How can I use regen for braking only ?

[GCinDC]

man, i don't want to blow a controller this week... wife's away and i've got to drop off/pick up kids in a trailer...

so, just for the sake of asking... if i have 3 470uf 100V caps from another decommissioned controller, could i just swap them out? or would there be a lot of other mods, hardware and software, to make?

may be easier to recommision other controller, but was told it could only handle 75V max...

of course, i don't have to use regen either... i rigged up a switch on the BK - GND jumper..

man, things get dicey up at high voltages...

i've got a 1 mile uphill stretch. halfway up it the other day, my CA Amp reading jumped to 79A and stayed there -- which was odd cause my max up to that point was 36A -- I reset the CA, but it stuck at zero. no more amp readings. after thinking about this i reasoned that the wire must have fallen off the shunt. i was right! point is, does the shunt get that hot!?

anyway, i'm stalling, cause the kids aren't down and i can't do anything anyway...

 

[DISH]

if i have 3 470uf 100V caps from another decommissioned controller, could i just swap them out?

If you do this MAKE SURE you know which side is positive and which side is negative.

I'm a bit concerned as you indicated you had difficulty finding the markings on the capacitors. Installing the capacitors with reversed polarity will indeed cause quite a bit of excitement as you let the smoke out of them.

Unfortunately putting the smoke back in is a LOT more difficult!

 

[GCinDC]

MAKE SURE you know which side is positive and which side is negative.

good tip....... :lol:

 

[GCinDC]

1k, 4.7K, and 22K in parallel will give you 795 ohms. Close enough if you can't find the 56Ks.

i twisted these guys up tight (in parallel), but am only getting 786 ohm... is it me or the math?

 

[GCinDC]

is it me or the math?

neither.. the resistors are off. (at least according to my multimeter...)

i'm going w/ 330 + 330 + 150 in series = 798 ohm, real life measurement.

 

[DISH]

i twisted these guys up tight (in parallel), but am only getting 786 ohm... is it me or the math?

All electronic components have a tolerance, typically +/- 10% or +/- 20% and when it comes to capacitors it's not unusual to see -10% +100%

Your ohmeter also has a tolerance spec. Add all these variables together and you can see why the results are sometimes not quite as expected but usually close.

I suspect anything in the ballpark around 800 ohms would be fine +/- 10% in this case would be +/- 80 ohms and all the values you came up with are withing that tolerance.

You can always trim the value a bit one way or the other if needed...

 

[GCinDC]

thanks. i've learned about resistors today! it's a happy day that i learn something. thanks!

so okay, i tested it out...

background. in addition to the '90V mod' (removing r12 and adding 800ohm between ? and ? [per photo]), i've got BK & GND jumped. i've got an ebrake hooked up - didn't notice where the wires went in the controller.

so, road test observations:
if i touch the ebrake, any throttle power is cut. so that works.
if i accelerate and then let go of the throttle, there's a tiny amount of drag. and regen on the CA shows ~11%. seems always on when not accelerating.
there's no noticeable jump in voltage on the CA when i ebrake.

and here's the bummer: hitting the ebrake doesn't slow me down, not until the pads touch at least... that was the point of this exercise...

so it's gotten too late, but tomorrow i'll try the rig at 60V, instead of 83V...

mike, i'd be curious if there are voltage tests i can run on the board to see if things are as they should be.

 

[amberwolf]

exactly right! the 470uf was glued over. but there is no voltage marking i can make out... except for what looks like a mini cap, right between the huge resistors and the two free standing FETS, and THAT says 63V! it's right next to another even smaller that says 25V.

There are places in the controller that don't run on full pack voltage, so you will likely have several caps in there that are much lower voltage ratings than the ones on the FETs.

The one by the FETs might or might not be on the actual pack voltage, but probably is if it's that close to them.

The important ones (that have the most work to do most of the time) are the largest ones, the 470uF. It is very likely that they do have a voltage marked on them but that it is too far down the side to read next to something else. If it is not readable on one, perhaps it is on another, but it might require desoldering it and removing it to read it.

Unless you are absolutely sure you want to do that, I would leave it alone.

so, just for the sake of asking... if i have 3 470uf 100V caps from another decommissioned controller, could i just swap them out? or would there be a lot of other mods, hardware and software, to make?

If they are the same size, and you are *sure* you want to try it, and sure of polarity, then it might be safe to do so.

If you do this, first desolder just one of the caps off the "new" controller, and verify if it is or is not already high enough voltage spec for you. Pick one you can't see the side with the value/voltage marking at all. If it is only 63V, then it would be worth changing out, along with the other two.

Keep in mind there may be other caps on there that are also supposed to be higher ratings, that might also need to be changed to be sure it won't cause a problem running at higher voltages. Changing the "big three" would be most important, as they do the most work and are most likely to cause a problem if they fail, but the others are also there for a reason, and should be spec'd the same.

i twisted these guys up tight (in parallel), but am only getting 786 ohm... is it me or the math?

If those resistors are the common type, there is a gold band at the far end of each, signifying that they may actually be as far as 5% of their value off in either direction. So a 100ohm might be as little as 80, or as high as 120. Generally that's "close enough" for most things, but when you are using them to create a setpoint (like for LVC) sometimes it's important to be precise. Measuring like you're doing is a good idea in cases like that.

may be easier to recommision other controller, but was told it could only handle 75V max..

That could be the FETs, the caps, or both. Or even a voltage regulator design for the lower voltage section of the controller.

man, things get dicey up at high voltages...

Sometimes they can. Even at low voltages, if the parts aren't up to the task, the same things result (but are not as spectacular because lower energy levels are involved, most of the time).

i've got a 1 mile uphill stretch. halfway up it the other day, my CA Amp reading jumped to 79A and stayed there -- which was odd cause my max up to that point was 36A -- I reset the CA, but it stuck at zero. no more amp readings. after thinking about this i reasoned that the wire must have fallen off the shunt. i was right! point is, does the shunt get that hot!?

I suppose a shunt could get pretty hot, depending on how it's enclosed. Is it in open air, or inside a bag or box?

 

[Mike1]

thanks. i've learned about resistors today! it's a happy day that i learn something. thanks!

so okay, i tested it out...

background. in addition to the '90V mod' (removing r12 and adding 800ohm between ? and ? [per photo]), i've got BK & GND jumped. i've got an ebrake hooked up - didn't notice where the wires went in the controller.

so, road test observations:
if i touch the ebrake, any throttle power is cut. so that works.
if i accelerate and then let go of the throttle, there's a tiny amount of drag. and regen on the CA shows ~11%. seems always on when not accelerating.
there's no noticeable jump in voltage on the CA when i ebrake.

and here's the bummer: hitting the ebrake doesn't slow me down, not until the pads touch at least... that was the point of this exercise...

so it's gotten too late, but tomorrow i'll try the rig at 60V, instead of 83V...

mike, i'd be curious if there are voltage tests i can run on the board to see if things are as they should be.

Be careful about reducing voltage, you can blow your controller up just as easily as increasing it. What are the values of the main power resistors?

A few of things can affect how much braking effect you get.

If your battery has a BMS this may not allow regen current to flow.
The controller may be programmed for the lowest level regen.
You may not be getting the full 90V and the resistors may need adjusting.
The resistors may be soldered in the wrong place - check resistance across original R12 pads (should be a little under 800 OHMs).

You really should make a programming cable.

 

[DISH]

and here's the bummer: hitting the ebrake doesn't slow me down, not until the pads touch at least... that was the point of this exercise...

One observation with a 36 volt system that I have with regen E-brake:

When I leave the house the Li-Ion battery pack is at 42.0 volts and the E-Brake is next to useless.

As I ride and the battery voltage falls the E-brake becomes more and more effective. By the time the battery is down to 38 volts or so the E-brake works about as well as any other rear brake.

The controller on that particular bike is NOT programmable so I don't have a clue as to how it's set up but I know LVC is around 31.0 volts. If I hit the E-Brake when the battery is that low the rear wheel looses traction and I'll leave rubber on the road!!!

I live in a mountainous area (3200 foot ASL peaks) and regen works VERY well. I sometimes arrive at my destination with more charge in the battery pack than when I left depending on the terrain.

I guess I should have mentioned the fourth color band on resistors which indicates % tolerance.

Gold = 5%

Silver = 10%

No fourth band = 20%

So something like BROWN, BLACK, RED, GOLD would be a 1000 ohm resistor with 5% tolerance.

BROWN = 1
BLACK = 0
RED = (number of zeros to be added to above) In this case RED = 2 or two 00
GOLD = 5% tolerance

 

[GCinDC]

What are the values of the main power resistors?

No idea. How can I identify them? Are they shown in a pic somewhere? There are 3 'large' resistor standing up at weird angles between two free standing fets and two caps(?)... pardon me, it all looks like the New Jersey Turnpike in there, approaching NYC... :lol:

If your battery has a BMS this may not allow regen current to flow.

no bms battery. i have a 20S2P lipo pack (74V nominal, 83 hot, i never let it get lower than 75V, 10Ah)

The controller may be programmed for the lowest level regen.
You may not be getting the full 90V and the resistors may need adjusting.
The resistors may be soldered in the wrong place - check resistance across original R12 pads (should be a little under 800 OHMs).

You really should make a programming cable.

Ok, programming cable is on the list. Will check R across R12 pads tonight.

By the time the battery is down to 38 volts or so the E-brake works about as well as any other rear brake.

The controller on that particular bike is NOT programmable...

THAT is awesome. I would LOVE that kind of performance, albeit at higher speeds... What kind of controller/motor/kit is it??

YOUR icon is getting me excited about putting on my snow & ice tires!

Re the resistors, I put on the reading glasses, and went through the whole lot, Identifying and tagging each one... I figured it was time to 'embrace the technology', though i can't imagine how colored stripes work better than numbers....

 

[Mike1]

My regen has at least 3 times as much stopping power as my rear drum brake. Before I did the resistor mod it would only work after I had ridden about 8 miles. Now it works from a full charge.

Look at the "Infineon Technical" thread in the technical reference area. It will show you where the power resistors are and what values will support what voltage range.

 

[GCinDC]

My regen has at least 3 times as much stopping power as my rear drum brake. Before I did the resistor mod it would only work after I had ridden about 8 miles. Now it works from a full charge.

Look at the "Infineon Technical" thread in the technical reference area. It will show you where the power resistors are and what values will support what voltage range.

which resistor mod did you do?
what voltage are you at?
same controller?

is this the section you're referring to?

Nominal Battery Voltage = 72V
LVC = 60 volts
R01A = 330 ohm (2W)
R01B = 330 ohm (2W)
Maximum Safe Operating Voltage = 90 volts

 

[Mike1]

I effectively did the mod you have done (reduced R12 to 800 ohm). I did have the same controller until I melted it, regen wasn't quite as aggressive as with the 18 fet I've got now but it was still pretty good.
I run 72V.

Yes, that looks like the section.

 

[GCinDC]

I did have the same controller until I melted it...

How'd you melt it? Love to learn some lessons the easy way...

Was that your 18fet controller that you showed me the resistor mod? Or your 12fet -- is it 12? or is it 9?? Anyway, it seemed you've got your AUX holes filled.

Do you have a thread for you build?

 

[chet]

Here is my problem.
I jumped from 36v to 60v. on my infineon controller.
At 36v. regen braking was effective but at 60v. it is almost non existent while my battery is > 50v.
Is there a way I could use the braking power of the motor without the regen to the battery ?

TIA

http://endless-sphere.com/forums/viewtopic.php?f=2&t=583&start=45#p45794

 

[Mike1]

I did have the same controller until I melted it...

How'd you melt it? Love to learn some lessons the easy way...

Was that your 18fet controller that you showed me the resistor mod? Or your 12fet -- is it 12? or is it 9?? Anyway, it seemed you've got your AUX holes filled.

Do you have a thread for you build?

I upgraded to a 2KW motor and then tested it by trying to start on a 20% hill. Solder melted in two places and shorted to case, one FET exploded.

The photo I posted was from my 12 FET board (EB812, the same as yours)
What do you mean by "AUX holes filled"?

 

[GCinDC]

I upgraded to a 2KW motor and then tested it by trying to start on a 20% hill. Solder melted in two places and shorted to case, one FET exploded.

uphill testing motor, or downhill testing regen? i'm guessing uphill.. how many amps were you pulling when that happened? did only smoke come out of the controller or was it more spectacular? dangerous?

The photo I posted was from my 12 FET board (EB812, the same as yours)
What do you mean by "AUX holes filled"?

on this part of your photo, the 5 'holes' on the right are filled. mine are empty, and i think it says AUX right under where your brown shrink tubing is... i was just wondering about other possible mods...

http://endless-sphere.com/forums/viewtopic.php?f=2&t=583&start=45#p45794

Thanks for the link... Reminds me that I may have hijacked a thread... Terribly sorry, if so...

 

[Mike1]

Yes, it was up hill. Only smoke so nothing dangerous. Having to push 250 pounds of bike up the rest of that hill wasn't fun though.

Those 5 holes are where the programming header is soldered in. That is all detailed here http://endless-sphere.com/forums/viewtopic.php?f=2&t=8317

 

[GCinDC]

The controller may be programmed for the lowest level regen.
You may not be getting the full 90V and the resistors may need adjusting.
The resistors may be soldered in the wrong place - check resistance across original R12 pads (should be a little under 800 OHMs).

You really should make a programming cable.

Ok, programming cable is on the list. Will check R across R12 pads tonight.

Embarrassing, but I've still not had a chance to do this...

But holiday coming up, and I'm expecting a programming cable in the mail soon... 8)

It's not possible to download the current infineon settings, right?

Before mucking around too much, I thought I'd put in the default settings first, test the bike to see if performance is the same, and then start making changes. Is there a screenshot somewhere of the default values? The one in the Tech Ref Area doesn't have a lot.

What if not all settings are made and the controller is programmed? Where's the programming for Dummies thread? :lol:

Considering I'm running 84V hot off the charger, and am happy pulling 35-40Amps for now, and am happy with the torque... What are all the settings should I make?

PS: one change I WOULD like to make is to soften up the throttle. It sure is jumpy at the start...

 

[ZapPat]

It's not possible to download the current infineon settings, right?

Right.

Before mucking around too much, I thought I'd put in the default settings first, test the bike to see if performance is the same, and then start making changes. Is there a screenshot somewhere of the default values? The one in the Tech Ref Area doesn't have a lot.

Check one of Method's monster controller threads, but this is for the newer 116-based boards (not the infinion MCU based ones).

What if not all settings are made and the controller is programmed? Where's the programming for Dummies thread? :lol:

Never tried, but wouldn't recommend it.

Considering I'm running 84V hot off the charger, and am happy pulling 35-40Amps for now, and am happy with the torque... What are all the settings should I make?

Check Method's screenshots, but lower the settings to something more sane to start off with.

PS: one change I WOULD like to make is to soften up the throttle. It sure is jumpy at the start...

This is kind of unavoidable with this type of controller (doesn't have a motor (phase) current throttle control mode. However, if you want to lower low speed torque, you could lower the battery current to phase current ratio to something lower than the usual 1:2.5 ratio (like 1:1.5 or so). You could also use a two speed switch and start off with the lower speed setting (a 33% speed setting for example will triple your throttle resolution).

 

[GCinDC]

my infineon is an EB812.

i just installed the parameter designer. it has a EB212. should i use that?

i have a feeling more questions will be coming soon....

 

[Arlo1]

my infineon is an EB812.

i just installed the parameter designer. it has a EB212. should i use that?

i have a feeling more questions will be coming soon....

In your parameter design it should let you choose EB812 instead of EB212 it is a drop down menu.

 

[GCinDC]

There are only 5 options:

EB206
EB209
EB212
EB215
EB218

I downloaded the program from the bottom of the "Keywin E-bike Lab Parameter Designer Software Manual" thread.

Is there another version w/ older models?

 

[Chalo]

You would have to switch in some sort of load to dump the regen energy into. This would require high power relays and BIG (expensive) high power resistors.

How do you reckon? High wattage light bulbs in the right maximum voltage range do not have to be very expensive. A few of these six dollar 1200W/120V bulbs should do it, and they can be used in series or parallel as necessary to get the desired effect:

http://www.topbulb.com/find/Product_Description.asp?intProductID=43471

At 10A each, you can just use as many of them as your controller amperage dictates. They'll suck up almost as much current at lower voltage as they do at their intended voltage, but without producing the brilliant light that their 1200W rating would suggest. I'd solder their pins right in and save the cost of sockets.

Likewise, decently potent relays are not necessarily horribly expensive. The following relays aren't rated for 60V, but coupled to a sufficiently low-impedance resistive load, they might not exceed their rated voltage anyway:

http://www.robotmarketplace.com/products/0-SL1106.html
http://www.robotmarketplace.com/products/0-SL1120.html
http://www.robotmarketplace.com/marketplace_solenoids.html

Chalo