H0TR0D my controller! help me out please...

h0tr0d

1 kW
Joined
Apr 28, 2012
Messages
460
calling for the "gods" of controller/electronics knowledge!!! :D

I've bought a kit with this controller. I really need some help to up the current to 30A continous.
I'm really interested in more torque, speed is fine for the moment.
20121020_143937.jpg
Who's the maker of my controller?
Is it possible to reprogram? If so, where can I buy the necessary cable and download the software?
Praying that its an Infineon... :roll:


What is known so far:

Thanks to a 1kW pair of legs, I know that the motor/controller starts braking/regen/something @ ~333rpm. :lol:

Mods intented:
  • More copper in traces
    Thicker fase wires up to hub motor
    thinnest kapton possible behind the FETs for optimal thermal transference
    Thermal paste wherever needed
    Low ESR caps, possibly bigger capacity
    Shunt mod if programing is not possible

Questions:
  • Mods REALLY needed?
    Wire gauge to beefup the rear traces?
    Suggestions for quality caps, voltage, capacity, etc.
    Fan necessary?
    What else do I need for a REALLY realible controller?
20121020_144007.jpg20121020_144136.jpg
 
BUMP
 
Sorry, I've not seen one of these before, the last time I saw a white board like that was in the old Crystalyte controllers, but this doesn't look to be one of those. Looking at the photos the daughter board seems to be a switched mode voltage regulator for the FET driver supply - it's marked 13V which is typical for the sort of voltage they normally run at.

The FETs aren't great, and will be pretty much limited by their internal die temperature, I doubt that adding anything to improve heatsinking will help much. They have an Rdson that's more than double that for something like an IRFB4110, and close to three times that for an IRFB3077, either of which would work OK in this controller.

Similarly, improving the caps probably isn't going to do much, as the FET switching times are probably modest, and increasing the current limit by the limited amount that might be possible isn't going to make changing the caps really worthwhile. Thicker wires is always a good thing, but the traces are probably fine. Patrick (methods here) did some testing a while ago that showed there wasn't a great deal of point in going OTT with soldering up tracks. The port next to the pad marked XS2 looks a bit like a programming port, but my guess from the pad marking is that it isn't. This may well be a hard-coded controller, without a programmable capability. This means either doing a caveman shunt mod to increase the current limit or tracing the schematic and finding a way to fool the current sense circuit. My guess is that it's going to get pretty warm at 30A continuous with those FETs, so this may not be a realistic target to aim for. Changing the FETs is expensive, and if you're going to go to the trouble of doing that I'd suggest buying a bare 6 FET Xiechang board (no FETs) and case from Keywin for around $20 and playing with that instead. Fitted with IRFB 3077s the Xiechang will run at 30A OK, stay pretty cool and is programmable using the software on this forum, meaning you don't need to play around modding the board.
 
I don't recognize that board, but it looks pretty typical in most respects.
No clue what the flying board is. Possibly a voltage regulator.
30A is probably within the range of the existing FETs, but it might get a bit warm.
There are quite a few FETs out there that have much better specs. Replacing them does require some pretty good soldering skill.

What voltage are you running?

In the picture below, I arrowed the shunt, which measures the current. The simplest way to increase the current would be to solder a wire across the shunt. The shunt material has much more resistance than copper wire, so you have to be careful not to lower the resistance too much or you are likely to blow up something.
One way to mod it would be to solder a very short section of copper wire to the top of the shunt. Wild guess, maybe a piece about half the length of the shunt. Another way would be to solder a longer piece of something like 12ga. wire from one end of the shunt to the other. Maybe a 6" piece. Changing the length of the 12ga will change the resistance.

If you try modding the shunt, it would be important to take some kind of current measurement before and after to make sure you don't increase it too much.9FET controller.jpg
 
Jeremy and fechter, thank you very much for your answers.

I'm running a 12S LiFePO4 ~40V

I'll try the "caveman" solution then... But first, the hall wheel sensor isnt working, so probably a new controller is in order.

Not Jeremy's sugestion for now because because of the warranty.
With some luck, I'll keep both controllers and take one to the limits.

Jeremy, where to buy that controller in the UE? And quality fets and caps? What about selling your DIY controller?

Regards,

RNM
 
Hot rod your wallet first I'd say.

What are the chances he's got a battery up to 30 amps continous?

Hot controllers are fairly cheap when bought from sketchy china ebay vendors. I just got a 48v 40 amp for less than 50 bucks shipped. But you better have the battery for it. That's where the money flies out your wallet.
 
Dogman,

Thanks for the answer. I've got a 12S headway 10Ah, so 3C continous, 10C peak. No problems there I think.

Could anyone tell me places to buy those cheap controllers everyone talks about?

Regards,

RNM
 
Lyen's $130 12-FET is a great controller. The 6-FET and the 9-FET are less expensive and "might" handle 30A continuous, but if you flow the 30A through 12-FETs, it will run cooler, and it won't overheat. That means it will last a very long time.

If you add a temp probe to your motor, you can limit the amps with a CycleAnalyst. Then you can slowly raise the amps with the push of a button until things are at the safe upper temperature limit.

If your current controller works, I recommend you sell it for whatever you can get for it. Hotrodding it would be an education for you, but once you fry it, you can't sell it at all. Best of luck with whatever you choose...
 
spinningmagnets said:
Lyen's $130 12-FET is a great controller. The 6-FET and the 9-FET are less expensive and "might" handle 30A continuous, but if you flow the 30A through 12-FETs, it will run cooler, and it won't overheat. That means it will last a very long time.

If you add a temp probe to your motor, you can limit the amps with a CycleAnalyst. Then you can slowly raise the amps with the push of a button until things are at the safe upper temperature limit.

If your current controller works, I recommend you sell it for whatever you can get for it. Hotrodding it would be an education for you, but once you fry it, you can't sell it at all. Best of luck with whatever you choose...

The sound of reason. Thanks for the great tips SP!

Since I have some warranty issues with this controller (no speed in the J-LCD display...), perhaps I can get another one for free. With this one, I'll educate myself. ;)
 
I haven't ever boosted a shunt to raise the controller amps, but...the more of the length of the shunt is bypassed, the more you will raise the max current the controller can put out. The classic method is to add solder to it.

I wish I could remember who recommended the method I'd like to describe. First you add some kind of meter so you can verify the stock max amps. Then wrap copper wire around the shunt at one end. Start with two wraps, and solder it down onto the shunt. At this point, about 1/4-inch of the shunt will be bypassed, which creates the effect of a shorter shunt, with less resistance than stock. Ride it again and measure the max amps. You will then know how many amps are added by two wraps of wire, and then proceed...
 
Please do not PM people to come look at your thread.

If they want to do so they will find it and help you.

Below is the conversation regarding this. Since it is about a public matter and not a private one, it is being posted in the thread you PM'd me to respond to, in the hopes that you and others will read it and perhaps understand the problem better, and to not do it in the future.

It's especially a problem since much of the information you asked for in your OP is available in various controller mod threads already existing around the forums.

h0tr0d said:
Hello there!

I'm having trouble with my controller, so I was wondering if you could help.
http://endless-sphere.com/forums/viewtopic.php?f=2&t=44859

Thank you very much in advance!
Regards,

Rodrigo


PS- If you are not the right person for this task, could you please tell me who to talk to? :)

amberwolf said:
The right person is the entire forum. Please read the first line of my signature (and that of several other members).
"Got a question that isn't personal or private? Post it in the forums, don't PM it. ;):
If I hadn't had to take the time to read and reply to your private message about a public matter, I might have had time to look at your thread. Now I don't. Sorry.


h0tr0d said:
I sent a PM to you and others just call for atention to the topic that I sent. People known for the great knownledge they possess. All others responded in the topic or to my PM.

For them, many thanks.
For you...

Amberwolf said:
I'm sorry you don't understand this, but it is considered rude to PM people just to respond to a topic they could already see. It's been brought up publicly by multiple people many times that it's not a good idea to PM for such things, and everyone has been asked (via such threads and via people's signatures) repeatedly to not do it.

I get dozens of PMs about this kind of thing when I am already following such threads or would probably peek in and respond if I have any info, and if I have the time. I also get PMs asking to help troubleshoot things in private when there is no reason for it to be private, and should be a public thread, and worse, the same PMs are sent to multiple people, who could much better help the person together, in a single thread rather than using many people's time individually at cross-purposes.
 
AW, FWIW I got PM'd
h0tr0d, Not sure where in the world UE is, but pretty much all of these controllers come from China. There are a lot of unknown controllers around, plus a few made by big Chinese companies that we get to see a lot, mainly from either Wuxi Technology or Xiechang. The latter controllers are the most flexible, as we have access to the programming software and collectively we've learned a lot about them. There are lots of resellers who brand Xiechang controllers as if they were the manufacturer, including e-crazyman on ebay (Keywin), Ed Lyen on here, Crystalite and several others.

You can buy a controller from Ed Lyen here already modded and programmed, as suggested above, or you can buy the same controller for less than 1/2 the price direct from e-crazyman on ebay and mod it yourself. The end result won't be any different, as they use exactly the same parts (from the same source) the choice is down to whether you want to pay a lot more and just fit a controller and ride or save a lot of money and experiment a bit yourself.

Soldering the shunt is a rough and ready way to increase the current limit on a controller, normally it only makes sense to do this if you have no other way, as it's easy to get it wrong and blow the controller and also very often disables the controller fault protection circuit. If you understand the risks and are good with a soldering iron then, as a few here have found, you can make it work tolerably well. My personal view is that there are much better and safer ways of modding controllers to increase the current limit, but to do them means tracing out at least a part of the controller schematic and working out how the two current sense circuits work.
 
Jeremy Harris said:
Sorry, I've not seen one of these before, the last time I saw a white board like that was in the old Crystalyte controllers, but this doesn't look to be one of those.
Perhaps if the OP was to poke around the build/test threads of some of the people he PM'd, he might have seen something very like that controller opened up and described to some degree, including some manufacturer info. ;)
 
Jeremy Harris said:
AW, FWIW I got PM'd
h0tr0d, Not sure where in the world UE is, but pretty much all of these controllers come from China. There are a lot of unknown controllers around, plus a few made by big Chinese companies that we get to see a lot, mainly from either Wuxi Technology or Xiechang. The latter controllers are the most flexible, as we have access to the programming software and collectively we've learned a lot about them. There are lots of resellers who brand Xiechang controllers as if they were the manufacturer, including e-crazyman on ebay (Keywin), Ed Lyen on here, Crystalite and several others.

You can buy a controller from Ed Lyen here already modded and programmed, as suggested above, or you can buy the same controller for less than 1/2 the price direct from e-crazyman on ebay and mod it yourself. The end result won't be any different, as they use exactly the same parts (from the same source) the choice is down to whether you want to pay a lot more and just fit a controller and ride or save a lot of money and experiment a bit yourself.

Soldering the shunt is a rough and ready way to increase the current limit on a controller, normally it only makes sense to do this if you have no other way, as it's easy to get it wrong and blow the controller and also very often disables the controller fault protection circuit. If you understand the risks and are good with a soldering iron then, as a few here have found, you can make it work tolerably well. My personal view is that there are much better and safer ways of modding controllers to increase the current limit, but to do them means tracing out at least a part of the controller schematic and working out how the two current sense circuits work.

UE is a typo, european union... :D

Lyen's stuff is Xiechang "Infineon" type right?

Again, thanks for the great info!
 
spinningmagnets said:
I haven't ever boosted a shunt to raise the controller amps, but...the more of the length of the shunt is bypassed, the more you will raise the max current the controller can put out. The classic method is to add solder to it.

I wish I could remember who recommended the method I'd like to describe. First you add some kind of meter so you can verify the stock max amps. Then wrap copper wire around the shunt at one end. Start with two wraps, and solder it down onto the shunt. At this point, about 1/4-inch of the shunt will be bypassed, which creates the effect of a shorter shunt, with less resistance than stock. Ride it again and measure the max amps. You will then know how many amps are added by two wraps of wire, and then proceed...

Once I resolve my warranty issues, that's what I'll do. Thanks for the pointers!
 
amberwolf said:
Please do not PM people to come look at your thread.

If they want to do so they will find it and help you.

Below is the conversation regarding this. Since it is about a public matter and not a private one, it is being posted in the thread you PM'd me to respond to, in the hopes that you and others will read it and perhaps understand the problem better, and to not do it in the future.

It's especially a problem since much of the information you asked for in your OP is available in various controller mod threads already existing around the forums.

h0tr0d said:
Hello there!

I'm having trouble with my controller, so I was wondering if you could help.
http://endless-sphere.com/forums/viewtopic.php?f=2&t=44859

Thank you very much in advance!
Regards,

Rodrigo


PS- If you are not the right person for this task, could you please tell me who to talk to? :)

amberwolf said:
The right person is the entire forum. Please read the first line of my signature (and that of several other members).
"Got a question that isn't personal or private? Post it in the forums, don't PM it. ;):
If I hadn't had to take the time to read and reply to your private message about a public matter, I might have had time to look at your thread. Now I don't. Sorry.


h0tr0d said:
I sent a PM to you and others just call for atention to the topic that I sent. People known for the great knownledge they possess. All others responded in the topic or to my PM.

For them, many thanks.
For you...

Amberwolf said:
I'm sorry you don't understand this, but it is considered rude to PM people just to respond to a topic they could already see. It's been brought up publicly by multiple people many times that it's not a good idea to PM for such things, and everyone has been asked (via such threads and via people's signatures) repeatedly to not do it.

I get dozens of PMs about this kind of thing when I am already following such threads or would probably peek in and respond if I have any info, and if I have the time. I also get PMs asking to help troubleshoot things in private when there is no reason for it to be private, and should be a public thread, and worse, the same PMs are sent to multiple people, who could much better help the person together, in a single thread rather than using many people's time individually at cross-purposes.

To Amberwolf:

Replying to your last PM: You considered it rude. Some people agree would with you. 1 person, out of 10 that I PM'ed and replyed on topic and PM, found it rude.

Please stop replying off topic or in a mockery fashion. If you took the time to lecture me about forum etiquette, please act like you preach.

If I offended you or anyone else, I'm sorry, it was never my intention.


To all:

If you agree with Amberwolf, please ignore me and keep building great ebikes like he does!
Don't take the time and energy to respond...

If you don't mind a "call for attention" in the form of a PM, count me in!
For me, a small "call for attention" can do wonders. Sometimes, I can get back to you really fast, its a subject the takes me 5s to think and 2 min to reply and save you 2 months of online searching and/or lots of €€€ spent for nothing.

Knownledge sharing is Awesome!
 
h0tr0d said:
UE is a typo, european union... :D

Lyen's stuff is Xiechang "Infineon" type right?

Again, thanks for the great info!

Again, FWIW, I very nearly didn't bother replying on this forum, because, like AW, being PM'd a lot does annoy me (and despite the sig I have on the bottom of EVERY post here I STILL get lots of PMs asking questions). The only reason I replied is that by pure luck I was stuck with an hour or so to kill this morning.

Anyway, here's a brief history of the odd naming that these controllers have been lumbered with and some background that might help:

When they first hit this forum around 4 or so years ago they were being sold by e-crazyman (Keywin,) and labelled externally as "Shenzen Sucteam" on a paper sticker (that I'm pretty sure he just added himself). At that time they used an Infineon badged microcontroller chip (the XC846), so they got the (incorrect) nick name on here of "Infineon" controllers. Then, as now, they were actually made by Wuxi Xiechang Technologies Ltd. These were the first cheap programmable controllers we saw in the West. Around two or so years ago Xiechang changed the design and stopped using the Infineon badged chips (which almost certainly weren't made by Infineon in the first place, in my view). The new version of the controller was very similar in performance (pretty much identical in most respects) could still be easily programmed, but used a proprietary microcontroller (maker unknown) labelled XCKJ 8B116A. These controllers are still (very misleadingly, in my view) referred to as "Infineon" controllers too, even though they have no connection with that company.

To make matters more confusing, we know that the same controllers have also been badged as Crystalyte and a few other names, and Ed Lyen (who has a very close connection with e-crazyman) gets his versions of the same controller badged as "Lyen Edition". All of these apparently different controllers are actually the same core units (you can clearly see the Xiechang logo on all the boards), with minor changes to the wiring etc brought out and, in the case of the ones Ed Lyen sells, with your choice of FETs fitted. Any of them can be modded easily, as the collective knowledge on this forum about these is probably greater than for any other cheap controller that's readily available.

When searching the forum for info on these, it pays to search using both the wrong name "Infineon" and the manufacturer name, Xiechang, as a lot of people still call these things by the wrong name. FWIW, if you're good at reading Chinese, there is some useful info on the manufacturers web site: http://www.xie-chang.cn/cn/index.asp
 
FWIW, some of us regulars do get too many PM's. It does get bothersome to get them when it's not personal. You might want to say shit about a vendor, or another member private, send an address for a purchase shipment, or just do private talk on any subject.

But technical discussions don't benefit lurkers if they take place in PM.

Great that you have a good battery! So many will buy a funky little frog or something, and then want to run a 3000w controller on it.

Sketchy ebay controller vendors are found on,,,,,, wait for it,,,,,,,ebay.
 
h0tr0d said:
If you don't mind a "call for attention" in the form of a PM, count me in!
For me, a small "call for attention" can do wonders. Sometimes, I can get back to you really fast, its a subject the takes me 5s to think and 2 min to reply and save you 2 months of online searching and/or lots of €€€ spent for nothing.

Knownledge sharing is Awesome!

If you try it again I can pretty much guarantee you'll just get put on my ignore list, I'm sorry to say! Life's too short and my time is normally too precious to spend it trying to answer PMs or be pushed into being polite by having to respond to a thread to avoid being seen as rude.
 
1,2,3, Everybody PM Hotrod.

This is a very fun and friendly forum, till you rub the fur the wrong way a few times. Then you find out the moderators don't edit everything here.
 
to all:

Thank you for your great replies and I can assure Jeremy, AW or Dogman that I won't send another PM like yesterday. My intention was that people replied on topic, not waste time with my PM, since I agree 100% with your signature.

As for my behaviour, I've done this over the years, online and IRL business and personnal relationships with good success. Of course, I will respect everyones choice.

For all who might have mechanical questions (my line of expertise), please do like I did. First an Topic, then a PM. I'll do my best to try and help like others helped me with this topic.

Jeremy:
Regarding your DIY controller, I fell in love with its simplicity and great spec components. Sure you aren't interested in selling a few?


Dogman:

LOL!
I was asking for some specific vendors, that's why I wrote that. But apparently, like Jeremy pointed out, they are all selling controller made in 2 factories. Going to find the cheapest one then... thanks!
 
PM'ed

by Lyen

Hi Rodrigo,

You may simply bypass the shunt to increase the current. You may try to put/solder a piece of wire between the shunt and reinforce the power related traces with additional stack(s) of wires, thicker wires, and change out with higher power capable MOSFET transistors. Be aware doing so may have the potential to cause premature failure due to the MOSFETs and overall design are for around < 15A only.

The controller is likely not user programmable since it was designed to be programmed only once from the factory if it is a digital controller.

Regards,
Edward Lyen
 
Also PM'ed

by John in CR

Reduce shunt resistance at shown here http://endless-sphere.com/forums/viewtopic.php?f=2&t=31643&p=458366&hilit=controlled+shunt+modification+with+controlled+results#p458366

Don't go to far, maybe wrap only 1/4 to 1/3 of the length with copper and try that.
 
This is definetly getting strange... Jesus!

YO! NOT TRYING TO ANNOY ANYONE! just sharing what others shared with me, with their written consent.

Please lets stop and just focus on topic!
 
Back
Top