Hello
My name is Brian. I have started a Fet Matching service. I supplied FETs for the recent AlephP1.7 group buy. I currently have the following matched Fets in stock. I match using a temp. regulated heatsink assembly and use control Fets to check precision.
Test currents are regulated.
A few sets for the Aleph P 1.7 -8x IRF610 + 4X IRF9610. At least 2 pairs of 610s /set with 1mv or less match for use in the gain stage. Typically other pairs are within a few MV. Cost 25$ Can.
I have a good stock of closely matched pairs (1mv or less) of IRF 9610 for the front end of Aleph 3 etc.
Matched at 20ma, 40C temp. 6$Can/pair
I also have some IRFP240 matched at 1 amp and 40C. I have 13 pairs with a match closer than 5mv. Cost 10$Can./pair
I am located in Canada and ship worldwide. Please note that right now the Canadian $ is about .80 US dollars so prices are about 25% lower for US$.
If anyone has any requests for other devices, please let me know.
Thank You
Brian
I
My name is Brian. I have started a Fet Matching service. I supplied FETs for the recent AlephP1.7 group buy. I currently have the following matched Fets in stock. I match using a temp. regulated heatsink assembly and use control Fets to check precision.
Test currents are regulated.
A few sets for the Aleph P 1.7 -8x IRF610 + 4X IRF9610. At least 2 pairs of 610s /set with 1mv or less match for use in the gain stage. Typically other pairs are within a few MV. Cost 25$ Can.
I have a good stock of closely matched pairs (1mv or less) of IRF 9610 for the front end of Aleph 3 etc.
Matched at 20ma, 40C temp. 6$Can/pair
I also have some IRFP240 matched at 1 amp and 40C. I have 13 pairs with a match closer than 5mv. Cost 10$Can./pair
I am located in Canada and ship worldwide. Please note that right now the Canadian $ is about .80 US dollars so prices are about 25% lower for US$.
If anyone has any requests for other devices, please let me know.
Thank You
Brian
I
I am using a custom test fixture, mounted on a large temp. regulated heatsink that is temp. controlled by a PID controller. The heatsink is inside an enclosure that is also heated to the regulated temp. I use control FETs to make sure that precision is maintained and that results are reproducible at the 1mv tolerance level.
I am matching Vgs at a given current and temp. All currents are regulated. For the 1 amp tests of the large FETs I use a pulse method to obtain a VGS value for 1 amp. This allows the determination of VGS at 1amp without the temp. swings normally involved if you use a continuous current. A stable reproducible reading can be obtained this way. Measurements are made with a recently calibrated 6.5 digit DMM.
The biggest challenge to getting good precision is the control of temp. Since the Vgs will change at about 4mv/degree c. you can see that a tolerance of 1mv requires very close regulation of temp. and taking the time to let the temp. of the tested parts stabilize.
I hope this answers your questions. I have not investigated going into further testing. I may do that in the future but I need to look at the time and extra equipment needed to do it properly and how that would impact the cost.
Regards
Brian
sounds good for matching Idss.
but some devices have variations in linearity/gain and tempco. in practice this might mean that as used in a "hot" running amp there could be deviation of Idss set points for paralleled devices as the devices and heatsinks warm up. many devices seem to be remarkably well behaved and "track" each other nicely.
in practical terms I guess it may not make all that much difference, unless one is really trying to get a "gilded" amp implementation.
thanks for the rundown on your setup. btw, what do you use to generate ur pulses? And do you monitor on a scope?
but some devices have variations in linearity/gain and tempco. in practice this might mean that as used in a "hot" running amp there could be deviation of Idss set points for paralleled devices as the devices and heatsinks warm up. many devices seem to be remarkably well behaved and "track" each other nicely.
in practical terms I guess it may not make all that much difference, unless one is really trying to get a "gilded" amp implementation.
thanks for the rundown on your setup. btw, what do you use to generate ur pulses? And do you monitor on a scope?
In terms of tempo I think the difference in typical devices would be much smaller than the difference in the tempo of the thermal interface to the heatsink. The source resistors help to flatten out the small differences in VGS. I didn't mention this above, but all my matches are from the same batch code and often they are very near to each other within the batch. This should create the conditions for most of the parameters to line up close to one another. If you start trying to test for multiple parameters it gets complicated pretty quickly. If you are simply looking for low distortion you could swap them in and out of a board until finding a golden combination. I believe that is done at the Pass factory. It's not something that's easily done in my situation.
I use a fixed duty cycle gated current source to generate the pulses.
I use a fixed duty cycle gated current source to generate the pulses.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.