Author Topic: Behringer EPA900 Repair (Read 3900 times)

Gabriel · « **on:** March 09, 2019, 11:42:02 am »

Hi All,

From my school, I received a Behringer EPA900 that was affected by a power surge. Of course, I took the opportunity to try and repair it. I ordered the parts that were visibly damaged, these being D30 and D31, U2 and C22. The PCB was also damaged with one of the traces completely coming off... The 2 diodes were blown in half and U2 had the die exposed... C22 the lid had just started to come off so I thought I might as well replace it. For the PCB I just used some wire to connect the broken trace.

So I then replaced the parts and applied power... It blew up again.

I have inspected the board again, it's only D31 that blows up now, but I have no idea where to go next. I have attached a few photos and the service manual that I managed to find for it. If anyone could help me out I would much appreciate it

P.S. Here is the service manual with the schematics: http://bee.mif.pg.gda.pl/ciasteczkowypotwor/SM_scena/Behringer/Behringer_EPR900_Powered_Speaker_Schematics.pdf

BurningTantalum · « **Reply #1 on:** March 10, 2019, 02:47:27 pm »

It looks as if you have other failures that you have not found. I would concentrate on the power supply circuit(s). For D31 to die again suggests that something is still seriously amiss, and for C22 to have blown up would suggest that the surge event went a long way into the circuit. What was the cause? Lightning transients?
Worth fixing but there could be all kinds of damage on other PCBs.
BT

Gabriel · « **Reply #2 on:** March 10, 2019, 11:42:50 pm »

I am not sure what exactly caused the surge. My school has far from reliable power... I tried looking for other damage but I couldn't find anything else. This is why I came to you guys. Any ideas where I should start searching?

Gabriel · « **Reply #3 on:** April 07, 2019, 10:58:03 am »

Twoflower · « **Reply #4 on:** April 07, 2019, 11:31:14 am »

Have you checked the BR1? If this one is shorted you might have enough reverse current on the diodes to fry them. And that would also put the D30 in danger. And is probably damaged as well, even if it seems to be OK.

What's connected to J4? That might help to identify why there's such a high current flowing through that poor diode D31.

Gabriel · « **Reply #5 on:** April 07, 2019, 11:59:52 am »

I'm not sure about BR1 a visual inspection it seems fine but when I whipped out the multimeter it is going wild... I'm not sure if that is just because it is still in the circuit I will desolder it and check tomorrow. As for J4, it is a switch with J3 for 120V/240V operation (it is closed for 240V (shorted to J3)).

Twoflower · « **Reply #6 on:** April 07, 2019, 12:05:46 pm »

Visual doesn't tell you anything. And can you explain 'going wild' a bit clearer? My multimeter never growled at me ;-) But since there are many caps around the rectifier: Yes, the reading might be confusing (e.g. charging the caps, reverse measurement then shows negative resistance, ...).

Thanks for the J3/J4 explanation.

Gabriel · « **Reply #7 on:** April 07, 2019, 12:11:56 pm »

For example, probing pins 1 and, 2 + on 1 and - on 2 shows less resistance then + on 2 and - on 1. It is weird. And then some of the combinations the resistance just keeps falling or resistance.

Gabriel · « **Reply #8 on:** April 09, 2019, 01:02:49 pm »

I got around to desoldering the rectifier today. It's dead. Time to order a new one! I'll install it when it gets here and let you guys know if it works.

Gabriel · « **Reply #9 on:** June 12, 2019, 11:04:09 am »

So the rectifier arrived today. I replaced it and.... it doesn't explode!!!! YAY! I then assembled the rest of it and no power.

I checked the power on CN3 (the output of the PSU) and there is nothing. Any ideas where to go next?

LateLesley · « **Reply #10 on:** June 12, 2019, 12:00:09 pm »

Well you know the bridge had failed - was if failed short? If so, now it's replaced, next step would be to check D30, D31, C2, C3 (since they're polarised), and D28 D29. I would also check U2 and Z1 zener diode, its at the bottom left of the second page of the PDF, but it does seem it's connected indirectly to VDC-H - which is the power line directly after the rectifier.

Gabriel · « **Reply #11 on:** June 12, 2019, 12:10:43 pm »

I have replaced D30 and D31 (they kept blowing), C2 and C3 seem fine on a visual inspection and C3 hold it to charge but I don't know about C2 I will check it. D28 and D29 are fine on a visual and I checked them with a multimeter and they read correctly. I have also replaced U2 and it is fine. As for Z1, it seems to be shorted? I will desolder and double check and replace it if needed. Thanks for the help!

LateLesley · « **Reply #12 on:** June 12, 2019, 01:11:49 pm »

yeah, see the diodes would have kept blowing as they would have got reverse voltage across them with a blown bridge, thats why I was worried about the polarised caps too, as they'll have seen reverse voltage, same with the zener through U2, which is why I suspected them. Basically any polarised component (read polarised capacitor or semiconductor) which could have seen the reverse voltage, should be treated as suspect, and checked. I think I was suspicious of U4 too, as it had an op amp input tied to VDC-H through resistors, but then the resistors may have saved it. Worth checking it's operation though. It may also be worth checking Q9 Q10 FETs, but I have a feeling it's possible they survived. (if the diodes failed short).

Gabriel · « **Reply #13 on:** October 25, 2019, 12:53:38 pm »

Hi All,

I'm back again!! I have ordered and now received the new lot of parts! Good news, it is halfway to functional, the front panel now fully works. This means that all the voltages coming off CN4 are fine. On the other hand, all except +15 and -15v on CN3 (going to the amplifier board) do not work... After some probing with the ocilloscope I am even more lost than before. It seems there is no voltage on the transformers secondary side (~4.5V AC).

Any help with what to try and probe is much appreciated...

Many Thanks!

LateLesley · « **Reply #14 on:** October 27, 2019, 10:02:50 pm »

YAY! Well at least it's not catastrophically dying. If you look at it, it is basically two PSU's in one. You seem to have fixed one half of it.

Looking back at the circuit diagram, if you look at the right had side of page two, halfway down (section C marked in the right border) you will see the pinouts for CN3 and CN4.

I would say the first thing to do is verify the Voltages on CN4 - you say it's all working, but without measuring, you do not know, and it has 4 rails on it. It may be only one is working. So Multimeter set to DC 100V or higher (And be careful probing, there's voltages there that will give you a tingle) Black lead on pin 1 (GND) , check -
Pin 2 - Should be +5V or thereabouts.
Pin 3 should be +48V.
Pin 5 should be -15V.
Pin 6 should be +15V.
Pin 4 is another ground.

On CN3 - There are 6 voltages here. Some high ones which will wake you up if you touch them. So, take your time, set up your measurement, and make sure you touch nothing bare metal when testing these. (with your hands or body I mean),

Pins 4 and 5 are ground. Black probe here.
Pin 1 - +80V
Pin 2 - +40V
Pin 3 - +15V
Pin 6 - -15V
Pin 7 - -40V
Pin 8 - -80V

Take your time, set up each measurement, and test them individually, preferably with test clips rather than having to probe live. Report back what voltages you have, and what ones are missing, then we can delve back into the circuit diagram, and give you components to check.

But if, as you say the + -40V and + -80V are missing, there is a protection circuit, set up with U4 which has 4 comparators, and that could be disabling the high voltage supply because it detects a fault. (Don't poke around U4, it has line voltages around it, 120V on inputs). U5 controls the high voltage supply, and is powered on pin 15 and 13 (should be 16V), and has a voltage reference on pin 16, though I'm not sure of the voltage of this.

That should get you started. Let us know what voltages you have, and how you get on.

Gabriel · « **Reply #15 on:** October 28, 2019, 12:33:44 pm »

I have tested CN4 and all the voltages are fine. (I had funny things happening with the not using the right ground but when compared to the right ground they all work). As for CN3 nothing... on any of them except + and - 15v. I did order a replacement op-amp (U4) but as it is SMD I put off replacing it... Maybe it is broken?

Anyway, that's the current state of things

Thanks so much for your ongoing help! I really do appreciate it!

LateLesley · « **Reply #16 on:** October 29, 2019, 02:14:54 am »

OK, So that means some of the supply is working.

The part that is faulty is going to be centered around U5 at the top half of the PSU schematic on page 2.

Now we need to check U5 is powered, so you check pin 15 to see if it has 16V or thereabouts on it. Pin 12 and pin 1 have Gnd on them. Check pin 16 if you have 16V on pin 15. pin 16 should have 5V on it, if the chip is powered. It's the Vref voltage, which i've seen going into the NE555 (U6) and into the protection circuit (U4)
_____
You should also check pin 10 - this is the SHDN which i suspect is a shutdown pin, active low. If it is high (has voltage), then it'll not power up the +-40 +-80 supplies. That pin is controlled by U6, a trusty NE555 chip, which in turn is controlled by the protect signal by the looks of things. I think it gives a delayed power up, and also shuts it down if the PSU is in protect mode.

The protection circuitry seems to rely on several things, some opto isolator inputs, and some stuff off board, on CN1, and CN5. It's kinda melting my brain trying to figure it all out.

it has inputs from 120V, SC (From Transformer/L1 T106 -2), VRef from U5, +16V from transformer T3-D/ D27, U8 Opto which seems to be connected to CN1/CN5, VDC-H which is the line rectified voltage. oh, and there's a thermal switch in the mix too. If all is well, it seems to output a protect signal and an opto signal on U10. The protect signal seems tied into the NE555. Which looks like it controls a relay and the high voltage supply.

I'm trying typing it out to try and get my head around it, cos parts of the schematic are confusing, with some parts looking like inputs when they are outputs.

OK..... Things to check D14 (powered off, it has 120V on it! YOU HAVE BEEN WARNED

), D12, D13 and zener Z3 (5.1V), Q5 (PNP), Q6 Q8 Q11 (NPN), U8 Opto, And then finally U4
the quad comparators. Just check them for shorts with the power off, and check the Z3 zener voltage if you can. I think Q8 sends out the protect signal (it grounds if in protect) And that in turn turns on Q7 (PNP) near the NE555 (U6) So I think those two transistors are worth a check. It'll stay in protect if those are faulty.

That should give you a few things to look at.

Good luck.

Gabriel · « **Reply #17 on:** January 28, 2020, 03:52:36 am »

Hi Lesley,

Thanks for your reply! Sorry for getting back so late, I have been overseas and travelling and haven't had time to look at it. Also, I lost my multimeter while travelling, so I got myself a nice new Uni-T 139C

.

I got it in the mail today and sat down with it straight away. Both pin 15 and 16 have ~9v on them and SHDN also has ~9v on it... FYI there is nothing actually on CN 1 or 5 (or 6). It is just an unsoldered header (possibly used in another model?)

Also checked some other things:

D14: OK - 0.49V
D12: Dead? 0V both ways...
D13: OK - 0.59V
Z3: OK - 0.72V
U8 (opto): Dead? Get around 9k\$\Omega\$ both ways

I am unsure how to go about testing the transistors. They are all SMD so desoldering them is just asking for trouble.

Anyway thanks again for all your help! Hopefully, we can fix this mess.

LateLesley · « **Reply #18 on:** January 28, 2020, 09:37:45 am »

Welcome back! And Congrats on the new meter, commiserations on the loss of yer old one.

Well, if you were measuring the Diodes in diode mode, and D12 was 0V each way, yep, it's dead.

Also, Z3 seems low... you might need to check that out of circuit. Sometimes it's hard to measure stuff in circuit, as other parts can affect the reading. And I think with zeners, your meter's test current could affect the voltage measured too. SO the reading could be very dependent on the zener and the meter. Maybe someone with more experience than me can chime in here.

As for the transistors, for starters, just test to see if any are shorted on all 3 connections. Many times they fail short, so one which is shorted out on all 3 connections instantly draws more attention. It's not always the case, because again, you can get other components around them with low resistances, especially if they are used in switching applications (EG to control a motor), But it is a kinda shortcut way to focus in on potentially problematic transistors.

And the opto may be dead, I think it's meant to be open circuit (when off). You might need to pull that and build a wee test circuit for that.
https://www.antimath.info/electro/how-to-test-an-optocoupler/
http://www.electronicrepairguide.com/optoisolator-ic.html

EDIT : And after looking at the schematic again to refresh my memory, If D12 is toast, U4 is highly suspect. you might need to pull that component too. And check D27 that seems to be where U4 gets it's 16V supply from, on the bottom left of the schematic page 2. U4 and Z3 are on the bottom right of page 2 of the schematic.
And if you measured high on SHDN the protect might be activated, possibly due to a faulty U4. You have to check all around the protect circuit, and the signals into it too.

Gabriel · « **Reply #19 on:** January 31, 2020, 12:30:50 pm »

Hi Lesley,

I desoldered the opto, it's dead...

I do have a spare quad op-amp (U4) that I previously ordered just put off replacing it.

The transistors seem fine, none of them are shorted out.

Apart from the diode and the opto is there anything else I should try and source and replace?

Many Thanks,
Gabriel

LateLesley · « **Reply #20 on:** February 02, 2020, 08:10:55 am »

It's hard to tell, I would check all the transistors and diodes, around the U4 protection circuitry. It's always the problem when supplies blow, massive amounts of power can flow through unintended paths, and take out quite a few components. And there really is no way around isolating and testing various parts of the circuits. I'm afraid it'll be you and your meter who will decide if a part is dead or not.

Gabriel · « **Reply #21 on:** April 18, 2020, 10:35:43 am »

Well, this repair has come to a sad end... I had a bit more time now with school off and through I would take another look at it. While attempting to replace U4, I managed to rip off the PCB pads for most of the pins.

I was too rough and the heat wasn't enough.

I got into contact with Behringer about a spare board and the referred me to my local dealer for help. I am waiting for a response from them. Still might be able to bring it back to life (depending on the cost of the spare board).

Oh well, lessons have been learned. $:-\$

aqibi2000 · « **Reply #22 on:** April 18, 2020, 01:44:07 pm »

The traces can be rebuilt!

LateLesley · « **Reply #23 on:** April 18, 2020, 10:25:39 pm »

Quote from: aqibi2000 on April 18, 2020, 01:44:07 pm

The traces can be rebuilt!

I agree, back in the day, we used to get pad repair kits, which could rebuild traces on the board. I wonder if they still do them....

AAAAnnnnndddd I got a shock, they still do them, but they are bloody expensive, uneconomical now.

https://www.jensentools.com/search.aspx?f=1%3A763

We used to get stencils for ICs with all the pads, and traces, on clear sheets ready to lay on the boards. You technically could build a board that way, though etching was better.

Anyway, its a sore lesson to learn, we all go through it. You always need to make sure things are heated through, before moving them. We've all done it, and still do from time to time. That's where the infamous bodge wires come in. You still have that as an option. :-)

https://www.instructables.com/id/How-to-Repair-Damaged-Printed-Circuit-Board-Pads/

Though it depends how bad the damage is, and if you want to put time into it. :-)


EEVblog Main Site	EEVblog on Youtube	EEVblog on Twitter	EEVblog on Facebook	EEVblog on Odysee

Author Topic: Behringer EPA900 Repair (Read 3900 times)

Share me