I am just finishing up my F5 rebuild after blowing my IRF outputs after 3+ years of trouble free service. Please don't ask how I blew the mosfets, it was an error on my part 
I decided to use Toshiba 2SK1530/2SJ201 outputs this time around. When I set the bias to to 0.6 volts the outputs are barely warm to the touch. I was hoping someone could advise where I went wrong on this rebuild. Of course when I built the my first F5 I had no issues on power up (beginners luck I guess).
I decided to use Toshiba 2SK1530/2SJ201 outputs this time around. When I set the bias to to 0.6 volts the outputs are barely warm to the touch. I was hoping someone could advise where I went wrong on this rebuild. Of course when I built the my first F5 I had no issues on power up (beginners luck I guess).
These MOSFETs turn "on" at about half the voltage of the IRF (now Vishay) or Fairchild MOSFETs (same with the Renesas lateral mosfets). Before you solder them in make sure that the potentiometers will have a really low Vgs voltage -- and remember that the P-channel device Source is connected to the positive rail and vice versa.
Thank you, I rechecked my boards aginst the BOM and didn't find any mistakes. The pots were at 0 ohms before I started, then gradually I increased resistance to obtain .3, .4, .6 volts across R11 and R12 then I backed off. The mosfets did not produce significant heat ~ 90 degrees only. I did not test the outputs before installing them. Still baffled.
OK. Will check those voltages on outputs at 0.3 volts when I can get to it. I am using alum. Oxide insulators and the outputs are electrically isolated. I will probably try to replace the Toshiba's with the irfs cause I have a spare pair that I know are good. Then will try again to bias and see how it goes from there.
ok great - time to measure the voltages as suggested in post #5. If you have 0.6 volts across the source resistors and those are 0.6ohms then you have 1 amp flowing through them and presumably thru the fet.
Since your fets are not heating up with this much juice flowing through, we have to measure their Vds to try and understand why.....
Since your fets are not heating up with this much juice flowing through, we have to measure their Vds to try and understand why.....
Success!! I thought a little more deeply about this build verses my previous F-5 build. I used the same type of board (CViller V-2-had an extra set). I compared the old boards with the new populated board and the components were all in order and matched the old board. The only difference on the new build were different output FETs and the use Artic Silver instead of silicone grease. I was pretty sloppy with the application of Artic Silver and when I pulled the FETs off of the heatsink to check there was a glob of the Artic that had squeezed out on the side of the FET and was in contact with the heatsink. Also, the Artic was on the screw that held the FET in place. I had checked for short between the drain and heatsink before I pulled off the FET and my meter didn't read a short. Hmm? I cleaned off the Artic Silver and used silicone grease and Alum oxide insulators and replaced the FETs, fired up one channel and the die of the FET measured 49 degress centigrade after about 1/2 hour or so. D/C offset 4 mV at .595V across R11 and 12.
I don't want to "diss" Artic Silver and it is not supposed to be electrically conductive, but I won't use it again for this type of application. Thanks guys for helping me with this. Greatly appreciated.
I don't want to "diss" Artic Silver and it is not supposed to be electrically conductive, but I won't use it again for this type of application. Thanks guys for helping me with this. Greatly appreciated.
I know this is an old thread, but it is worth noting that several Arctic Silver products are electrically conductive, although the company makes other greases which are not supposed to be. Which one did you use?
Also, how do you like the F5 with the Toshiba FETs in place? I am in the parts-gathering stage of my F5 build and am trying to decide between the usual IRF parts and these Toshibas on ~16V rails with 1A bias or so.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.