Behringer DDX3216 - Trying to repair, please help!

[Steve Conner]

I definitely agree replacing the crystal is a wise move. Farnell should have one, but if you want to do it in real ghetto style, bust open any battery powered digital clock or watch, and take the 32.768kHz crystal from that. Old computer motherboards often have one too.

Sometimes you can check a crystal oscillator with a scope. But other times prodding it with the scope can make it stop, so it appears faulty even though it's perfectly functional the rest of the time.

[fznuk]

I definitely agree replacing the crystal is a wise move. Farnell should have one, but if you want to do it in real ghetto style, bust open any battery powered digital clock or watch, and take the 32.768kHz crystal from that. Old computer motherboards often have one too.

Sometimes you can check a crystal oscillator with a scope. But other times prodding it with the scope can make it stop, so it appears faulty even though it's perfectly functional the rest of the time.

Cheers again guys!

I will see what i have kicking about to try it before ordering new. Not had much chance to have a look over the past few days due to getting in a lot of PC work.

I will get back to you all after i have replaced the crystal and done a long test.

Cheers again!

[fznuk]

Still Failing?!

Hi guys,

Bought and fitted the new crystal (Q3) but i still have the original error where after a random amount of time the master LED's light up full and you get no sound.

One good thing is that you don't need to wait 20-30mins before you can switch it back on to clear the error?!

My probes still havent come for the scope as yet so cant check for any power ripples etc.

It will also crash at boot-up on the logo screen if switched on and off too fast.

and since i still get crazy hi pitched noise from the PSU, almost like digital modem noise, but higher, im still wondering if the PSU is struggling somewhere or input Caps are bad on the primary?? I bought a Cap meter which will go up to 20,000Uf, and i know it isnt brilliant but will give me an idea.

Would still love to fix this beast!

Cheers again for all your help! and any more would be appreciated!

[km6xz]

Sorry but replacing parts randomly or on advice on a web site often results in worse operation. Diagnosis the problem and fix that, not throw parts at the problem until it magically goes away.
You have a very strong clue that no one is considering in the rush to replace a crystal. The power supply has a normal switching frequency of way above human hearing between 200khz and 1Mhz depending on the load. Did you try to isolate the supply or reduce the load to see if the supply is singing at a low frequency by itself or due to excessive load? A low switching frequency means very low transformer efficiency and more heat in the switching transistors, and lower output, higher ripple etc.
What is causing the singing? The mixer or the supply not able to switch at normal rates? You do not need probes for testing DC output ripple level, but any measurements in high Z circuits do require the probes. So far, it appears that the most basic measurements have not been taken due to lack of gear. The supply performance should be isolated by testing on loads, without the mixer connected. From there, you will know where to look for the problem.
The biggest problem is that you are attempting to repair a complex device that 80% of current pro techs would not even try to get into due to its complexity and lack of service information. Even the warranty stations had little more than a schematic and little other assistance. Behringer USA knew NOTHING about the unit, which was about 2.5 light years ahead of anyone's electronics knowledge at the US based distributor.
When we were the warranty station for those, most of the repairs we did, were correcting problems induced by other Behringer warranty shops. There was no one to turn to for help so we reversed engineered much of it, and built test jigs for isolating different pc boards and the supply. That is what it took to successfully repaired hundreds of them.
Starting a repair career on a complex digital board is not going to be easy unless a basic problem with the supply, or a shorted bypass cap is found. I think you would be better off sending it to a shop that has built up a reputation for being experts in these, it will probably be cheaper and a lot less frustrating.
Good luck

[Steve Conner]

Good point, a SMPS in a piece of pro audio equipment should not squeal or fizzle, as the noise is electrical as well as acoustic and would get onto the audio. So if it does, it's probably faulty, or being overloaded by a faulty part elsewhere in the mixer.

Electrolytic caps in SMPS are wear items, but a capacitance meter isn't much use for testing them, you need to know the ESR. (However if the capacitance is way below spec, it's a fair bet that the ESR will be bad too.) The output filter caps are more likely to cause instability than the input ones.

Random crashes sure sounds like a bad power supply. You can check with a meter to see if the rails sag to weird voltages when the problem appears. This is a sure sign of a problem, but it doesn't tell you whether the fault is in the PSU, or just some part of the mixer overloading it. For that you also need to measure the current.

The secret of complex repairs is to keep dividing the system in half and testing which half the bug is in. Complicated systems should be designed as a bunch of modules to allow for this. In this case, we want to find if the fault is in the PSU, or the "rest of the mixer".

This doesn't catch "systems" faults, where the system doesn't work even though the modules all test OK. The fault is in the interaction between them. An example would be if Behringer accidentally made the power supply too small to meet the worst-case demand.

[fznuk]

Well, i hear what you guys are saying and appreciate your help and advice. I really thought the changing of the output caps would cure it like with a few other SMPS power supplies on other items i looked at, but i suppose it all depends on why they failed.

Im not new to electronics but i am quite a bit rusty since im not repairing electrical items every day, and usually its simple repairs...and can vary depending on the job. My main job now is PC repair which never really requires any component level repair unless im just doing basic diagnosis rule out a bad motherboard....and then its just thrown away and replaced. Google is my friend at the moment for any literature in brushing up on some refresher basic electrical repair knowledge....im just too lazy sometimes to do a full read into deep specific repair....im more of a practical than a theorist...if you get what i mean

I would like you at least try to repair this desk as i dont like to give up to easily...and still have a hunch that it may be a PSU problem....but as said, it could be a bad component on one of the other boards thats bringing down the PSU?

a Couple of things before i dive right in again...

1) The PSU audible noise, could this be bad coil design / transformer vibration and part cured by resin?

2) Is the noise due to instabilities with the PFC control to the switchers?

3) The noise was there when the PSU was powered up without the desk connected so does this point more to the PSU than a problem on the other boards or could the damage still have been done by another board fault so now the PSU components have been weakened in some way?

3) I only tested the analogue power output rails so will test all next time. Is there an accepted tolerance with the voltages before being too high or too low?

I think a more thorough test on the PSU is needed or do you still think im out of my depth? Please remember i have nothing to loose going deeper since the desk is faulty now anyway? But i still dont want to go round in circles.....and im enjoying learning more on SMPS PSU's

Thanks again guys...and thanks for helping a novice comparison to yourselves always great to learn from experts - this is why apprenticeships should make a come back!

[Nisse]

In an effort to help here, I just checked every DDX3216 repair in our database. Overwhelmingly, the solution we took to the problem described here was the replacement of crystal Q3, 32.768kHz. Over and over, the problem reads the same: runs for awhile, then locks up, no control, etc. Sure, there are also the ubiquitous cap issues and poor soldering, but this fits the bill. Hey, it's worth a shot. You need to eliminate this common issue first. We stock these crystals. Mouser has them as well, part# 559-NC38-LF.

THAAAAANK YOOOOU John R. Frondelli!!!! You made my day! I read this, and found this crystal on an old, oblolete PC-motherboard I had, so I took that and replaced it on my DDX-board, and...... TADAAAA!!! It worked again!!
I have repaired the PSU, fixed bad connectors, but the problems always came back. The last time, it did not start at all! I have to make a "long-time-test" now, but I'm very positive!
Once again, thanks for your information!


/Nisse

[jrfrond]

THAAAAANK YOOOOU John R. Frondelli!!!! You made my day! I read this, and found this crystal on an old, oblolete PC-motherboard I had, so I took that and replaced it on my DDX-board, and...... TADAAAA!!! It worked again!!
I have repaired the PSU, fixed bad connectors, but the problems always came back. The last time, it did not start at all! I have to make a "long-time-test" now, but I'm very positive!
Once again, thanks for your information!


/Nisse

My pleasure. That's what we are here for.

[Nisse]

No power on ANAOUT on ddx3216 PSU

Hi again,
Now, when there is an expert available , I have another question/issue on my ddx PSU.
The thing is:
I power up the DDX, let it be on for a while (more than 15 min. so it gets warm). Everything is working fine.
The I turn off and then on. Now it's very likely that there is no power on the ANAIN02. (Sorry, wrong in the title) (It's the +-17V and +5V anaolg power)
If I then turn off, wait for a while, and the turn on, it might work again.
I have replaced all the electrolytic capacitor (except the two big ones on the high-voltage side)
Do you (or any one) know about this, and what component(s) I must replace?

Best regards,
Nisse

[gonzosaxman]

I'm new to this forum, and am currently having problems with my desk. I have taken it apart and it seems that it has the usual capacitor problems with the PSU, but I can't get any voltage out of any of the rails at all as they all seem to be shorting to ground. My Dad was an electronics engineer and typically I didn't ask him enough questions before he died, so I'm wishing he was alive right now. If there are any friendly electronics engineers near Aylesbury, Buckinghamshire, UK who wouldn't mind helping me out, please let me know.

Anyway, I am going through the process of trying to fix it currently using my Dad's test gear, and I came across this forum. I have uploaded the complete schematics for the DDX3216 (including the PSU) as a zip file onto the following url for anyone to download if they wish - the file may only be available for a limited time though so if it ain't there, just email me and I will upload it again;

http://www.sendspace.com/file/if20vu


Cheers

[Nisse]

Fantastic!! Where/how did you get the schematics? I've searched a lot on the net for this, but never found anything! Thanks a lot!
Well, as mention earlier, it's quite usual that there are bad solderings on the PSU board. The first step would be to check them. It's very hard to see if the soldering is bad just with your eyes. You will need at least a good magnifying glass, or microscope.
Here's an image of a typical bad soldering that I had on my PSU.
(Click on it to enlarge)
/Nisse

[fznuk]

Fantastic x 2! Thanks for the schematics!!

Still had no joy in fixing the original problem but now got the power supply on my test bench within my workshop so i just need to put some time aside to look at the schematics, test the voltages & stability, then have a look into why a ceramic disk is 'hissing'?!

The Cap (C4 on the Schematic, 223 1KV) starts to hiss/high frequency vibration?! after about 30 seconds of the PSU being switched on (without load or connection to the desk). - this part of the circuit also shares a Transient Voltage Diode and large 18K resistor across it.

This Circuit is repeated twice again on the other two switch mode circuits (I see this PSU as having three seperate SMPS PSUs in one)

I did notice the voltage across the resistor varying 79 - 85V while the other two similar circuits were stable if thats anything to worry about?

Will reply again once i have tested all output voltages.

Cheers again all! - your all stars!

[MarkusBass]

Doesn't the "PSU & DUMMY LOAD" document suggest that the PSU cannot be switched without a (quite complex) dummy load? I think that the noise that you get is due to the fact that you don't have a dummy load. There must be a reson why they provided the schematic of such a complex dummy load, don't you think?

Mark

[gonzosaxman]

Replaced all electrolytics and ceramics in the 17 volt section

Just replaced all electrolytics and ceramics (and CX and CY caps) in the 17 volt section and all the electrlytics throughout the rest of the board (I used 125 degree ones wherever available).

Before this, all of the signal clip lights on the desk lit up and that was it. My problem now is that all of the lights on the desk are lighting up alternately. All of the meters, every alternate mutes and select button, every other button and every rotary encoder and all of the fader motors are on and under tension - after doing research, this seems to be a typical problem.

Whilst this is extremely pretty, the desk is still not working. My research tells me that this is due to fluctuating voltages on some of the rails when under load (possibly the 12 volt rails to the CPU board). I'm now testing to see what is causing the problem. My C4 cap is also hissing. Speaking to several repairers about the board (which it turns out behringer nicked from a plasma TV) the zener diodes (zd1 - zd4) are common failiures, followed by some of the rectifiers and regulators on the board (U3 and U4, KB1M0880). Also, CX1, 2 & 3 and CY caps are also prone to failiure causing shorting to ground and overheating of the in line CX variety.

I may go through the whole board and replace more or less everything (excluding the trannies) as I reckon I could do this for around £100, which seems pretty cheap. Most of the components are readily avaiable or modern better performing equivalents are readily avaiable.

Something I did notice when replacing the ceramics, several of the 104 ceramics on the 17volt rail were actually 103 caps instead. If I replace them with 104 caps as it says in the schematic, will this have any adverse effect?

Also, J7 (AG - DG) isn't present and the screw connecting the heatsink to the board accross the AG and DG rail isn't present? Is this normal? I can't see how it would affect the board and I assume you would want to keep DG and AG separate wherever possible.

Watch this space

Glad my schematics are of use to people.

[Enzo]

Generally on any mixer or similar piece, when I see all the peak lights on, it usually means one of the power rails is missing. So check the +/-15VDC to the analog ICs.