to induce back emf for the regen current.
Simply have to put a wire from the abs pin to one of your brake handle cut off switches and another wire back to the ground connection pins in right hand bottom corner of board and you have regen braking when brake handle is applied.

I think its not a very efficient method of making regen but makes a hell of a good brake, will hardly make any difference to range unless you start getting into supercaps.

 

[The7]

the abs pin on the board is for regen braking ( not related to reversing function),
it uses the inductive method I assume via pwm to induce back emf for the regen current.
Simply have to put a wire from the abs pin to one of your brake handle cut off switches and another wire back to the ground connection pins in right hand bottom corner of board and you have regen braking when brake handle is applied.

I think its not a very efficient method of making regen but makes a hell of a good brake, will hardly make any difference to range unless you start getting into supercaps.

Thanks for clarification.
"abs" is similar to the method (b).

 

[calinb]

Individual gate resistor for FET?

From the picture, it seems that FET 3 and FET 4 use a common gate resistor E2 ( 51 ohm).
So V2 would have common gate resistors for the parallel FETs.
Glad if Fechter could verify this.

These FETs are MOSFETs and they require individual gate resistors to suppress "parasitic oscillation".
Parasitic oscillation will cause irregular firing and damage the FETs themselves.
Common gate resistor for parallel FETs is not good enough.

Besides V2, Shenzhen controller seems also has common gate resistors.

Common gate resistors seems to be the case. How bad is this? Dave, with ElectricRider.com, says only the 72 V controllers have all 12 FETs installed.

I'm trying to decide whether to refit my board with 6 FETs or 12 FETs with 4130s or 4110s.

six?
twelve?
4130s?
4110s?

@monster, I noticed your C14 is missing. Did you remove it? Mine is populated.

-Cal

 

[monster]

thats just a picture i stole from earlier in the thread.

heres my board. do you spot anything missing?

 

[calinb]

thats just a picture i stole from earlier in the thread.

heres my board. do you spot anything missing?

Oh! Maybe it was The7 who posted the pic orininally:
http://endless-sphere.com/forums/download/file.php?id=10358
Interesting that C14 is missing.

I'll compare your board to my board this evening. Was your board a 12-FET board originally? My board shipped with only six 4310s. Do you think the "parasitic oscillation" problem, mentioned by The7, may have been the cause of your failure?

Edit: Hey--it looks like your board is missing C14 too. I wonder why. Look at your top pic, just above the big orange cap and between V5 and V6. The two pads are there, but no cap. In fact, are any of those caps between every fourth FET populated on your board? They're installed on my board.

-Cal

 

[The7]

Oh! Maybe it was The7 who posted the pic orininally:
http://endless-sphere.com/forums/download/file.php?id=10358
Interesting that C14 is missing.

They're installed on my board.

This pic is originated from Fecther.
I added some comment on his picture.
Maybe Fechter could tell if there is such C14 in his tear-down board.

What is C14 on your board? Could you read its value?

"Common gate resistors seems to be the case. How bad is this?"

It is difficult to anticipate the outcome because it depends on several factors as:
a) switching reponse of the FET used;
b) inter-capacitance between any two of G,D and S;
c) parasitic capacitance and parasitic inductance;
d) drain current or source current;
e) biasing gate voltage;
f) frequency of operation (parasitic oscillation may only occur at a particular range of operation frequency) .

Individual gate resistor is used such that the inter loop-gain of the FET is less unity in order to prevent (parasitic) oscillation. In the design of audio amplifers using MOSFETs, most will be using individual gate resistors. This also applies to switching circuit using MOSFETs.
Common gate resistor will not so predictable because there is direct interaction between the parallel MOSFETs.
Parasitic oscillation (if occur) will be at a very high frequency (order of MHz or higher) and sometimes the magnitude of oscillation might reach a large value to damage the MOSFET and the relating circuit.

 

[solarbbq2003]

Adapted a v2 controller for a customer, installed a dual winding switch, added low voltage var resistor to adjust low voltage cut out so he can adjust to suit his lithiums ( used 0-20k var pot in parallel with existing r15 resistor). A couple of other mods, put a regulator in instead of transistor as have had one board where transitor blew, will be used on 408/4012 crystalyte motor on 36v so Have put smaller caps in, can fit controller with dual speed switch into smaller controller box than supplied by crystalyte.
No on/off switch just using low voltage circuit red wire as on/off externally, no led 'on' light being used.

 

[solarbbq2003]

v2 modded to dual winding type in smaller caseView attachment 1

 

[The7]

solarbbq2003,
Good work.
Is it possible to switch between the two windings during riding?

 

[rkosiorek]

depending on the location of the controller with that huge switch yes you can.

instead of the switch i use an external relay to switch the windings. i have a handlebar mounted thumb switch that i use to select the 408 or the 4012 winding.

the handlebar mounted switch makes it very convenient to switch windings, especially when you are moving. i like the dual speed motors. but i think that the original controllers with the switch are a half assed solution. the manufacturer chose the switch instead of the relay as a cost measure. even though anyone who has tried it with a switch agrees that the relay/handlebar mounted switch is the way to go.

rick

 

[monster]

anyone know what to replace T5 with? part number?

 

[calinb]

anyone know what to replace T5 with? part number?

Maybe someone has a recommendation but I was contemplating finding a transistor curve tracer to borrow and then matching to a good fit from a datasheet, but I think my transistors survived.

 

[solarbbq2003]

when switching windings make sure throttle is in off position, if motor is under load the analog controllers could be damaged, assume same for digital type, there was lengthy dicussion about it on 'power assist' yahoo forum, one solution so could switch when under load was too use voltage suppressing diodes, to stop the back emf spike when you disconnect a winding.