• Howdy! we're looking for donations to finish custom knowledgebase software for this forum. Please see our Funding drive thread

Extreme moded 18 Fet Lyen Controller

markcycle

10 kW
Joined
May 14, 2007
Messages
528
Location
Farmingdale, NY
I been working with the 18 FET controller to see how far I can take it
My goals are to remove current from the PC board. I didn't want to beef up the traces as I feel this is both labor intensive and prone to solder shorts. I utilized the Tabs of the FETs where I could instead of the leads, this allowed me to greatly reduce the heat load in the PC board without much labor
I also added addition decoupling caps

D336079_7141548_605574


Battery voltage connected to tabs
D336079_7141548_605552

To my surprise when I measured the current in the red wire to the tabs, almost all the B+ current is in the Red wire
Added Caps
D336079_7141548_605577


Opinions and technical review welcome
 
:shock: :shock: :shock:

I love you Mark!!..

man.. this work is al la kelly method right?.. i mena with the bus bar.. :mrgreen:

please keep us informed about your test and how hot it become at different current.. i guess to preserve it you'll stay in the 80% of max fet voltage region right?

Does the heat sink is connected to the 3 x 3 fets directly and is one of the battery connection by the same time?

Doc
 
Ypedal said:
Very nice !!!!! 8)

So har far have you taken it so far ? ( Voltage, and amps.. )

I have built a few controller this way ones with 4110 Fets and 4115 fets
As far as Dyno testing 10KW 100 volts 100 amps continuous. On the road I set one for a block time of 15 sec and on this unit I recorded peaks of 150 amps but that is with a 72 volt LIFEPO4 pack

I hope to get this unit in my motorcycle which fully charged sits at 114 volts
I have two units in the hands of customers but I've haven't gotten back any data

Thermally they are performing, well I'll have hard data soon
 
That's pretty sweet.

Subscribed.
 
I want to show how I connected the B+ to the tabs of the FETs from the first set of pictures it looks like I soldered the wire to the tabs Not so see picture

D336079_7080457_19629


The terminals go up against the tabs if you try this be careful, use ring lugs large enough to allow the shoulder of the plastic insulator to fit through the hole of the ring terminal
 
i thought that remote mounting of additional caps at the end of a wire is a big no-no.
don't the caps combined with the stray inductance from long leads create a tank circuit?
i guess if the fets don't turn into plasma then the proof is in the eating.
have to wait & see how it holds up long term.
 
Toorbough ULL-Zeveigh said:
i thought that remote mounting of additional caps at the end of a wire is a big no-no.
don't the caps combined with the stray inductance from long leads create a tank circuit?
i guess if the fets don't turn into plasma then the proof is in the eating.
have to wait & see how it holds up long term.

That would be true if they were only at the end, that's why I mount them all along the wire path. Otherwise you would be correct.
 
I'd still be adding some MLCC's or something - those long leads make effective inductors.

Have you scoped the output? Fortunatly the gate drive on these controllers is pretty wimpy, so you shouldn't get extremely fast rise times.
 
heathyoung said:
I'd still be adding some MLCC's or something - those long leads make effective inductors.

Have you scoped the output? Fortunatly the gate drive on these controllers is pretty wimpy, so you shouldn't get extremely fast rise times.
Can you reference which leads you are referencing. I didn't remove any of the original caps
I only added caps and conductors. So I bring the power supply to the tab but all decoupling
happens at the fets leads.
I should add high freq caps mica or polyproplene caps to surpress high freq noise. I've built six units so far
with lower fet temps and most important much less heating in the PC board when running the controller
at 70 amps continuious (battery current). Today I ran repeated peaks of 150 amps duty cycle of 150 amps 10 sec
70 amps 45 sec @ 96 volts. Did this for a half hour.
I want to make this design as good as possible so please let me known where you would add caps and how you would expect it to improve the
design. I did scope the output I'll try to do it again and post pictures.
 
I know this thread has been dormant awhile, but I was hoping the pictures are still available somewhere. It would be much appreciated.
 
Back
Top