was going through the doc from semelab
http://products.semelab-tt.com/pdf/ApplicationNoteAlfet.pdf
on the page 3 its given as
>
No requirement for source resistors for
single device or parallel device operation
>
so when we parallel generally we need source resistors. So can we completely eliminate the source resistors?
If so how do we measure the bias current ? generally I measure the bias by putting multimeter across the source resistor and measure the voltage then calculate the current through it.
how to measure the bias in the case?
http://products.semelab-tt.com/pdf/ApplicationNoteAlfet.pdf
on the page 3 its given as
>
No requirement for source resistors for
single device or parallel device operation
>
so when we parallel generally we need source resistors. So can we completely eliminate the source resistors?
If so how do we measure the bias current ? generally I measure the bias by putting multimeter across the source resistor and measure the voltage then calculate the current through it.
how to measure the bias in the case?
Pull the DC rail fuse if there is one and use your DMM to measure the total current for that channel, then divide by the number of paralled P or N devices. If you don't have rail fuses then you'll have to be creative in making this connection.
Source resistors are not needed with lateral mosfets. I have tried adding them but found it sounded better without them.
If paralling I use mosfets from the same batch number. I do check the switch on voltage of each mosfet but my experience is that lateral mosfets from one batch have similar switch on voltages and it is probably ok to use mosfets from the same batch without testing the switch on voltage.
I run a separate power supply to the mosfet stage from the amplifier power supply. To check the bias I remove the fuse and check with a multimeter on amp setting.
As a final check I test each mosfet temperature when the amp has been running for a short time. Switch the amplifier off and check the temperarure of each mosfet. They should be similar if sharing current.
Don
If paralling I use mosfets from the same batch number. I do check the switch on voltage of each mosfet but my experience is that lateral mosfets from one batch have similar switch on voltages and it is probably ok to use mosfets from the same batch without testing the switch on voltage.
I run a separate power supply to the mosfet stage from the amplifier power supply. To check the bias I remove the fuse and check with a multimeter on amp setting.
As a final check I test each mosfet temperature when the amp has been running for a short time. Switch the amplifier off and check the temperarure of each mosfet. They should be similar if sharing current.
Don
Yes one of the PSU fuses, there should be two per channel one for pos rail and one for neg. Not to be confused with the main AC fuse(s). This is the procedure that many commercial paralleled lat mosfet amps use. Do a google search for one of the Hafler amps, i.e. XL-280 and read the biasing procedure and it might make more sense.
regarding bias setting got it thank you very much.
So when paralleling when source resistors are not used then does the current sharing is balanced by itself?
So when paralleling when source resistors are not used then does the current sharing is balanced by itself?
In theory yes.
Paralleled pairs still have to be closely matched. Because laterals have a pos temp coefficient if one of the paralleled devices is passing more current it should heat up more than the others increasing the internal resistance of that device and self-limiting. According to info from others on this board there will still be imbalances between matched paralleled devices.
Last edited:
from 2007
The device temperature MUST be DIFFERENT if the current is different. That puts the devices into different parts of their operating region.
Source resistors help to maintain near equal currents and that in turn helps to maintain near equal temperatures.
This was with selected Vgs devices AND using Source resistors and still there is imbalance in device current and device temperature.
The device temperature MUST be DIFFERENT if the current is different. That puts the devices into different parts of their operating region.
Source resistors help to maintain near equal currents and that in turn helps to maintain near equal temperatures.
Since transconductance of typical lateral power mosfet is roughly 1S, each mV disbalance in Vgs results in roughly 1mA disbalance in Iq.
Mounting each MOSFET on separate heatsink could possibly facilitate more even current distribution.
Id vs Vgs curve in datasheet suggests that thermal self regulation works better for higher quiescent currents (above 0.1-0.2A per mosfet).
Mounting each MOSFET on separate heatsink could possibly facilitate more even current distribution.
Id vs Vgs curve in datasheet suggests that thermal self regulation works better for higher quiescent currents (above 0.1-0.2A per mosfet).
- Status
- Not open for further replies.