Behringer DDX3216 - Trying to repair, please help!

[Jazz P Bass]

You really do need a scope & an LCD datasheet to see what is going on with the LCD.
Initializing an LCD is a tad more complex than simply applying voltages.

[dutch_anykey]

Thanks for the help guys, I appreciate it.

@nikb: Yes it is a lcd with this elastomeric connectors between the glass and pcb. But I think the connections are o.k. Because the screen turns dark with no "empty" lines between them (specially with the potentiometer disconnected, which means maximum contrast setting)

@ Jazz P Bass: The tools are no problem, I got a microscope and a descent scope. But yes indeed without the specs of the display it is difficult. I really would like to have the complete schematic with the contrast setting part. It's only partially there in all the schematics that i have found online. And on the lcd itself there is no info at all.
Since the rest of the mixer is running well (tested with sound on input and headphones connected to the mixer), I'm really suspecting the problem is on the display pcb itself. The contrast regulating part in particular. Normally you would just order a new display and be done with it, but these are just no where to be found. And a drop in replacement wouldnt have gone unnoticed on the web.
I don't even know what voltage the contrast voltage has to be. Normally this is a higher negative voltage compared to the +5 volt Vcc. I think the SCI7654 chip on the lcd PCB is responsible for generating this voltage (dc-dc converter that can quadriple the input voltage and outputs it as a negative voltage) But the schematic only shows half of the chip and connections But I think this chip drives the nagative voltage needed to bring the contrast. This voltage goes to the next chip (LM324) which devides the voltage in 4 steps and sends each "step" to the 4 control IC's (display driver and character processor) that are connected to display itself. If only I had some more info on the display itself and a complete schematic haha So each pixel voltage would be the negative contrast voltage plus the small voltage that comes from the display driver for each pixel. I even wonder now if this measured vcc 5.4 Volts instead of the exact 5 volt would be enough to disturb that voltage balance.

When the mixer is turned of for a few hours and then turned on, for a few seconds I can see the startup logo and somewhat the menu for a few seconds after that. But then after a short while it goes black (like it is shown in the video) turning it off an on again doesnt help, it needs to be off for a while again and then I get some picture again for a few seconds.

It's driving me nuts haha. I want to fix this thing, its keeping me occupied.

It's 4 in the morning here, just got back from an emergency callout from work, I probably shouldnt have rattled on here with my sleepy head

br
Maurice

[g1]

At this point I would agree with guitardad and just recap the psu anyway. If nothing else it will be preventive maintenance.

My opinion of SMPS just got adjusted. I recently replaced the smps in my 8 yr. old vista computer. It had been running fine though a little slow I thought. Then it was just dead.
11 of 14 electrolytics in the supply were dead. Not just a bit off, shot. Yet it had seemed fine till it quit. Those 11 caps did not die overnight, one of them maybe. To top it off, about 6 of them were the same value, voltage, and (strange) brand. Every one of those was a different physical size! For all I know they were salvaged from a junk pile and re-labeled.
Sure changed my opinion on cheap caps in smps though. I don't think longevity is a design concern.

[guitardad]

Yeah the caps have been a source of many oddball symptoms with these. Start there.

[dutch_anykey]

Hi,

The caps will be on my doorstep tommorow according to tracking. So I will go ahead wih recapping the PSU anyway. Will let you know what the results will be, probably tomorrow evening

In the meanwhile I have been reading up on datasheets of the chips that are on the LCD PCB. Just trying to figure out the whole picture on how this LCD circuitry works. Like these voltage steps created around this lm324 chip I mentioned earlier on. After recap is done and it still isn't working i will go from there. But man it would be great if it is the PSU afterall.

[dutch_anykey]

O.k. the recapping job didn't bring anything. Still the exact same symptoms. I dialed the voltage down from 5.4 to 5.0 volts. But no difference either.
I suspect the LM324 or else the SCI7654 chip on the LCD PCB. Going to hold the pcb under my microscope and trace all routes to complete the missing parts in the schematic. From there figure out what voltages should be present in those cuircuits.

br
Maurice

[dutch_anykey]

Hey guys,

This is going to be a difficult story. Seems like the lm324 and SCI7654 are doing their jobs. At least if the negative voltage should be around -10 volts. with contrast pot the voltage is between -9.36 and -12.4 volts. But I have no reference from a working display. So I'm more or less stuck here. Maybe the toshiba end drivers (4 big chips on the lcd display) are failing.
Suggestions are welcome to check something.

br
Maurice

[dutch_anykey]

Hi guys,

Some good news and some bad news.
So I did proper measurements on the LM324 circuit. The circuit is responsible for producing the various voltages need to drive the lcd. (v2,v3,v4,v5 and vee) it goes to the toshiba row and column drivers. the total sum of this voltage is this negative voltage that comes from the SCI7654 chip and this voltage is also altered by the contrast pot. (varies between -9.36 and 12.3 volts in my case)
To get the separate voltages v2.v3 etc. there is a series network of resistors R5,R4,R3,R6 and R2. So they work as a voltage devider. There you have the voltages, but you can not bring this voltages directly to the display since it will alter the voltage when a load is added to the resister. Her the LM324 comes in. It is wired to be a "voltage follower" it means it reproduces the input voltage and outputs it on the output. In this way it can handle the load request from the toshiba drivers.

What I found on my display is that one voltage was way off on the output of the lm324. Ordered a new one and put it in. Voila it works...happy dance. But only for 5 minutes. Because the display was turning darker and darker again. So I suspect the new LM324 is dying again. weeeeh.
I will go on for a little bit. But this thing is just about to get me down and give up. I suspect the the toshiba driver is going faulty and drains to much power, or the lcd itself is faulty. Not sure how i'm going to check this yet.

Another thing that i haven't mentioned yet, just before i found the lm324 fault I also decided to disassemble the LCD itself and check these rubbery connection...So I'm missing a few lines But I guess I can get those back with adjusting the alignment of the rubber.

I will put up my drawing this evening of the circuitry of the LCD that i have drawn up. Maybe some future reader can have an advantage of it.

br
Maurice

[g1]

there is a series network of resistors R5,R4,R3,R6 and R2. So they work as a voltage devider. There you have the voltages, but you can not bring this voltages directly to the display since it will alter the voltage when a load is added to the resister. Her the LM324 comes in. It is wired to be a "voltage follower" it means it reproduces the input voltage and outputs it on the output. In this way it can handle the load request from the toshiba drivers.

What I found on my display is that one voltage was way off on the output of the lm324. Ordered a new one and put it in. Voila it works...happy dance. But only for 5 minutes. Because the display was turning darker and darker again. So I suspect the new LM324 is dying again. weeeeh.

Those resistors and their solder connections are good? Can you verify that the voltages going in to the lm324 are remaining good while the outputs go bad?

[dutch_anykey]

Hi g1,

Yes the the voltages on the input resistors are stable. It's really the output voltage that is slowly slipping. BUT, i got a little further and I think I found the reason for this
I also told you that I disassembled the LCD and that I was missing some lines. So I took it apart again cleaned everything with 99% alcohol and carefully put it together again. And surprise the display is working o.k. now. Maybe in one corner it is still a tad off. So nickb was right on the money here with the comment about the elastomeric connector!!

From what I see now, this might have be the root cause of the display not working when I bought it and killing the LM324. What might have happened is that some connections got shorted through this elastomeric connector and a higher voltage was fed back to to lower voltage connection. The guy where i bought it from was using it "in the field" So I bet this unit has had it's fair share of beating and bumping when moving it around to and from gigs. I guess the elastomeric connector got shifted.

For now I'm afraid that this unit can only be used in a stationary studio environment. Not really a situation that I bought it for. But lets see how it goes.

So thank you all for thinking along with me! It was a great help. Next on the table will be a behringer EP2500 PA .... I already took a peek. I thought it would be simpeler then this mixer, but I noticed this 2 step power supply in the end stages...brrrrr

Here is the drawing i made up to understand what was going on:

[dutch_anykey]

One more thing,

Maybe people that own a ddx3216 can help me on this one: Do you have to play a lot with the contrast pot during normal use?
I build the display in the case and put the few hundred screws back in to have it completely finished. But I experience a rather high temperature drift that needs constant corrections on the contrast pot.
So, took the display out again, and then noticed when holding my fingers on the 4,7 uF caps that the contrast changes. blowing cold air over it changes it in the other direction. One of the first things during this repair was changing the caps for new ones because the old ones measured zip capacitance. I wondered if the new caps might be wrong.... It's the standard alu smd caps. So I changed them for normal electrolyte caps. But the result is more or less the same. I wonder if there are caps that have solid capacitance over a wide temperature range.

br
Maurice

[nickb]

Is there any chance C7 (on your schematic) is bad or inserted the wrong way round on the board?

Those voltages should not be so sensitive to temperature. The capacitor leakage current might vary a bit with temperature but the output impedance of the LM324 is so low that it should not matter. Quite possibly, with one output bad, that voltage is varying as the LM324 is not behaving as it should.

[dutch_anykey]

Hi Nickb,

The display is all garbled again (so it got worse)
The capacitors are new and just double checked that the orientation is right. I quickly measured with a component tracker (not sure if you know this device in combination with a scope) from there I can see that the output of the lm324 is most probably dead again. It is late so I will measure decently tomorrow.
What I have observed on the LCD glass plate itself in this faulty mode is that there are some darker lines. And I mean darker so that they stand out from other black lines. Maybe this indicates a short in the lcd itself, So it gets a combination of voltages on those lines and therefore are more black. Perhaps that's why I was seeing this abnormal voltage on the lm324. So that voltage not coming from the lm324 side, but coming back from the lcd plate.
If the LCD plate is shot, I'm done. No way I can find a replacement. But I will try to measure and find ways to confirm that tomorrow.

As for the mixer itself, during the time it worked I took the opportunity to fiddle around with it. I must say it is very nice with the compressors. limiters, FX's and more stuff build in. So the basic things really can be programmed without knowing to much about this unit. And it saves the need to have external effects and compressor/limiters.

[nickb]

The problem with messed up multiplex voltages on an LCD is that if on average there is a DC voltage then it can damage the LCD. It is possible you could isolated the 'bad' LM324 output from the driven part ( cut track) to confirm that the driven part is the problem? I would expect these voltages to do to a switch of some kind that actually does the multiplexing. If that is external to the LCD then you may be able to replace it. If internal, then game over.

I have heard of many kinds test equipment in this area but a component tracker is a new one on me. I suspect something might have been lost in translation. Perhaps you are mean a Bode analyzer (Vector network analyzer)?

[dutch_anykey]

Hi,

Regarding the tracker (anyway that's how we call it locally here) I did a quick search on youtube https://www.youtube.com/watch?v=BF0cdoetwRw , so it is more or less the first vid I came across that shows the working. It's not the device/brand that I have here. But it pretty much shows what it does.

Before I go measure again I want to check also another thing that came up in my mind. I disassembled the lcd and put it back together as I told you. As for alignment of the display, the contacts in the elastomeric connector are very close to each other. so the contact distances on the PCB and LCD are far bigger. Can't go wrong with making and unwanted cross connect was my idea. But I forgot that the PCB and the LCD itself should of course also be aligned properly to each other. I have to check if it can be miss aligned. But it's one thing to check before going measuring again.
And indeed these "V" voltages are fed to toshiba display drivers (three T6A39 column drivers and one T6A40 row driver). They are responsible for the multiplexing of these voltages to the right columns and rows. Now I have to find out if one of these chips is problematic or the LCD itself.