Hi all
I don't know if it is necessary for a mercury relay, unless you want to really test very short pulses?
Most SOA curves follow "DC" or 1 S tests then increase in dissipation for 1 mS...
I think that a transistor timer stage should be enough for pulse durations to 10uS or less if high speed circuit design is used.
And since the curent source transistors might take 1 uS to turn on maybe SOA to 10 uS is enough.
From what I have heard about fakes it seems to me that DC or 1S testing will determine how good they are ... as long as the crowbar turns the TUT off when it fails ...
cheers
John
I don't know if it is necessary for a mercury relay, unless you want to really test very short pulses?
Most SOA curves follow "DC" or 1 S tests then increase in dissipation for 1 mS...
I think that a transistor timer stage should be enough for pulse durations to 10uS or less if high speed circuit design is used.
And since the curent source transistors might take 1 uS to turn on maybe SOA to 10 uS is enough.
From what I have heard about fakes it seems to me that DC or 1S testing will determine how good they are ... as long as the crowbar turns the TUT off when it fails ...
cheers
John
another SOA tester paper from Motorola
http://www.datasheetarchive.com/pdf/685820.pdf
I think now the test time of SOA tester is 1mS, 100ms,1S....not less than 100ns.
but the tr(raising time) of test pulse is very short ,for example 50ns or less.
http://www.datasheetarchive.com/pdf/685820.pdf
I think now the test time of SOA tester is 1mS, 100ms,1S....not less than 100ns.
but the tr(raising time) of test pulse is very short ,for example 50ns or less.
All sounds pretty good to me 
I have some experience in my day job testing FETs. I havent done every test on the data sheet but I've been able to accomplish every measurement asked of me so far. I use nothing fancy but a suitable selection of lab quality bench equipment and a few custom made circuits and fixtures.
I think the idea of setting up these tests first with minamul equipment, evaluate the results, determine what will be necessary with a compresive set of prototype circuits and set ups, before building it After some experience you will know better what you need.
I test devices with BVds up to 30V and Id cont. to 60A. That sounds easy
But, take the SOA curve. The spec on my main supply is 40V 50A. This supply will support the "DC" line of that set of curves for all but our biggest device, but what about the 1, .1, 10mS, 1mS, etc tests ? The current I need to obtain for pulsed testing is 10X the Id Cont. The voltage on the supply is sufficient but how regulated is it during pulsed type testing? Truth is, I'm lucky enough to have a halfway descent supply. It will deliver about 180A for 100uS. I'm fairly happy with that When I go to do an Rdson test though, my supply dumps about 5V momentarily. The test must be done in less than 500uS or the device begins heating. I added about 50,000uF to the Test circuit and I only loose about 300mV when switching on a 30A MOSFET And then, there is how you connect to your D.U.T. to acheive accurate consistant results. My typical parameters are sometimes difficult to deal with, espesially because I deal with bumped die style smd devices.
I have some experience in my day job testing FETs. I havent done every test on the data sheet but I've been able to accomplish every measurement asked of me so far. I use nothing fancy but a suitable selection of lab quality bench equipment and a few custom made circuits and fixtures. I think the idea of setting up these tests first with minamul equipment, evaluate the results, determine what will be necessary with a compresive set of prototype circuits and set ups, before building it
After some experience you will know better what you need. I test devices with BVds up to 30V and Id cont. to 60A. That sounds easy
But, take the SOA curve. The spec on my main supply is 40V 50A. This supply will support the "DC" line of that set of curves for all but our biggest device, but what about the 1, .1, 10mS, 1mS, etc tests ? The current I need to obtain for pulsed testing is 10X the Id Cont. The voltage on the supply is sufficient but how regulated is it during pulsed type testing? Truth is, I'm lucky enough to have a halfway descent supply. It will deliver about 180A for 100uS. I'm fairly happy with that When I go to do an Rdson test though, my supply dumps about 5V momentarily. The test must be done in less than 500uS or the device begins heating. I added about 50,000uF to the Test circuit and I only loose about 300mV when switching on a 30A MOSFET And then, there is how you connect to your D.U.T. to acheive accurate consistant results. My typical parameters are sometimes difficult to deal with, espesially because I deal with bumped die style smd devices.
Hi Ashok,
I will try to get back to this tread.
Interest has been shifted away from designing and building testgear to repair of musical instrument amplifiers.
There hasn't been much time to build any thing.
Ideas enough but hands and time to short.
And the community is in need of a helping hand; so that's me then...
Cheers,
Tarzan
I will try to get back to this tread.
Interest has been shifted away from designing and building testgear to repair of musical instrument amplifiers.
There hasn't been much time to build any thing.
Ideas enough but hands and time to short.
And the community is in need of a helping hand; so that's me then...
Cheers,
Tarzan
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.