Instant Start 18 fet Infineon Boards are here...

methods

1 GW
Joined
Aug 8, 2008
Messages
5,555
Location
Santa Cruz CA
We now have 30 of the 18 fet instant start controllers.
These are MUCH better in many ways than the last batch of 18 fet slow start boards

18 fet fast start.jpg

Notable major improvements:

- Instant Start
- Programable
- High Power wire routing corrected (interference with caps/exit hole)
- 4 shunts
- Insane capacitors
- Good up to 24S lipo (100V)
- >>100A. We will test them to destruction :twisted:
- etc (4110's of course)

Gary will receive the first completed board.
Look to Gary to start a review thread on these fine controllers

Please dont start asking if you can "Get One"... Gary will be in charge of that.
I may release a few "kits" to people around here who can test them to the limits.

-methods
 
:twisted: :mrgreen:

Can you confirm that the case and mosfet holes footpring and dimensions is the same as the previous 18 mosfet controller?

In other words, does the alluminium bar for sinking heat of the mosfet and the distance between mosfets is the same as the older version with soft start?

Doc
 
I have had these for a month.
Now I can release them

-methods
 

Attachments

  • topSmall.jpg
    topSmall.jpg
    176.7 KB · Views: 7,505
  • bottomjpgSmall.jpg
    bottomjpgSmall.jpg
    224.8 KB · Views: 7,363
Doctorbass said:
:twisted: :mrgreen:
Can you confirm that the case and mosfet holes footpring and dimensions is the same as the previous 18 mosfet controller?
In other words, does the alluminium bar for sinking heat of the mosfet and the distance between mosfets is the same as the older version with soft start?
Doc

The heat sink and case are nearly Identical.
They can probably be interchanged.
They are hand tapped though, so each is slightly unique and requires tuning during assembly.

This happened last night so....


The builds may not go as fast as they were going.


-methods
 
Oooh, broken ring finger, hope you are OK, Methods!

Will Keywin be selling these, or do you and Gary have an exclusive? I want one.

Is there any chance you could get a less fuzzy pic of the front (SMT) side of the circuit board?
 
That's Keywins picture. I suspect it is fuzzy for a reason.
My boards have caps on them so you wont see the details.
I will take pictures later.

Keywin will have these but maybe not for a little while.
He is still developing the software from programming them.
I got these early on a special deal.

Gary and I are going to populate these with 4110's, Crystalyte compatible connectors, proper marine grade 10 AWG copper, a power switch, CA tap, built in USB, and a few other things.
I cant speak as to what version Keywin will be selling.

-methods
 
methods said:
Gary and I are going to populate these with 4110's, Crystalyte compatible connectors, proper marine grade 10 AWG copper, a power switch, CA tap, built in USB, and a few other things.

Sounds ideal, though in my case I would do with bare copper and skip the Crystalite connectors. Otherwise, perfect, and perfect for most enthusiasts on this board in any case.

Will you divulge if the uC is still the Infineon XC846, or is it upgraded to the '866?
 
XCKJ8B116A
09122
More memory, more I/O, more features.
At this point we dont even know what all the new features are.

Everything is bigger and better.
They even did smart things like grouping the Throttle, Brake, Hall, etc. all together right by the exit hole
Power and ground now clear the main capacitors without interference
More holes everywhere - for instance 3 holes for ebrake.

Instant_Start_18Fet_003.jpg
Instant_Start_18Fet_004.jpg
Instant_Start_18Fet_006.jpg
Instant_Start_18Fet_007.jpg
Instant_Start_18Fet_008.jpg
Instant_Start_18Fet_009.jpg
Instant_Start_18Fet_010.jpg
View attachment 2
View attachment 1
Instant_Start_18Fet_013.jpg

-methods
 
Instant_Start_18Fet_014.jpg
Instant_Start_18Fet_015.jpg


The mother-in-law just showed up from out of town for wedding planning :roll:
Maybe if I get them to drink wine they will pass out and I can get some more work done.

-methods
 
methods said:
Gary and I are going to populate these with 4110's, Crystalyte compatible connectors, proper marine grade 10 AWG copper, a power switch, CA tap, built in USB, and a few other things.

That sounds perfect. A number of us will be looking forward to these becoming available!

Is it possible to get an LED status light or one that blinks out error codes to the front of the case? Will the CA tap make current control from the CA? The CA connector keywin added to my 12fet infineon does not.

The thought of having 'no cut corners' in the connectors, wiring, caps actually glued down properly, etc is very appealing. Good work 8)
 
Note that the only wires crossing the board are for the CA.
Those are BAT+, GND, SHUNT+, SHUNT-
Everything else is grouped so nicely!

The only thing I have left to do on this controller is:
- Solder in the phase wires
- Solder on Anderson
- Solder wick copper on the traces
- Clean everything
- Test
- Re-glue the caps
- Drill hole for the button (which is sealed via a rubber nut cover)
- Heatsink compound
- Reassemble
- Test

We need to characterize the shunt and decide if it will be acceptable for high current
(not likely - 100A^2 * 1mOhm = 10W -> This may be ok for burst but certainly not for continuous)
I would like to find a high quality, high power, low cost shunt. Suggestions?

-methods
 
Looks great Methy!

This is exactly the product the E-bike world was missing.

You are doing a lot to help E-bikes and electric transporation to get a foot-hold in this world.

I can't wait to feel a pair of them driving my motors!

-Luke
 
voicecoils said:
Is it possible to get an LED status light or one that blinks out error codes to the front of the case?

There are at least 4 points on the board for LED outputs. We can consider an LED error code output but it would probably be in the case, not external. If enough people request this I can drill the case for an indicator.

voicecoils said:
Will the CA tap make current control from the CA?

Of course.

voicecoils said:
The CA connector keywin added to my 12fet infineon does not.

That is because he hooked the Throttle Control line to the Ebrake. This is the old-school way of doing it.
All you have to do is pull the wire from Brake and add the diode-resistor combo described in the CA manual.
I dont use the CA current limiting because it does not react fast enough with high power systems. I prefer to use hardware limiting.

voicecoils said:
The thought of having 'no cut corners' in the connectors, wiring, caps actually glued down properly, etc is very appealing. Good work 8)

There are places where we could be "more extreme" but I think it starts to dip into diminishing returns.
We considered 8 AWG wire but it is just too damn big, wont fit into 45A Anderson connectors, and basically requires 75A Andersons which are huge overkill.
We considered wiring in all the functions like Cruise Control, Regen, Multi-speed, LED Indicate, etc. but I think that simplicity is better, cleaner, and more reliable. We can add those features upon request. I was considering bringing all those features out on one large connector and leaving it to the user to chose what functions to use. For now we are going to keep it simple for the serious ebike folks.

I too am looking forward to another controller that I have confidence in. The Kelley is confidence inspiring because it is ROCK SOLD. I would like to create the same sense of confidence in a smaller, cheaper package.

-methods
 
Your choices all makes sense in terms of features to include and exclude. The Infineon boards seem to offer a wide range of functionality in their hardware & software.

I think an LED for error codes would be a valuable addition. Perhaps include it internally on long enough wire that the end user can drill their own hole and mount it themselves. It would save you the trouble of drilling and fitting it and save the buyer the trouble of figuring out which points on the board to use and soldering it up.

With the CA current limiting, while it's not instantaneous, it does function well as a 'energy saving' mode. It can let you run at reduced acceleration capability while maintaining the same top speed. I've found it useful for getting more distance from a battery pack and limiting heat build up in the motors. I can see how it would not be suitable as an absolute current limiting function for some purposes though.
 
Boy, you've been busy today. :wink:

As soon as we get a little testing done, they will be available. Eventually, I'll list the completed units on my website. For now, PM me if you are interested in getting one of these "Infineon II" finished models. The pricing is not fixed yet, until we get all the costs together, but it will be under $200. For custom and/or board-only option versions, PM methods directly.

As I mentioned in a couple threads, there will also be "package deals", with a controller and an Astroflight sensored 3210 or 3220 motor, at a pretty good savings over each one individually. We're also looking into adding a CA as an option.

Bummer about the finger, though. :shock: I hope you didn't get any blood on my board... :mrgreen:

-- Gary
 
Hmmm, it is not clear to me that this is actually an Infineon micro-controller. Xie-Chang (the OEM) clearly did build XC846-based controller boards, but the part number "XCKJ8B116A" only shows up on the Xiechang website. There's lots of other stuff there, too, for instance part placement diagram (attached), but it is mostly in Chinese and Google does only a fair job on it. Maybe this is an Infineon XC846 knockoff?
 

Attachments

  • EB218-A-1贴片位置图.pdf
    53.4 KB · Views: 826
GGoodrum said:
Bummer about the finger, though. :shock: I hope you didn't get any blood on my board... :mrgreen:

I am going to finish this board this morning and burn it it.
If it can do 88.8V @ 100A burst I will send it to you.
First I am going to try to blow it :twisted:

Hope I dont blow my 9C trying :shock:

-methods
 
This build took too long from start to finish.
Doing everything "right" takes a lot of time. . .

Wicking all the traces, mounting a switch, calibrating the shunt, working with larger gauge wire....
People who buy these "complete" will need to understand that there is a big difference between this controller and what you get off the shelf from Keywin.

Instant_Start_18Fet_Done_003.jpg
Instant_Start_18Fet_Done_002.jpg
Instant_Start_18Fet_Done_001.jpg

I am going to go out and break it in now.
I have a 24S 88.8V Lipo pack
I am going to run it on a 9x7 Front 9C
I have the current limit set at 100A so I should be able to see bursts to 8KW
Obviously a lot of this will be heat but the controller does not know that...

I ran the controller on my test stand and the free-wheel current was a little high...
Does anyone have test data for a 9x7 9C @ 88V ?

-methods
 
methods said:
This build took too long from start to finish.
Doing everything "right" takes a lot of time. . .

Wicking all the traces, mounting a switch, calibrating the shunt, working with larger gauge wire....
People who buy these "complete" will need to understand that there is a big difference between this controller and what you get off the shelf from Keywin.

View attachment 2
View attachment 1


I am going to go out and break it in now.
I have a 24S 88.8V Lipo pack
I am going to run it on a 9x7 Front 9C
I have the current limit set at 100A so I should be able to see bursts to 8KW
Obviously a lot of this will be heat but the controller does not know that...

I ran the controller on my test stand and the free-wheel current was a little high...
Does anyone have test data for a 9x7 9C @ 88V ?

-methods

I tested the 9C 9 x 7 at 75V in the delta mode and reached 3kW for one minute ,, the motor was like 45 degree C

at 100A you will certainly burn it... :twisted:

Doc

Doc
 
Very nice work Methods

Looks great!!

How did you fasten the 10guage wire in connectors? Crimped Soldered
How tight of a fit was it?

Reason I ask is that my 12 guage wire was a tight fit in the anderson connectors and bearly fit into the plastic plug.
Im thinking 10 guage was quite the challenge
 
Microbatman said:
How did you fasten the 10guage wire in connectors? Crimped Soldered
How tight of a fit was it?

It is a nightmare... it blows my mind that they market those 45A connectors as "10 awg"
Total nonsense! Maybe with the $50 crimper

I have to first crimp them with my crimper then carefully inspect the connections
The hole gets smaller side to side but not on top so any slack must go up
After I get them crimped just so I solder them.
After soldering I take the sanding wheel on the dremel and take a few thousandths off the very top

After doing all that work about 7 out of 10 will plug right in nice and loose.
The other 3 require inspection and correction

-methods
 
Method, What i do for real 10 gauge with 30A anderson is that i only solder them.. it is so thight than just soldering them give a very low resistance lost...

You can enlarge a bit the hole in the connector to allow the full size of the 30A anderson connector.. .I use a long nose plyer nd enlarge it and it fit well..

time saving!

I use them with my 37A CC-CV charger with my battery and at 37A it not really get hot.. just a little watm.. ( the 10gauge dissipate well the heat)


Oh.. I forgot... YOU DID A NICE JOB ON THE NEW CONTROLLER!! :wink:

Doc
 
First Test Results

So I had a few issues getting started :roll: but now I am getting down to business.

Current Limit: 30A for this first test
Test Motor: Front 9x7 9C from Amped Bikes in a 26" wheel
Test Battery: 24S 2P 88.8V 10Ah Lipo
Monitoring: CA calibrated to the shunt
Bike: The wife's woman's Walmart full suspension bike


I started with the battery charged up to 90V
@ 90V and 30A I am seeing between 2100W - 2600W

Lets take this opportunity to clear up a misconception!
There is absolutely NO ISSUE with rideability with this controller!
I can start off as smooth as butter and I can ride as slow as 2mph


And this is about the worst possible situation... I am running a high KV motor with high voltage.
Running a lower KV motor or a lower voltage will effectively increase the throttle resolution.


Throttle Response


Response is great!
Not as good as a Kelley, but good. With the Kelly you can tune the response to be explosive. The response of this throttle is more than adequate for racing. If you are an aficionado you can definitely detect the lag but it is on par with a good ICE response. Your brain will calibrate it out in no time. Twisting the throttle gives a solid snap with powerful authority.

In the next couple of days I will try 60A, then 90A, then 120A
If the motor does not smoke I will try 150A, then no current limit at all.
All at 88V

This will be slow going though as I worked my ass off to build this controller right. I dont want to smoke it without collecting some quality data first :p
This will be a D-Test though.
I am not going to stop till I show what this thing is capable of.

-methods
 
methods said:
First Test Results

So I had a few issues getting started :roll: but now I am getting down to business.

Current Limit: 30A for this first test
Test Motor: Front 9x7 9C from Amped Bikes in a 26" wheel
Test Battery: 24S 2P 88.8V 10Ah Lipo
Monitoring: CA calibrated to the shunt
Bike: The wife's woman's Walmart full suspension bike


I started with the battery charged up to 90V
@ 90V and 30A I am seeing between 2100W - 2600W

Lets take this opportunity to clear up a misconception!
There is absolutely NO ISSUE with rideability with this controller!
I can start off as smooth as butter and I can ride as slow as 2mph


And this is about the worst possible situation... I am running a high KV motor with high voltage.
Running a lower KV motor or a lower voltage will effectively increase the throttle resolution.


Throttle Response


Response is great!
Not as good as a Kelley, but good. With the Kelly you can tune the response to be explosive. The response of this throttle is more than adequate for racing. If you are an aficionado you can definitely detect the lag but it is on par with a good ICE response. Your brain will calibrate it out in no time. Twisting the throttle gives a solid snap with powerful authority.

In the next couple of days I will try 60A, then 90A, then 120A
If the motor does not smoke I will try 150A, then no current limit at all.
All at 88V

This will be slow going though as I worked my ass off to build this controller right. I dont want to smoke it without collecting some quality data first :p
This will be a D-Test though.
I am not going to stop till I show what this thing is capable of.

-methods

Hey methods

If i showed you my 12 mosfet 4110 shunt i had on my first x-lite controller you would of been :O ... at 66V .. i draw 110-115amps .. that controller never blow and i ended up selling it.....

My most powerful controller is my diy to-220 mosfet controller modded to fit 4568 to-247 package fets lol ... Max i draw was 125 amps through my watts up metter lol .. I could of gone more .. but wow .. i was surprised i didn't melt anything load testing .. my front weel was skitting on ashpahlt as i gave it full throttle.. and full brake to load test the controller lol

..

150amps on this 18 mosfet should be fine

What i can not stress enough is making sure you have 100% thermal contact ..

I've had over a dozen controller blow due to poorly machined sinks that where warpped .. CHECK THAT BUS BAR .. put it on a flat surface .. see if any of the ends rock .. put a flat edege to the bus bar and check if its flush by eye ..

properly thermal paste everything .. Done overkill on the paste .... just put enough!! can't wait to see your results!! .. What are you load testing with .. x5? lol .. don't burn out another one methods lol

-steveo
 
Back
Top