DiyMobileAudio.com Car Stereo Forum banner
1 - 20 of 37 Posts

·
Registered
Joined
·
18 Posts
Discussion Starter · #1 · (Edited)
Came back from a week vacation and my Banda 7k won't turn on. Instead of powering on normally the power LED turns blue and fans begin roaring, about 2-3 seconds later I can hear the protection relay switch and a pop on the output into whatever speaker is connected. The Clip/Short light is illuminated. For what its worth the amp was seeing 15.5v lately due to cold weather (never full output at 15.5v, though).

Here's what I've checked so far:

1) Power input and remote voltages - good
2) Load presented to the output - good
3) RCA input to the amp - good
(2&3 ruled out by unplugging all wires besides power)
4) Checked internal 400a fuse - good
5) Check continuity between +12v and ground - good
(very high reading, meter displayed high number at first, then nothing, cheap meter)
6) Inspected the board, found thick dust nearby the fan and was touching a leg of a transistor, but no heat or burn marks.

Update:
7) Found a single output fet and 6 22kohm resistors in parallel dead on the board. Resistors were nearby a BPT that we think generates the PWM to drive the output stage fets. The two dead parts so far are a single output fet and the 6 resistors. Still unsure why it died, but may continue by just replacing the 8 output fets and the resistors.

Update #2
Now we're thinking that the 2010 drive IC's, 90N20D's, and the resistors will be needing replacement. A little foggy on checking the drive PWM's at this point, I need to scope the low output on the drive IC and see what I get. Might just let someone else handle it, its getting complicated.

I have a DMM, Oscilloscope, AC/DC clamp, and power supply to work with. Unfortunately no capacitance or inductance testers. If it turns out to be something requiring an entire array of transistors to be replaced I'll probably send it out, but if it's something simple, I could do it myself.

Pics of are of my initial thought, bad caps. I think they're fine, if you don't think so let me know.



 

·
Registered
Joined
·
2,441 Posts
Check the resistance between the legs of all large fets/transistors in output devices and power supply fets.

Have 3 ways to check each, pin1-2, pin2-3, pin 1-3. Looking for anything suspiciously low.. usually under 100ohms, will indicate the part is blown.

If you find one or two that read low, cross check against nearby parts with a similar label. Ie if q47 is 31ohms from pin 1 to 3, and q48 next to it is not.. q47 is likely a blown part.

Diode bridge rectifiers are typically connected pins 1-3 fyi.


If you find something, be sure to change all parts in that circuit with a matched set, use sources like digikey, mouser, etc.

If you have a bad power supply fet, I'd change them all.. usually 4-8 of them. If it's a bad part in a channel, it's usually 2 or 4 parts.

Gl
 
  • Like
Reactions: wew lad

·
Registered
Joined
·
18 Posts
Discussion Starter · #3 · (Edited)
Check the resistance between the legs of all large fets/transistors in output devices and power supply fets.

Have 3 ways to check each, pin1-2, pin2-3, pin 1-3. Looking for anything suspiciously low.. usually under 100ohms, will indicate the part is blown.

If you find one or two that read low, cross check against nearby parts with a similar label. Ie if q47 is 31ohms from pin 1 to 3, and q48 next to it is not.. q47 is likely a blown part.

Diode bridge rectifiers are typically connected pins 1-3 fyi.


If you find something, be sure to change all parts in that circuit with a matched set, use sources like digikey, mouser, etc.

If you have a bad power supply fet, I'd change them all.. usually 4-8 of them. If it's a bad part in a channel, it's usually 2 or 4 parts.

Gl
Thanks for that

So far (this is a very quick check before I head to bed) from pin 1 to 3 all the power supply transistors are a consistent 1k ohm and the outputs are pin 1 to 3 around ~21kohm EXCEPT for one which shows about 41 ohms. I believe the middle pin is the drain i.e. the output, so do you think I should be checking these values as well? For the most part they have very low resistance, as in nearly zero.

I'm going to verify with my fluke meter tomorrow. I'm new to repairing any amplifier so can anyone tell me what kind of failure this is consistent with? What else could have caused just one to blow? Should I be looking further up the line for a dirty psu square wave or what? I'd like to find the source of the problem just in case. Could a bad inductor not be smoothing the output enough? Nothing appears broken or separated from the board due to vibration but I could check.
 

·
Registered
Joined
·
18 Posts
Discussion Starter · #4 · (Edited)
If I just unsolder the one bad transistor, I imagine the amp MIGHT power on normally. I won't load it of course. Should I check the waveform with a scope? Check the power supply square wave too?

So far I believe I have isolated it to the output stage, but my biggest concern is what caused the failure. Say I go through and replace the bad transistors, and then they blow again. Do they fail sometimes due to age? It doesn't appear to be heat related but it could be, I soft clip it but I never go into a hard clip because one, it sounds bad, and two, it draws too much current. I keep the amp above 12v at all times and usually below 14v fully loaded.

Can someone confirm or deny that the power supply caps and output filtering caps look okay? More pics are no problem just ask

edit: asked a coworker, said they look fine but might be on the way out.
 

·
Registered
Joined
·
18 Posts
Discussion Starter · #5 · (Edited)
More pictures, and a new find. Looks like R74 through R79 are dead.

Does anyone know what these resistors are for? Shortly after taking this picture I tried to probe them and they literally fell off the board. I hope this is the problem? How do I fix it. I need more information. It appears the chip next to them is a bipolar transistor JD 340 that read 166k ohm on pin 1 and 3 from my meter.



Here's the output fet and test pads. Can I utilize the test, or is it only for initial testing?



What type of failure is this, is my main question.
 

·
Registered
Joined
·
2,441 Posts
The mosfets are under more stress than anything else in the amp, when something goes, it's usually them. If the protection circuitry is good enough, a bad fet in the audio section wont cause the supply fets to go too. All depends...


The resistors are 22k, looks like 6 of em.. Either 1206 or 0805 package, not sure by pic.

Here's the dpak transistor.
https://www.digikey.com/product-detail/en/on-semiconductor/MJD340T4G/MJD340T4GOSTR-ND/919476


Unless you have a diagram of what the test point are, they're not of much use..


So you think the bad fet is in the supply or power amp?

Could post a larger pic with parts of interest circled.
 
  • Like
Reactions: wew lad

·
Registered
Joined
·
18 Posts
Discussion Starter · #7 · (Edited)
The mosfets are under more stress than anything else in the amp, when something goes, it's usually them. If the protection circuitry is good enough, a bad fet in the audio section wont cause the supply fets to go too. All depends...


The resistors are 22k, looks like 6 of em.. Either 1206 or 0805 package, not sure by pic.




Unless you have a diagram of what the test point are, they're not of much use..


So you think the bad fet is in the supply or power amp?

Could post a larger pic with parts of interest circled.
Yes, I believe one of the output/audio fets is blown as it is shorted on pins 1 and 3. I posted a picture that shows another area of concern which is the resistors R74-R79. They've come off the board because of how dead they are. Appears to be 6 223 ohm or is it 22k? in parallel driving something? Should I just replace the resistors and output fet and call it a day?
 

·
Registered
Joined
·
2,441 Posts
Yes, I believe one of the output/audio fets is blown as it is shorted on pins 1 and 2. I posted a picture that shows another area of concern which is the resistors R74-R79. They've come off the board because of how dead they are. Appears to be 6 223 ohm or is it 22k? in parallel driving something? Should I just replace the resistors and output fet and call it a day?

Honestly, you could.. if you have a current limited supply to play with afterwards. If you hook it back to a battery it might just smoke again before you know something is wrong :)
 

·
Registered
Joined
·
2,441 Posts
If you take the clamp off the bad fet, can you see more lettering?
 

·
Registered
Joined
·
18 Posts
Discussion Starter · #10 ·
Honestly, you could.. if you have a current limited supply to play with afterwards. If you hook it back to a battery it might just smoke again before you know something is wrong :)

If you take the clamp off the bad fet, can you see more lettering?
Found another code IRFP90N20D, didn't even think to look there. You've done this a few times, thanks!

https://xdevs.com/doc/xDevs.com/Neutron/pdf/irfp90n20d.pdf

Misspoke and said pin 1 and 2 were shorted, it was actually 1 and 3 relating to the output fet in a previous post.

I have a psu with a 3 or 5a fuse and it puts out about 15a. would that work?

i cant imagine it needs more than a few amps when it starts.
 

·
Registered
Joined
·
2,441 Posts
Found another code IRFP90N20D, didn't even think to look there. You've done this a few times, thanks!

https://xdevs.com/doc/xDevs.com/Neutron/pdf/irfp90n20d.pdf

Misspoke and said pin 1 and 2 were shorted, it was actually 1 and 3 relating to the output fet in a previous post.

I have a psu with a 3 or 5a fuse and it puts out about 15a. would that work?

i cant imagine it needs more than a few amps when it starts.
Inrush could be up to 5-8a for a split second, but would imagine it would idle under 2A.. it's a big guy though.

How many other fets are there that are the same type?
 

·
Registered
Joined
·
18 Posts
Discussion Starter · #12 ·
Inrush could be up to 5-8a for a split second, but would imagine it would idle under 2A.. it's a big guy though.

How many other fets are there that are the same type?
Eight total, four on each side. Looks like 4 fets per rail and 4 fets per 2 inductors? Just noticed the other (good) rail has fets that appear newer with more solder as if it has been repaired but the other (bad) one hasn't. Here's a full board view:


I'll admit I'm a little confused, these are DC to DC converters, wouldn't they be for the power supply then? Or is the output pulsed DC which results in the AC music source?

Sidenote: I'm guessing it'd be a good idea to change the entire rail of fets
 

·
Registered
Joined
·
2,441 Posts
Bummer on cost, that's unusually high. My suggestion is to do all 8 at once though, like i said before.

Still thinking over those burnt resistors.....
 

·
Registered
Joined
·
18 Posts
Discussion Starter · #15 ·
https://www.digikey.com/products/en?keywords=IRFP90N20D

This guy? It's a mosfet.. if we're talking the 8 larger devices (think we are), i'm pretty sure those are the output devices.
We're on the same page, just seemed odd to me but I think I figured out why.


Bummer on cost, that's unusually high. My suggestion is to do all 8 at once though, like i said before.

Still thinking over those burnt resistors.....
$40 isn't too bad, I mean I think I can do it myself so that saves shipping and labor. The resistors are definitely getting replaced too but I just wonder why they failed.
 

·
Registered
Joined
·
2,441 Posts
Did you check resistance between pin 2/backside (same connection) to pin 1 and 3 on U7?

Those are actually 3 pin devices, not just the outer 2 pins.
 

·
Registered
Joined
·
18 Posts
Discussion Starter · #17 ·
Did you check resistance between pin 2/backside (same connection) to pin 1 and 3 on U7?

Those are actually 3 pin devices, not just the outer 2 pins.
I did see that they're 3 and checked them, they didn't seem off because I got 230kohm on left and middle, and 180kohm on 2 and 3.

Maxed out resistence on my more expensive fluke clamp meter, maybe its maxed out.

I also tried a diode check and got fairly low drop of about .7v from left to middle and left to right, 0 drop from middle to right
 

·
Registered
Joined
·
2,441 Posts
I did see that they're 3 and checked them, they didn't seem off because I got 230kohm on left and middle, and 180kohm on 2 and 3.

Maxed out resistence on my more expensive fluke clamp meter, maybe its maxed out.

I also tried a diode check and got fairly low drop of about .7v from left to middle and left to right, 0 drop from middle to right
10-4, think your resistors are 0805's in reference to the dpak next to it.

Take your pick, thick film is fine.

https://www.digikey.com/products/en...=0&page=1&quantity=0&ptm=0&fid=0&pageSize=500


Occasionally you will get burnt gate resistors near bad fets when they short out too, likely 47 or 22ohm resistors. Will want to check various resistors in the area and compare to like resistors on another fet.
 
  • Like
Reactions: wew lad

·
Registered
Joined
·
18 Posts
Discussion Starter · #19 ·
10-4, think your resistors are 0805's in reference to the dpak next to it.

Take your pick, thick film is fine.

https://www.digikey.com/products/en...=0&page=1&quantity=0&ptm=0&fid=0&pageSize=500


Occasionally you will get burnt gate resistors near bad fets when they short out too, likely 47 or 22ohm resistors. Will want to check various resistors in the area and compare to like resistors on another fet.
Perfect, that looks like the parts I'll need. I'm guessing the thermal paste should also be replaced? I'll look for something general and order a tube.

The gate resistors are something like 22k ohm for one and 22 ohm for the other.

Still unsure of what caused this. Common clipping failure or?... Since the fets are 40 bucks I want to be sure I completely resolve the problem.

Just to verify, the diode and resistance sounded correct for the BPT by the dead resistors?
 

·
Registered
Joined
·
2,441 Posts
I use this, https://www.amazon.com/gp/product/B004Z9XG0I/ref=oh_aui_search_detailpage?ie=UTF8&psc=1

No, gate resistors should be low resistance, usually 22 or 47ohm. I still dont know wtf the 22k resistors and that circuit are lol.

It's usually pretty easy to spot a bad fet by checking resistance, if the one you are looking at is 41ohms between two of the pins, its a sure sign he's bad.. Anything above 1k, i just try to get an idea by checking other fets and comparing and going back and forth..


I would buy 10 fets, think the math is about the same that way. If something goes wrong on one again, you'll have a matching fet you can drop in. Thing to avoid is mixing and matching different lots of fets, when they are the same date code (almost always are from digikey), the performance of the parts is very similar. They will match/share the load better in circuit.
 
1 - 20 of 37 Posts
Top