Instant Start 18 fet Infineon Boards are here...

[methods]

Here is a messy overlay
I am sure one of you artists can put me to shame

View attachment xiechangtranslationJPG.jpg

This is awesome...

So I dont see any control for multiple current limiting.
This leads me to believe that there are simply jumpers on the board that work something like this
Jumper A
Jumper B

A to ground means low current
B to ground means Medium current
A + B not grounded = maximum current

If you look at Kenny's picture this makes sense too - his rocker switch has full power in the middle.
Now all we have to do is figure out which pins!
Doc - you have one - Wont take but a second to trace the pigtail in and check which pads it goes to.

Once Doc finishes up with his drag bike we will need to pump him for info!
Doc is the only one who has both the Crystalyte version and the Infineon version.
We must pour over these boards to detect every little difference.
With these differences we will learn more than we have ever known about the Infineon boards.

With translators at our side we can now unlock the secrets hiding on the website too

-methods

 

[Lyen]

Good catch on my mistakes SAM-Pilot!

Methods & SAM-Pilot & Doctorbass, thanks for your compliment. As for the program installation issue, I simply installed the software on a Chinese OS. However, according to the included documentation in the PDF file, there is a special instruction stating that you need to copy from an X folder to a Y location and such. Lastly, there is another instruction telling you what you hook this up first, second, and so forth. Let me know if you guys need more detail information on program installation & computer to controller hardware setup/interfacing + operation procedures. :wink:

BTW, Doctorbass I just want to let you know you are my admirer in electronics and batteries hacking. I have read nearly every single threads you have post. You have indirectly helped me built my ebikes and scooters and also helped me spent my fortune on it. Good job buddy!

I primary support American products first, others second.

 

[Doctorbass]

Thanks lyen for your appreciations.. your help is also appreciated as well! :wink:

 

[Doctorbass]

Here is a messy overlay
I am sure one of you artists can put me to shame



This is awesome...

So I dont see any control for multiple current limiting.
This leads me to believe that there are simply jumpers on the board that work something like this
Jumper A
Jumper B

A to ground means low current
B to ground means Medium current
A + B not grounded = maximum current

If you look at Kenny's picture this makes sense too - his rocker switch has full power in the middle.
Now all we have to do is figure out which pins!
Doc - you have one - Wont take but a second to trace the pigtail in and check which pads it goes to.

Once Doc finishes up with his drag bike we will need to pump him for info!
Doc is the only one who has both the Crystalyte version and the Infineon version.
We must pour over these boards to detect every little difference.
With these differences we will learn more than we have ever known about the Infineon boards.

With translators at our side we can now unlock the secrets hiding on the website too

-methods

Can we just get back to the HARDWARE a bit

I just finished to assemble my infineon 18 fets controller.... AND IT WORK !! lol

not very complicated but with all these NEW board and the update i did on it, that add a bit of risk..

I observed one imnportant thing that may fail on the board. The 3 smt 1Kohm resistor in serie near the 3 big resistor of the regulator become VERY hot at 100V.. I mean.. very very hot!.. I thing that they may unsolder if that is not corrected... is it normal?... cause smt component can not take so much heat!!.. I guess not.. Methods, did you noticed that? ( just to let you know i used the 3 parallels 1.8K you sent me for 100V.)

I also added two more shunt in parallel to the existing 4 shunt.. now total of 6 shunt parallel.
I used my precise 250W DALE 0.1ohm resistor as reference to calculate the real current.. and than the resistance of the shunt on the controller.

That give me 0.793mohm

Method.. I remember you said that when you have set the shunt to 250mohm, the limit of the current was the too small wires of the motor?.. can you confirm ?.. I want to set it to 100A first.... then i'll see how well it work before to increase current. ( I have acces to mant thermal IR camera at work.. I may examine that controlerl with thermal imaging to locate potential over heat problem... like mosfet not correctly thightened that overheat.. etc


Pics or video of the working controlelr SOON

Doc

 

[Doctorbass]

Also.. I guess that modifying the shunt will just lie the current value that the uC read and that we may need to calculate a multiplyer factor when setting the current limit o the software since we mod the shunt...

I wonder if it already have storred seting like the oem measured shunt value to calibrate the data that the programing software read?.. and where may be these data.. string.. etc

Doc

 

[methods]

I observed one imnportant thing that may fail on the board. The 3 smt 1Kohm resistor in serie near the 3 big resistor of the regulator become VERY hot at 100V.. I mean.. very very hot!.. I thing that they may unsolder if that is not corrected... is it normal?... cause smt component can not take so much heat!!.. I guess not.. Methods, did you noticed that? ( just to let you know i used the 3 parallels 1.8K you sent me for 100V.)

This concerns me. Please measure the temperature though before getting too worried... The three 1.8K resistors are the correct resistors and they should be good to 107V. My guess is that you will find that the surface mount trio are getting very hot - but not so hot as to de-solder. We can look up the max rated temperature. Worst case we can replace that trio with a single power resistor.

Method.. I remember you said that when you have set the shunt to 250mohm, the limit of the current was the too small wires of the motor?.. can you confirm ?.. I want to set it to 100A first.... then i'll see how well it work before to increase current. ( I have acces to mant thermal IR camera at work.. I may examine that controlerl with thermal imaging to locate potential over heat problem... like mosfet not correctly thightened that overheat.. etc

Yes, my 100A current limit was certainly governed (to some degree) by the 16AWG wires. Please test at your current shunt resistance and do not proceed directly to 0.250m. Gary has a controller with 0.250mOhms. Maybe he will test this weekend and give us an idea of how much current this will allow.

Anyhow, thanks for the feedback on the 12V regulator circuit. This is party why I have been delaying on releasing kits to the general public. I want experienced guys like you and Gary to gracefully run into any possible bumps in the road that may appear.

Keep us updated.

-methods

 

[Doctorbass]

I observed one imnportant thing that may fail on the board. The 3 smt 1Kohm resistor in serie near the 3 big resistor of the regulator become VERY hot at 100V.. I mean.. very very hot!.. I thing that they may unsolder if that is not corrected... is it normal?... cause smt component can not take so much heat!!.. I guess not.. Methods, did you noticed that? ( just to let you know i used the 3 parallels 1.8K you sent me for 100V.)

This concerns me. Please measure the temperature though before getting too worried... The three 1.8K resistors are the correct resistors and they should be good to 107V. My guess is that you will find that the surface mount trio are getting very hot - but not so hot as to de-solder. We can look up the max rated temperature. Worst case we can replace that trio with a single power resistor.

Method.. I remember you said that when you have set the shunt to 250mohm, the limit of the current was the too small wires of the motor?.. can you confirm ?.. I want to set it to 100A first.... then i'll see how well it work before to increase current. ( I have acces to mant thermal IR camera at work.. I may examine that controlerl with thermal imaging to locate potential over heat problem... like mosfet not correctly thightened that overheat.. etc

Yes, my 100A current limit was certainly governed (to some degree) by the 16AWG wires. Please test at your current shunt resistance and do not proceed directly to 0.250m. Gary has a controller with 0.250mOhms. Maybe he will test this weekend and give us an idea of how much current this will allow.

Anyhow, thanks for the feedback on the 12V regulator circuit. This is party why I have been delaying on releasing kits to the general public. I want experienced guys like you and Gary to gracefully run into any possible bumps in the road that may appear.

Keep us updated.

-methods

OK... So I tested the voltage on the trio resistor with 75V input on the controller

I got 60V on the 3 serie resistor wich give 3kohm and 20mA cuurrent.. that's 1.2W dissipation on 3 smt resistor at just 75V input!

I have in the package you sent me, a pair of 1.5K resistor... guess what :wink: .. i'll put them in serie and will replace the 3 smt with this set.. it should be fine!

this is very impressive how i installed these 5x 160V low ESR 1000uF caps on that board !!.. that will certainly help!!

I also have 8 gauge power input.. the max size foldable between the free area of on the board.. !!

As usuall.. "few" pics for you guys :wink:

Doc

 

[SAM-Pilot]

I am reading here since months and I love this forum. Now I decided to be an active part of endless-sphere and write some posts.

Doctorbass wrote:
I got 60V on the 3 serie resistor wich give 3kohm and 20mA cuurrent.. that's 1.2W dissipation on 3 smt resistor at just 75V input!

The 3 SMD 1Kohm resistors are not dimensioned for 100V. If you have 100V on the input heat dissipation is
(100V-12V)^2/3000Ohm=2,6Watt which means 0,9Watt each. This size of SMD resistors can handle only 0,5Watt.
So these SMD 1Kohm resistors will work till 75V battery voltage. May be this is why the max regen voltage is also limited to 75V.
But we will have a solution, soon... :wink:

 

[dnmun]

you will need more than the 600R for the input power resistor if you are up at 107V. you should have 3 2k4R resistors in parallel to get the input to the voltage regulator down close to 52V to keep it in spec. assuming 65mA then 107-(800x.065)= 57V.

if you start at 3.2V for 32S lifepo4, and use 3 2k2R, 102.4-(734x.065)=54.7V input to the regulator.

same voltage as which it did when i used the 540R input resistance for the 72V 9FET infineon upgrade. i actually pulled the 12V rail up to 19V when running at 86V then. 5V rail was 6V too. but it ran ok.

 

[Jeremy Harris]

Just a thought on the power/heat dissipation problem when running at ~100V or so. Has anyone looked at just stripping the board out of a small plug-in switched mode appliance power supply/charger and using that as a low voltage regulator?

I did this for a power supply for another project a while ago and was surprised at how compact the little switched mode supply board inside the plug-in case was. I reckon it would probably fit where the power resistors are.

The advantage would be much less heat, as these switched mode supplies are very efficient. Most will run from about 60-80V minimum, up to around 250V maximum. Might get around a lot of the heat problems if you can find one that is readily available with the right sort of output voltage.

Jeremy

 

[liveforphysics]

I agree. This resistor nonsense seems like the way to come up with control power voltage without spending more than $0.02.

My brain won't let me believe that there is not an elegant simple single IC bucking regulator that could take ~40-150vdc in, and output a steady 12v or whatever the circuit needs to see. It's not like we are looking for a lot of current on the control power.

 

[julesa]

Yeah philf designed one about the size of a quarter here:

http://www.endless-sphere.com/forums/viewtopic.php?f=2&t=8457&st=0&sk=t&sd=a&start=46

 

[dnmun]

methods can measure the rails when he puts it together, i suspect it will run up there, and the other side is that the 800 ohms eats up 52V under high load so you would have the LVC up at 66V and could not run a 24S lifepo4 to the end of capacity.

so to get the LVC down to 52V he has to use 600R.

the shenzen had a third step down regulator on top of the 12V regulator.but this board doesn't have holes for another one like the shenzen pcb.

 

[Doctorbass]

I agree for the switching.. but generally they are more fragile than linear circuit..

For my concern, loosing 10W in heat for a drag race is not so important from now :wink: .. i'll see if i need to save few 1/100sec in the 1/4 mile !

But that work well now since i replaced the little 3 serie smt rsistor for big 3W flameproof resistors.

 

[Doctorbass]

About the current limit, I think there is another way than soldering the shunt to modify the current limit..

The Isense circuit is simple and interact with the Uc. The circuit have few resistor that can be modified to re-adjust the amplification factor from the Isense..... just like we did 2 years ago with fechter on the old V1 crystalyte controller...

Here is some schematic i did to help to understand it...

Fechter, Camlight, liveforphysic Gary ?.. any opinion?

Doc

 

[Doctorbass]

Here is the old crystalyte curent control ( using an op amp)

Similar to the infineon desing

According to it i guess that on the infineon we just need to replace the 10K resistor with a 20K potentiometer?

Doc

 

[methods]

I believe this 12V regulator is in a slightly different configuration than the version found on your 9 fet board.
I am fairly certain that it requires different resistor configurations.
I could be mistaken. I believe that SAM-Pilot has schematics that he pieced together.

-methods

you will need more than the 600R for the input power resistor if you are up at 107V. you should have 3 2k4R resistors in parallel to get the input to the voltage regulator down close to 52V to keep it in spec. assuming 65mA then 107-(800x.065)= 57V.

if you start at 3.2V for 32S lifepo4, and use 3 2k2R, 102.4-(734x.065)=54.7V input to the regulator.

same voltage as which it did when i used the 540R input resistance for the 72V 9FET infineon upgrade. i actually pulled the 12V rail up to 19V when running at 86V then. 5V rail was 6V too. but it ran ok.

 

[methods]

Doc - nice overkill! I am sure that board will put out some serious power.

Too bad you replaced the 3 SMT's before we got a chance to test the heat...
Maybe I can power up a board later and do a test.
I bet they dont get over 300F


SMT's and rated value -
I understand that the SMT's are not "rated" to do 100V in this config but is ANYTHING rated for what we do with it?
Running 7.5KW into a 750W motor is certainly not "rated" :wink:
Running 10KW out of a 1.2KW controller is not rated
This is certainly not a "conservative" build This is the thread for people who are insane and want to shoot fire out of the sides of their bikes
So lets figure out what we can get away with

- Switcher for the 12V-

I looked into doing a switcher. The problem is finding one that can run from around 30V to 100V. That is a very large input and if bought as a package, will probably cost $30. Agreed that if one wanted to (Philf) one could build the circuit for much less but times is money. If I spend 20 minutes building each switcher (+pcb) the prices of the controllers go up $20.

-methods

 

[geoff57]

hi
R 44 was always the amp limit controling factor when we were working out how the 12 fets work but when the computer program came online we stoped further investigation.

Geoff

 

[Doctorbass]

hi
R 44 was always the amp limit controling factor when we were working out how the 12 fets work but when the computer program came online we stoped further investigation.

Geoff

Yess!.. excellent!.. I've also thought that this 10K R44 resistor can adjust the current limit! :wink:

I think that by adding two more shunt like i did.. for a total of 6 would solve the heating problem for up to 150A

0.793mohm * 150A^2 = 17.8W heat..... at 150A.. but we all know that this 150W is certainly not constant and continuous! so i believe that calculating 100A avg for hard use is ok.. so the heat drop to 7.93W total that seems alot better for 6 shunt !.. close to 1.33W per shunt is not too high...

So 6 shunt + 20K pot instead of R44 should be the key!.. but to avoid too high value with the pot adjusted on the extreme we may need to ass a resistor before the pot and one after to set the max limit even though the pot or min or min..

My controller is ready.. i just did not make the pot current limit since i want to know what will be the actual limit on that 0.793 mohm with the original 10K R44 before to mod it to set the min and max serie resistor linked to the pot..

Doc

 

[geoff57]

So 6 shunt + 20K pot instead of R44 should be the key!.. but to avoid too high value with the pot adjusted on the extreme we may need to ass a resistor before the pot and one after to set the max limit even though the pot or min or min..

My controller is ready.. i just did not make the pot current limit since i want to know what will be the actual limit on that 0.793 mohm with the original 10K R44 before to mod it to set the min and max serie resistor linked to the pot..

Doc

good idea
a control is vital if you are going to know what value the amp limmit is, also when the program comes online you will need to know what to use as a multiplier for the amp limit value setting to get a true value.
When the early 12 fet boards were sent out from china I had to calibrate (find the multiplier) them as the shunt values were all over the place and the program was expecting a set value these was for Team Hybrid.

Geoff

 

[Doctorbass]

Something VERY strange between the crystalyte and the Infineon is that the C2_1 capacitor close to the 78L05 regulator (the only small cap that is rated 100V) have 10x factor uF value between the crystalyte and the infineon!!

On the crystalyte the value is 10uF 100V.. and the infineon have 100uF 100V !!.. I just come from the electronic part store to buy a 10uF 160V that i checked on the crystalyte... and when i came back to replace the same infineon caps.. I discuvered that it's 100uF on it ... i'll need to go back and buy another one... lol...

i'll also replace the actual 3x 1.8K 3W resistor with 2.2K.. that will make the regulator current to be lowered and make it safer.. since we use this beast controller for future ultra high power multi kW ride, we must make sure that nothing can be weak or too fragile!.. 150A 100V blast make too much dammage!

Doc

 

[Doctorbass]

So 6 shunt + 20K pot instead of R44 should be the key!.. but to avoid too high value with the pot adjusted on the extreme we may need to ass a resistor before the pot and one after to set the max limit even though the pot or min or min..

My controller is ready.. i just did not make the pot current limit since i want to know what will be the actual limit on that 0.793 mohm with the original 10K R44 before to mod it to set the min and max serie resistor linked to the pot..

Doc

good idea
a control is vital if you are going to know what value the amp limmit is, also when the program comes online you will need to know what to use as a multiplier for the amp limit value setting to get a true value.
When the early 12 fet boards were sent out from china I had to calibrate (find the multiplier) them as the shunt values were all over the place and the program was expecting a set value these was for Team Hybrid.

Geoff

Geoff, where can we connect the famous disgnostic LED ?. cause i connected the + of the led to the LED true hole and the neg to the gnd it not blink.. the led is always off...

can you specify please?

 

[BatteryMooch]

I agree. This resistor nonsense seems like the way to come up with control power voltage without spending more than $0.02.

My brain won't let me believe that there is not an elegant simple single IC bucking regulator that could take ~40-150vdc in, and output a steady 12v or whatever the circuit needs to see. It's not like we are looking for a lot of current on the control power.

How about the Texas Instrument's three-terminal adjustable linear regulator, the TL783?
Good to 125VDC input, 700mA output,and less than $2 at Digikey.
Just add a 1uF input cap and a 10uF output cap, 2 resistors, and you're done!

 

[steveo]

I observed one imnportant thing that may fail on the board. The 3 smt 1Kohm resistor in serie near the 3 big resistor of the regulator become VERY hot at 100V.. I mean.. very very hot!.. I thing that they may unsolder if that is not corrected... is it normal?... cause smt component can not take so much heat!!.. I guess not.. Methods, did you noticed that? ( just to let you know i used the 3 parallels 1.8K you sent me for 100V.)

This concerns me. Please measure the temperature though before getting too worried... The three 1.8K resistors are the correct resistors and they should be good to 107V. My guess is that you will find that the surface mount trio are getting very hot - but not so hot as to de-solder. We can look up the max rated temperature. Worst case we can replace that trio with a single power resistor.

Method.. I remember you said that when you have set the shunt to 250mohm, the limit of the current was the too small wires of the motor?.. can you confirm ?.. I want to set it to 100A first.... then i'll see how well it work before to increase current. ( I have acces to mant thermal IR camera at work.. I may examine that controlerl with thermal imaging to locate potential over heat problem... like mosfet not correctly thightened that overheat.. etc

Yes, my 100A current limit was certainly governed (to some degree) by the 16AWG wires. Please test at your current shunt resistance and do not proceed directly to 0.250m. Gary has a controller with 0.250mOhms. Maybe he will test this weekend and give us an idea of how much current this will allow.

Anyhow, thanks for the feedback on the 12V regulator circuit. This is party why I have been delaying on releasing kits to the general public. I want experienced guys like you and Gary to gracefully run into any possible bumps in the road that may appear.

Keep us updated.

-methods

OK... So I tested the voltage on the trio resistor with 75V input on the controller

I got 60V on the 3 serie resistor wich give 3kohm and 20mA cuurrent.. that's 1.2W dissipation on 3 smt resistor at just 75V input!

I have in the package you sent me, a pair of 1.5K resistor... guess what :wink: .. i'll put them in serie and will replace the 3 smt with this set.. it should be fine!

this is very impressive how i installed these 5x 160V low ESR 1000uF caps on that board !!.. that will certainly help!!

I also have 8 gauge power input.. the max size foldable between the free area of on the board.. !!

As usuall.. "few" pics for you guys :wink:

Doc

Hey Doc,

I love the work you did on that controller.. Its starting to remind me of my monster irfb4568 125amp controller !!. I like how you use the bits of copper for the capasitors .. I was going to do that but didn't .. I haven't had any problems at all since i used copper strands to link each fet leg to the controller..

I can't wait to hear her pump out 150 amps!!! .. I'm very curious to see how much heat it will put out!

I envy your copper bus bar lol

P.s. When will we see a test drag run!!
-steveo