Kicker 2500.1 Protect Mode Issues [Repair]

Status
Not open for further replies.
And here are the FB31's in the same manner.
 

Attachments

  • Pin 1.jpg
    Pin 1.jpg
    206.2 KB · Views: 106
  • Pin 2.jpg
    Pin 2.jpg
    211.5 KB · Views: 105
  • Pin 3.jpg
    Pin 3.jpg
    216 KB · Views: 101
As a side note, using DC coupling is important when checking waveforms, especially when checking the gate drive for the power supply (not applicable here but would be if you were troubleshooting the power supply drive circuit).

That's the option on my scope in the 'CH'(annel) setting that says Coulping, yeah? I didn't even think to change that. It has been on AC this whole time. As for the BP1200 you posted a picture of, that's funny. I actually run one of those in my second car myself. Good little amp for what it is.

#17:

What were you using for ground?

I was using the center tap indicated, is that the correct thing to use? Sorry for so many posts, I didn't even catch yours in the process of posting pictures.
 
I should have caught this earlier,and the reason why I mentioned using the center tap as reference because this amp has a bridged output, which means the neg speaker terminal is not connected to the center taps.

I will study the photos later tonight and post any observation. This may take some back and forth, and the number of posts can't be helped,especially when dealing with this beast!

Sent from my HTC Desire 626 using Tapatalk
 
Recheck the waveform on the LM361 using the secondary center tap and post screenshots.

You will find 4 U-shaped bars on either side of the board near each bank of outputs,8 in all.The bars carry the +/- rail voltages. With your DC voltmeter,measure the the rails with negative probe on the center tap post.

You might try removing the driver board on the bad output side and recheck the drive signal on the good side. I suspect the driver board may be bad.

Sent from my HTC Desire 626 using Tapatalk
 
Okay, so I did a bit more testing as recommended. For the following tests, both driver boards have been removed from the amp, and the preamp/tone board is also removed from the amp. The amplifier has none of the FB31N20D FETs in it, but the IRF9640's are still installed.

I checked the voltage on the u-shaped bars, and I got ~58.0v and ~ -58.0v on the bars near the 9640 and FB31, respectively. This had very little variance, ~0.1v at most. All banks matched. I used B- as ground reference.

The first set of waveforms is from the outputs. All waveforms matched nearly identically (or at least identically enough I could not distinguish a difference) between banks. The first 3 are the 9640, second are the FB31'S. The secondary center tap is used as ground reference here.

The next waveforms are from the power supply. I did note that one side was different than the other on one of the legs. B- is used as ground reference.

The last waveforms are of the LM361, pins 4, 9, and 11, respectively. The secondary center tap is used as ground reference here. I do apologize, as it appears I only have the waveforms of 4 & 9. I will try to grab 11 next time I have a moment to work on this, if it is needed.
 

Attachments

  • LM361 4,9.jpg
    LM361 4,9.jpg
    105.4 KB · Views: 54
  • PS2.jpg
    PS2.jpg
    174.1 KB · Views: 45
  • PS1.jpg
    PS1.jpg
    163.4 KB · Views: 44
  • FB31.jpg
    FB31.jpg
    190.3 KB · Views: 81
  • IRF9640.jpg
    IRF9640.jpg
    186.3 KB · Views: 74
Try one driver board at a time. If you happen to know which driver board was on the good side, use it. If not try one and then the other. Whichever driver board works best then move it to the other side. I suspect that the driver board on the "bad side" is the issue. Once you have determined if this is the case we can take the next step.

Sent from my HTC Desire 626 using Tapatalk
 
You can remove all of the outputs FB31's and 9640's included.This will allow you to check waveforms and determine which driver board is bad. Did you say you have another 2500.1? If so you may be able to substitute the driver board's. Post pics of any driver boards you have,there were some revs made and you don't won't to mix and match.


Sent from my HTC Desire 626 using Tapatalk
 
I have come bearing waveforms! All output transistors have been removed. I am using secondary center tap as ground. Driver boards have been labeled 1, 2, 3, and 4, to keep track. 1&2 are original, 3&4 are spares.

In the following waveforms, I have board 1 in the 'Good Side', labeled by me as banks 3 & 4. Both sets of waveforms were matching between banks.
 

Attachments

  • IR-CARD1.jpg
    IR-CARD1.jpg
    185.8 KB · Views: 75
  • FB-CARD1.jpg
    FB-CARD1.jpg
    179.5 KB · Views: 75
Check the DC voltages around the LM361 and the MC14060 which is right beside the 361 and post. Comparing the pictures you posted in #17 and #27 it appears you may have
A) Damaged the MC14060 and possible the LM361 or

B) The regulated +/-15 volt supply
 
PapaZBill, you were correct. I swapped the LM361, and got no new results. I did not remove the donor LM361, and then swapped the MC14060 (labelled CD04060DE on my amplifier), and now I have waveforms that look clean and exactly like what Perry's link shows. No messy waveforms whatsoever as were previously observed on 4/9/11. I will install driver boards tomorrow, and re-post the waveforms from the outputs, and see if you think it might be safe to install FETs again. Do you by chance have any resources on what the waveforms on the output should look like with the driver boards in? It doesn't have to be exact, just would like to know about what I might want to/expect to see. Thank you again for all of your help, I greatly appreciate all the advice and explanations you've given thus far.
 
Last edited:
The attached photos show the drive signal and rail voltage. Your rail voltage will be different but you can see how the drive signal drives up and down from the respective rails and shows the amplitude of the drive signal.
 

Attachments

  • P1010012b2.jpg
    P1010012b2.jpg
    93 KB · Views: 41
  • P1010014b2.jpg
    P1010014b2.jpg
    92.8 KB · Views: 43
  • P1010009b2.jpg
    P1010009b2.jpg
    93 KB · Views: 37
  • P1010011b2.jpg
    P1010011b2.jpg
    92.5 KB · Views: 34
Okay, I popped the driver boards in. I now get this for the IRF and FB31. Note that the last two unlabeled waveforms are from bank 2, pin 2 of both IRF and FB; they were the only discrepancy between banks as far as waveforms go, the other three banks matched the corresponding waveforms almost identically. I believe this is the channel that was previously damaging FB31's, just to note.
 

Attachments

  • 1007160436.jpg
    1007160436.jpg
    193.2 KB · Views: 38
  • 1007160437b.jpg
    1007160437b.jpg
    201.5 KB · Views: 41
  • 1007160438b.jpg
    1007160438b.jpg
    46.1 KB · Views: 36
  • 1007160439.jpg
    1007160439.jpg
    46.4 KB · Views: 34
Without FETs in the circuit, the gate drive is the only waveform that you will have. When your scope is at 50mv/div, you're just looking at noise. The drain is at 0v and the source is at rail voltage. You cannot have any significant signal (only noise) on the drains without the FETs.
 
Status
Not open for further replies.