Steinberg MR816x distorted sound

[camarada78]

Hello,

I have a MR816x on my bench that works normally on analog ins and outs, but digital out is distorting every channel and has some constant digital noise when synced through adat. I think its something on the digital converter, but i have no clue where to start. No schems either.

Anyone had similar problem?

Thanks, have a good day

Leandro

 

[RoadrunnerOZ]

Hello,

I have a MR816x on my bench that works normally on analog ins and outs, but digital out is distorting every channel and has some constant digital noise when synced through adat. I think its something on the digital converter, but i have no clue where to start. No schems either.

Anyone had similar problem?

Thanks, have a good day

Leandro

I assume you have set the clock source to ADAT when syncing and the ADAT port to ADAT and not ADAT (6)+SPDIF(2) and the sample rates of the 816 match the source or destination? I have the service manual for this but it’s 25MB if you want it. The schematics cover many pages. All the settings are available in the MR Editor - part of Tools for MR816 - you need to have the drivers loaded as well.

 

[RoadrunnerOZ]

Perhaps if you can let me know what you’re sending the ADAT signal to - maybe a clocking issue with that device.

 

[camarada78]

I dont think its just a clocking issue, might be related, but not just that. The problem happens also with its internal clock.
- using it with internal clock, bad digital distortion in most channels (specially 3-8) but not in 1 and 2
- using it with adat clock (synced with a RME Fireface), bad distortion in most channels also.

I just found the service manual on Scribd, reading it now

 

[RoadrunnerOZ]

Run the tests - depending which version of the manual you have, it should be on p90
Edit: mistake above - from another thread

 

[RoadrunnerOZ]

Check supply 5V and 3.3V rails to IC’s 154 and 103 - DM board diag 002 digital I/O - IC103 is a 272 pin DICE II

 

[camarada78]

Did some tests and now i dont think its a "simple" digital problem anymore.

Analog Ins to Headphones have the same noise. Very distorted as if it is very loud. The noise i was hearing through the digital out (1-8) is also present on individual analog channels. The noise is modulated by the analog pot.

Channels 1, 2 and 4 are normal.
I changed connectors internal wiring to use Analog 8 input controlled by channel 4 pot and it works normally but Analog 4 controlled by channel 8th pot has the same distortion. The distortion is after

I though it could be some power issues but +/-15v are normal. 5v and 3.3v rails also.
The normal channel (4) has exactly same measures as channel 3 but channel 3 is heavly distorted.

The problem is present either as stand alone or using the PC to control.

 

[camarada78]

Update: factory reset no result.
Lowering volumes digitally (through software) changes nothing. I get the same terrible distortion but lower volumes.

I will start the test protocols now but i think its an analog problem on the hardware.

 

[Khron]

How's the Vref of the ADC / DAC chips?

Also an oscilloscope could help with tracing the signal from the analog ins "inward" and see where the signal starts misbehaving. Likely most visible with a plain sinewave.

 

[RoadrunnerOZ]

Update: factory reset no result.
Lowering volumes digitally (through software) changes nothing. I get the same terrible distortion but lower volumes.

I will start the test protocols now but i think its an analog problem on the hardware.

I have an MR816 and had a similar problem - it turned out to be 27KΩ smd resistors on the front and back of the main board in the area of the gain pots, the op-amps immediately following. It’s easy to see with a magnifier if they’re burnt but replace them anyway.
Ch1 & Ch2 the IC to look for is just behind Ch2 combo input socket, it’s IC602, replace resistors R747 &R748 - these are the feedback resistors for IC602 which may also be faulty. I replaced all 4 IC’s immediately following the gain pots and the smd feedback resistors all of which were cooked.
You need to get at the underside of the board next - you’ll need to remove all the screws holding down the DM board, take off the front plate, remove the 8 pot nuts and front combo connector clamps, the two front panel ribbon cables, the insert connector and two power supply cables - leave the rest connected.
On the underside of the board in between each pair of pots 3-4, 5-6, & 7-8 is a set of 4 rows of smd resistors and capacitors for each pair of pots - the resistor numbers are all in little boxes adjacent to their location. Replace R755 R756, R1041 R1042, R1043 R1044 - I just tacked little metal film resistors to the parallel smd capacitors. Their associated IC’s are directly above on the top of the board - these are IC’s 603, 650 & 651 - NJM2068

 

[RoadrunnerOZ]

Just another point - the faults in these areas feed into everything else tied to the same power rails - with no signal coming in you can see the input metering on all the dodgy channels registering in the MR Editor mixer. There’s nowhere for the heat to escape to on the underside of the board and it builds up, or the wattage of those resistors is just simply too low, or both - that’s why I didn’t use the smd resistors to replace the blown ones. I confirmed all this with the service manager at Yamaha and he agreed it would be better not to replace them with low wattage smd - they’re only 1/16W.

 

[RoadrunnerOZ]

The same could also apply in the output stages to the analog output op-amps - there is an 18KΩ resistor in the feedback path of each 5532 - both channels of each for 1 x balanced out. R1415 through R1430 and also just prior to the 5532’s 4 x NJM2068 with the same setup, but using a 7.5KΩ resistor also with an additional 0Ω in the feedback path - see DM Circuit Diagram 011. The DAC feeds to the headphone board use an NJM2068 then a 4556. It’s worth putting a scope across all the outputs to make sure they’re all functioning correctly using a signal injected into a known good channel and routing through all the outputs.

 

[Khron]

I have an MR816 and had a similar problem - it turned out to be 27KΩ smd resistors on the front and back of the main board in the area of the gain pots, the op-amps immediately following. It’s easy to see with a magnifier if they’re burnt but replace them anyway.

What could have lead to those burning up though?

 

[RoadrunnerOZ]

What could have lead to those burning up though?

Bad design - no input protection - can’t handle continuous high signal level? Often signal coming in will be running right to threshold like from sawtooth/triangle/square wave synth, or synth bass etc which can output fairly harsh and hard to deal with waveforms and those resistors just burn out - maybe electric guitar even normal bass with a lot of attack. The unit I inherited had been in a studio with 2 synths on inputs 3 - 6, gtr & bass on 7 & 8 and vocals via tube pre’s on 1 & 2. I have had others for service with the same problem. It seems that inputs 3 - 8 are most prone.

 

[RoadrunnerOZ]

Also the Neutrik combo connectors on the front and rear panel inputs suffer from jack failure - seems to be associated with cheap cables and the shape of the plug tip - Neutrik jack plugs go in and come out easily. I’ve had to replace these in several units, mine I just pulled down as input 1 has failed (my fault using a crappy cable), I already had replaced #2 when I first got it.

 

[NoisyIndividual]

I've had that same problem with Neutrik Combis and cheap jack plugs that get stuck in them.
So stuck there is no alternative but to replace the Combi socket.
Yes, something to do with the shape of the tip on these cheapo jack plugs.

Funny story was I had to replace one of these Combis for a customer of mine, and I thought I'd demonstrate the problem to him on a brand new piece of equipment belonging to the Music Shop I was working out of at the time, using one of their cheap off the shelf jack to jack cables.
The cheap jack immediately got so stuck I had to replace the Combi in that one too. (I thought I'd just be demonstrating how much stiffer a cheapo jack was than a proper Neutrik)

 

[RoadrunnerOZ]

I've had that same problem with Neutrik Combis and cheap jack plugs that get stuck in them.
So stuck there is no alternative but to replace the Combi socket.
Yes, something to do with the shape of the tip on these cheapo jack plugs.

Funny story was I had to replace one of these Combis for a customer of mine, and I thought I'd demonstrate the problem to him on a brand new piece of equipment belonging to the Music Shop I was working out of at the time, using one of their cheap off the shelf jack to jack cables.
The cheap jack immediately got so stuck I had to replace the Combi in that one too. (I thought I'd just be demonstrating how much stiffer a cheapo jack was than a proper Neutrik)

Expensive way to sell cables and a repair!

 

[RoadrunnerOZ]

I’ve used combo connectors in creating FX breakout panels for bantam patchbay I/O’s and the first thing I do is go through all cables that will be used in the rack and get rid of offending plugs. The worst seem to be the moulded plug 1/4” leads - I usually make up custom TRS cables using quad cable and Neutrik plugs and heave the moulded cables - they only seem to work with the metal jack sockets you find in guitar pedals and the cable is usually low grade as well. The tips seem to have a profile on the trading edge that gets caught in the combo jack and tears it apart - first warning is they’re hard to push in.

 

[Khron]

I have an MR816 and had a similar problem - it turned out to be 27KΩ smd resistors on the front and back of the main board in the area of the gain pots, the op-amps immediately following. It’s easy to see with a magnifier if they’re burnt but replace them anyway.
Ch1 & Ch2 the IC to look for is just behind Ch2 combo input socket, it’s IC602, replace resistors R747 &R748 - these are the feedback resistors for IC602 which may also be faulty. I replaced all 4 IC’s immediately following the gain pots and the smd feedback resistors all of which were cooked.

That is so weird though - even if those resistors were across the +/-15v rails, so 30v DC across them, they should "only" dissipate like 33mW each. Are quarters really THAT cramped that they can't dissipate even 1/3rd of their nominal 0.1W rating? Or half, if we're taking the 1/16W mentioned on the schematic...

 

[RoadrunnerOZ]

It could be heat transfer from the IC above though the legs and feedthroughs to the resistors below adding to their thermal breakdown when the IC is overdriven and heats up. I can’t find a really good explanation but I do know that when driven hard the chips get really hot.