methods
1 GW
I repaired a Sevcon and need to re-assemble.
It appears that a green material is used on the heat sink... obviously a thermal conductor and electrical insulator.
It is soft enough that it has "smeared" a bit onto the back of the TO-220's - I thought that was very weird.
I could re-assemble with the existing insulators... and it would work... but its for a high reliability application and I dont want to risk it.
I could of course use Kapton or some other insulator... but I am hesitant to change the thermal characteristics of the controller.
In my experience I have seen:
* Gray thick woven material - in the old ebike controllers
* Kapton tape in various thicknesses
* Dry Kapton with no adhesive
This assembly is particularly gnarly because it is spring loaded with LOTS of fets... very easy to scrape a corner going in and get a floating pin-hole...
A pinhole could pass a continuity test and fail hi-pot later.
I am hesitant to hi-pot a customer controller... but we do have a 350V supply so I suppose I could do it at 150 or so.
As for a jig to load the fets... I dont have the time or resources to build a proper jig so I will probably bring them in at an angle and drop the three sets down step by step.
There are 6 spring rows to unload... and "pushing through" once the lip is clear looks dodgy at best.
-methods
It appears that a green material is used on the heat sink... obviously a thermal conductor and electrical insulator.
It is soft enough that it has "smeared" a bit onto the back of the TO-220's - I thought that was very weird.
I could re-assemble with the existing insulators... and it would work... but its for a high reliability application and I dont want to risk it.
I could of course use Kapton or some other insulator... but I am hesitant to change the thermal characteristics of the controller.
In my experience I have seen:
* Gray thick woven material - in the old ebike controllers
* Kapton tape in various thicknesses
* Dry Kapton with no adhesive
This assembly is particularly gnarly because it is spring loaded with LOTS of fets... very easy to scrape a corner going in and get a floating pin-hole...
A pinhole could pass a continuity test and fail hi-pot later.
I am hesitant to hi-pot a customer controller... but we do have a 350V supply so I suppose I could do it at 150 or so.
As for a jig to load the fets... I dont have the time or resources to build a proper jig so I will probably bring them in at an angle and drop the three sets down step by step.
There are 6 spring rows to unload... and "pushing through" once the lip is clear looks dodgy at best.
-methods