i am working on another jbl bp1200.1 . this one is number 6 and i am getting a mountain of old mosfets. irf3205,irf9640 and irf640. most all of them are between .5 and .9 on my diode check. i am wondering if i can just put this amp back together with my old transistors? if i did how much is recomended tolerenece between them?
also there are 4 other transistors labeled MUR3040. they are bigger then the rest and i am wondering if it is good practice to pull these along with the rest?
also there are 4 other transistors labeled MUR3040. they are bigger then the rest and i am wondering if it is good practice to pull these along with the rest?
The IRF3205 will be used in the power supply, these are very high current (60 Amp+) low RDS on mosfets (55 - 60V) with a lowish breakdown voltage, these will be switching the battery voltage into the transformer
The IRF640 is a 200V N channel mosfet, this will be part of the amplifier. Likewise with the IRF9640 which is the P channel compliment to the IRF640 (still 200V). These are probably about 10 Amp.
The MUR3040 are not mosfets at all but high speed high current rectifiers that are part of the power supply. These rectify the output of the transformer.
You can't stuff transistors in where mosfets are used without a massive redesign of the circuit
The IRF640 is a 200V N channel mosfet, this will be part of the amplifier. Likewise with the IRF9640 which is the P channel compliment to the IRF640 (still 200V). These are probably about 10 Amp.
The MUR3040 are not mosfets at all but high speed high current rectifiers that are part of the power supply. These rectify the output of the transformer.
You can't stuff transistors in where mosfets are used without a massive redesign of the circuit
Last edited:
If the parts are original and weren't defective, you can reinstall them. If you're installing new outputs, you should only use components with the same date code (640s match and 9640s match). Typically, components with the same date code are matched closely enough.
Typically, the 9640s will read ~0.9v and the 640s will read ~0.6v.
Typically, the 9640s will read ~0.9v and the 640s will read ~0.6v.
well just all my 3205s seem to measure up into 3 groups that vary by about +-.002 and the 3 groups are .506 .515 and .535. i am wondering if this is normal and if they can work together in harmony? what is normal variance?
the date codes are all over the place and are all different. i know they should all be the same but doesnt measuring them give me an accurate description of the performance of the transistor?
the date codes are all over the place and are all different. i know they should all be the same but doesnt measuring them give me an accurate description of the performance of the transistor?
i didnt buy them they are the mountain i am ammasing from the other amps identical to this one that i am fixing. when i test all the transistors of the same part number i am finding that i usually only have 1 or 2 that are bad of any given part number. i am hoping i can just replace the 1 or 2 bad ones with ones that measure similar to the ones that have not failed.
i have a bag of all new ones but this amp is just for me and im not terribly concerned about the longevity of it. it is more for the fact that it is fun for me to frankenstein a broken amp with broken parts and make it work.
im not sure what this diode check/test is telling me but i thought it was proof positive that i was finding a correct match of 2 transistors with different date codes.
i have a bag of all new ones but this amp is just for me and im not terribly concerned about the longevity of it. it is more for the fact that it is fun for me to frankenstein a broken amp with broken parts and make it work.
im not sure what this diode check/test is telling me but i thought it was proof positive that i was finding a correct match of 2 transistors with different date codes.
If you're not concerned with reliability, you can use whatever you have.
If you can find 2 groups of 5 transistors with matching date codes, install all of the ones that match on the same side. Although it's probably better if both sides match, it's not nearly as critical as having matching FETs in each bank of power supply transistors.
If you can find 2 groups of 5 transistors with matching date codes, install all of the ones that match on the same side. Although it's probably better if both sides match, it's not nearly as critical as having matching FETs in each bank of power supply transistors.
You can't test mosfets with diode check like a transistor. All you see is the body drain diode which is there to prevent reverse polarity dumps frying the device. You need to test drain-source resistance (should be very high with gate shorted to source) and gate-source resistance (should be infinite). Also work on a grounded ESD mat and ground yourself, mosfet gates fry easily.
is this test an accurate way to test these things or is it just a way to find ones that are shorted out?
with this test what is an acceptable varience between transistors? or does it even matter?
aside from date codes is there a test to match these things together? or does it require lab equipment
what has happened is i keep ordering new transistors and everytime i order i get a couple extras, now i have a big bag of unused ones but none of them are matching date codes. does this mean i can never use them?
i was hoping to use all these odds and ends on this amp.
with this test what is an acceptable varience between transistors? or does it even matter?
aside from date codes is there a test to match these things together? or does it require lab equipment
what has happened is i keep ordering new transistors and everytime i order i get a couple extras, now i have a big bag of unused ones but none of them are matching date codes. does this mean i can never use them?
i was hoping to use all these odds and ends on this amp.
The multimeter can tell you with ~99+% reliability if the FET has failed. It can't tell you much more than that.
Some amps have only two FETs in the supply. You can use them in those amps. You can use them in amps for testing to determine if there are other problems. If an amp is yours (not for a paying customer, not to be sold to an unsuspecting buyer), you can use as many mismatched parts as you'd like but expect reliability to suffer if you are abusive or drive the amp to the limits (especially if the amp has a weak power supply to start with).
It's not going to bring about the end of the world as we know it if you use mismatched parts but it's not good practice for someone who is being paid for the repair work or intends to sell to amp to someone who expects to get an amp in perfect working order.
Some amps have only two FETs in the supply. You can use them in those amps. You can use them in amps for testing to determine if there are other problems. If an amp is yours (not for a paying customer, not to be sold to an unsuspecting buyer), you can use as many mismatched parts as you'd like but expect reliability to suffer if you are abusive or drive the amp to the limits (especially if the amp has a weak power supply to start with).
It's not going to bring about the end of the world as we know it if you use mismatched parts but it's not good practice for someone who is being paid for the repair work or intends to sell to amp to someone who expects to get an amp in perfect working order.
ok thats great news. im still just learning and i wouldnt trust my repair work to fix for other people any time soon.
can i use this test to "match up" FETs?
is it normal for them to vary from .502-.537? does it matter? or just as long as they are not shorted out? what is the allowable varience and does it matter?
can i use this test to "match up" FETs?
is it normal for them to vary from .502-.537? does it matter? or just as long as they are not shorted out? what is the allowable varience and does it matter?
I don't think the multimeter on diode check is going to provide any useful information regarding the matching of FETs.
FETs with the same date code are likely to be virtually identical in every way. When semiconductors change in temperature, their electrical properties change (sometimes, significantly). FETs from the same production lot will be more likely to track closely (their operating parameters/characteristics will change in the same way) as the temperature changes. This will keep them closely matched under normal operating conditions. The electrical characteristics of mismatched components may change in different ways which could cause some to take more of a load than others at different temperatures. This is only one way that mismatched, parallel components can cause problems.
Again, if these are your amps and you don't intend to sell them, use whatever parts you want. Just don't expect reliability to be 100% of what it would be with matched parts.
FETs with the same date code are likely to be virtually identical in every way. When semiconductors change in temperature, their electrical properties change (sometimes, significantly). FETs from the same production lot will be more likely to track closely (their operating parameters/characteristics will change in the same way) as the temperature changes. This will keep them closely matched under normal operating conditions. The electrical characteristics of mismatched components may change in different ways which could cause some to take more of a load than others at different temperatures. This is only one way that mismatched, parallel components can cause problems.
Again, if these are your amps and you don't intend to sell them, use whatever parts you want. Just don't expect reliability to be 100% of what it would be with matched parts.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.