I was working on my fuel pump and had to remove the sub from the truck and shorted the sub channel on a 5 channel amp. Needless to say, I heard a boom a few seconds after turning the ignition on to check that the fuel pump was running.
I removed the amp and disassembled and found that the 2 internal fuses were blown. I replaced them and powered the amp back up only to have the fuses blow again ~5 seconds later.
My guess is I'm looking at replacing the transistors on the sub channel. Is this a good place to start? I didn't see any physical damage inside the amp without removing the board to see both sides of the transistors. I plan to give that a try tonight.
Any suggestions?
I removed the amp and disassembled and found that the 2 internal fuses were blown. I replaced them and powered the amp back up only to have the fuses blow again ~5 seconds later.
My guess is I'm looking at replacing the transistors on the sub channel. Is this a good place to start? I didn't see any physical damage inside the amp without removing the board to see both sides of the transistors. I plan to give that a try tonight.
Any suggestions?
If the output devices are MOSFET, then usually when MOSFET's blown, the driver many times becomes damaged too, so I suggest to replace driver stage together with output MOSFET's.
Ok. I'm going to pull it apart again and see if I can identify the driver stage. What type of components are the driver stages usually composed of?
Essentially, 3 types: an IC (integrated, like MICREL drivers, MICC442x as an example), a pair of PNP-NPN discrete followers, or a discrete P-N MOSFET's pair. It may also be incorporated in the chip itself like UC384x series, in the case, an internal totem pole driver.
As a rule of the trumb, if the MOSFET has a small (≈10Ω) resistor in series to the gate, and if this resistor is also blown, the drivers ARE blown too. If this R is good, drivers MAY be also blown, or not.
As a rule of the trumb, if the MOSFET has a small (≈10Ω) resistor in series to the gate, and if this resistor is also blown, the drivers ARE blown too. If this R is good, drivers MAY be also blown, or not.
Last edited:
I ohm'd them out and the 2 on the far RIGHT on the channel I shorted are both bad. What else do I need to test here and how do I go about that?
Visit here for full size pics: https://photos.google.com/share/AF1QipMrg2Z78X36RLJC1UCW5HVQxSmfGXEBkudj6F42ipaOhosjuUf33aI1xeuVa3QqzQ?key=cnRjVjZURmtVMkNYbVJuekVjczhoSHB3Rl9GMVV3
Visit here for full size pics: https://photos.google.com/share/AF1QipMrg2Z78X36RLJC1UCW5HVQxSmfGXEBkudj6F42ipaOhosjuUf33aI1xeuVa3QqzQ?key=cnRjVjZURmtVMkNYbVJuekVjczhoSHB3Rl9GMVV3
Last edited:
Another test you can easily do, is to get the failed mosfet in your table, and with a DMM, measure gate, source and drain, two per time. If gate is NOT shorted to drain (Say, more than 10KΩ), then replace the mosfets only, and try to switch the set on. In any other case, measure as follows.
If you have an oscilloscope, try to measure the output from the driver without the power MOSFET, and try to turn on the set. If you have a good square/rectangular wave, then driver is OK (Good means without strange cuts inside the cycle, good rail to rail (minimum 5Vpp)). If the driver excites the upper N MOSFET in bootstrapping, possibly no output at all in this channel.
If you have an oscilloscope, try to measure the output from the driver without the power MOSFET, and try to turn on the set. If you have a good square/rectangular wave, then driver is OK (Good means without strange cuts inside the cycle, good rail to rail (minimum 5Vpp)). If the driver excites the upper N MOSFET in bootstrapping, possibly no output at all in this channel.
Last edited:
This is beyond my current knowledge of how to do this. Can you walk me through it or point me in the correct direction?
Test a MOSFET in very easy. With a DMM in the diode position of the switch, first put negative (-) lead to drain (in most TO220/247/smd) is the center wire.
Then, positive (+) to the right lead, viewing with the tab to the table and the pins to you. It is the unique combination of a good mosfet that must give a diode drop (.5V aprox). If it gives conduction in both directions, the MOSFET is shorted, and must be discarded.
Now, put + to gate (left) and - to source (right), then the DMM will charge the gate capacity. Without as low delay as possible, move the + to to the drain, it must give low voltage readings, usually 0V or 0Ω. Now, re-do this but reversing the gate and source leads, then the negative will go to gate and positive to source, and the MOSFET must be in open circuit (>3V in the diode scale) when moving the positive to the drain and negative to source (Doing it you charged the gate negative with the DMM itself).
Some old MOSFETs have bigger resistance and may fail to give 0Ω or 0V, but in any case a low value will go. Also, some MOSFETs has lower gate leak resistances so make the movements as quickly as you can, or some charge will be lost, and you can see how the MOSFET is closing slowly as the charge is leaking, when charged positive.
I do it several times at a day, with the Fluke F87, but with other DMM may result in different grades of MOSFET opening, because it depends on MOSFET VGTH and DMM diode bias current/voltage.
It is large to explain, but with a few of practice it is very quick test, and 99.99% effective in my case.
Good luck.
Then, positive (+) to the right lead, viewing with the tab to the table and the pins to you. It is the unique combination of a good mosfet that must give a diode drop (.5V aprox). If it gives conduction in both directions, the MOSFET is shorted, and must be discarded.
Now, put + to gate (left) and - to source (right), then the DMM will charge the gate capacity. Without as low delay as possible, move the + to to the drain, it must give low voltage readings, usually 0V or 0Ω. Now, re-do this but reversing the gate and source leads, then the negative will go to gate and positive to source, and the MOSFET must be in open circuit (>3V in the diode scale) when moving the positive to the drain and negative to source (Doing it you charged the gate negative with the DMM itself).
Some old MOSFETs have bigger resistance and may fail to give 0Ω or 0V, but in any case a low value will go. Also, some MOSFETs has lower gate leak resistances so make the movements as quickly as you can, or some charge will be lost, and you can see how the MOSFET is closing slowly as the charge is leaking, when charged positive.
I do it several times at a day, with the Fluke F87, but with other DMM may result in different grades of MOSFET opening, because it depends on MOSFET VGTH and DMM diode bias current/voltage.
It is large to explain, but with a few of practice it is very quick test, and 99.99% effective in my case.
Good luck.
OK, then I suggest to do as follows. Replace all FET's and resistors damaged. Then, power the set in series in one of the leads (pos or neg), with a 12V lamp, of sufficient current (like high lights in a car, 12V 45 or 75W). If lamp brights too much when powering the set on, then there is another trouble to investigate.
If past some milliseconds, lamp goes off, and equipment looks OK, then the set is repaired.
If you have a laboratory power supply, with electronic current limited supply, then use generous current (3Amp) but do not need the lamp.
Good luck.
If past some milliseconds, lamp goes off, and equipment looks OK, then the set is repaired.
If you have a laboratory power supply, with electronic current limited supply, then use generous current (3Amp) but do not need the lamp.
Good luck.
- Status
- Not open for further replies.