HELP! - Soundcraft 2400 Sudden Meterbridge weirdness!

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Dr nEon

Well-known member
Joined
Jan 11, 2006
Messages
232
Location
Grate Britain
Hi Guys,

I'm halfway thru mixing an album,  and yesterday morning I switch on the board and see that all my VU meters are not resting at -infinity (..  they are all reading  around -17dB .  Sitting static at that level,  with no DAW playing.. 

As soon as I printed my mix, I pulled everything from the patchbay...and with the board switched off, I worked all the switches on the board..just to try to rule out the obvious. . when I switched the board back on with faders down and nothing playing, the meters were just the same... all metering a constant -17dB, instead of resting on their endstop below the -20dB marking.

I don't have my scope to hand and I'm in the middle of a session, but I'm looking for any advice on what may be causing this.  There doesn't seen to be any noticeable problem with the audio signal, so I'm hoping to be able to continue mixing....But it is pretty unnerving seeing something wrong , like that. I should add that when I play the DAW , the audio signals are appearing  on the meters..

I wondered if it is probably power supply related since it is common to all the meters.  The Soundcraft 2400 manual has great schematics, but it is rather hopeless in this department.  There is no info or diagrams for the meterbridge,  or the power distribution to the meterbridge,  and I'm not even sure which voltage rail feeds these mechanical meters.  Online, the only info I found relates to the LED bargraph version which ran on 24v.  Not sure that this is the case here, though...  My 2400 manual only explains how to change the meter bulbs!  Ho humm....

So being with a client, mid-project,  I'm in a position where I'm not immediately able to fully investigate , like I usually would,  and so I'd be grateful for any insights from anyone who is familiar with this version of the desk. 

I did snoop around for schematics from earlier Soundcraft boards with similar looking meters, but didn't find anything relevant.

Any help would be most appreciated.

Many thanks in advance,

nEon
 
yes a common fault sounds like PS (or ground issue).

Perhaps start by measuring PS voltages in meter bridge and console, maybe check that meter bridge ground is same as console ground 0V.

JR
 
Thanks so much for the advice, John!

I'll take some time out tomorrow to investigate the ps voltages and grounds

For now I have just discovered a couple of other symptoms which may be related...

I've just set up to overdub a ride cymb with my  trusty AKG451,  using a Soundcraft mic channel,  and the recorded waveform is massively negatively biased... Or you could say the positive going wave looks suppressed.  I've attached a pic.    I record this combo of cymb/mic/mic amp  quite often, and it is usually symmetrical.  This looks weird...

FWIW, trying other desk channels yielded the same weird result, so I quickly repeated the cymb part without changing anything other than patching thru an outboard Aphex mic amp, instead of the Soundcraft, and the resulting waveform was a normal symmetrical pattern, with no exaggerated transients.

Another thing which I've noticed while setting up the next track, is that many of the returns from my DAW are causing a faint crackling distorted breakthrough onto the mix bus...  I soloed an unused channel to kill all direct audio, and could hear pretty much every channel crackling on all the transients.    I reduced all channel line gains down from zero, down to between -5 and -10dB  , and this reduced most of the breakthrough..  This is VERY abnormal for this desk!  It's as if there's a reduced headroom. 

I'm old skool, and I always double check the gain structure from my DAW into the desk with line-up tones, at the start of every mix session...The onset of this breakthrough is happening at signal levels way below where I usually comfortably work, and way way below the level required to illuminate a channel peak light. 

So I'm starting to suspect the main audio rails...  But what do you make of this John?  I'd be interested to know your thoughts!

Many thanks in advance,

nEon.
 

Attachments

  • 451 SOUNDCRAFT mic amp.png
    451 SOUNDCRAFT mic amp.png
    46.8 KB · Views: 10
I'm not John but strongly suggest you not to power on the console 'till you check all voltages of the power supply.
 
moamps, pucho...  thanks guys!

I am going to check the psu voltages first thing in the morning.    I will also attempt to take some measurements inside the meterbridge.  Unfortuntely the desk lives up against a wall, so I can't access the meterbridge from behind, but  the front panel is modular, with 8 meters on each of two strip. Hoping to fold those down, and get some kinda access to the meter driver cards...

Thanks again for your wise words.

nEon.
 
Well guys, I've checked the power supply , unloaded , just with a DMM (no scope yet),  and all the voltages look correct at the psu outputs . I have +/- 7.5v , +/- 18v, +24v, and +48v rails. 

However, as soon as I connect the desk and power up again, the +7.5v rail at the psu output  is reading +0.745v  so correct me if I'm wrong, but this looks very suspect to me... and I've a feeling that I may have to disconnect batches of channels to try and narrow down the suspects as to whats causing this loading down.  Does this sound like a sensible approach?  I'm keeping the power-up periods very brief.  All the other voltage rails appear to be fine.

Next though, is to follow John's advice and take a look behind the meter panels, and check voltage there.

Weird thing is.. I believe the +/- 7.5v rails are for logic, and yet the solo and master mute switching is all working fine...

nEon
 
Thanks John...that makes sense.  For now I'm removing  the meterbridge panels..  as per your first suggestion.  There's a panel on either side of the stereo master meters... each panel holds eight vu meters...  If I patch a tone to any of the first eight on the left panel, and switch the tone  on and off...the ballistics of those meters appears normal.  However, on all of the meters on the right panel (9-16)  when I apply the same conditions, with the exact same tone gain level, the meters are rising quite sluggishly, and taking a second to settle.. they are settling at a reading about 4dB higher than the first eight meters on the left panel .  Also when the tone is switched off, they are not falling back quickly to rest, as a usual VU ballistic... They are falling slowly...about the speed of a LA2A or 1176 compressor on medium-slow release.  This is definitely odd, so I will start by removing that whole meter panel , and checking the voltage again... and then yeah, start pulling all the channels as you've suggested...

I'm thankful that the board is modular, and that this approach can be taken!  I once acquired a Racal HF receiver which was DOA, with an apparently dead PSU, until I pulled all the boards, found the PSU to be OK, and then discovered that most of the Thompson tantalums throughout the (logic controlled) set had failed s/c and smoked other local components.  Satisfying feeling when that one was nursed back to life.      This will be even more satisfying when we've cracked this one!

Many thanks for your ongoing help, John!

Cheers
nEon.
 
Further Update....

I haven't yet pulled all the channels  completely from the Chassis, which involves removing a few ribbon cables from each channel,  but I have pulled all of the ribbon connectors which carry power to the channels.

The problem remains.  With all of the harness disconnected from the channels,  I can probe any of the ribbon connectors with the DMM, and read a healthy -7.5von the negative rail,  but an unhealthy + 0.7v on the positive rail.  If I do a resistance check on the ribbon harness, with the power off,  between 0v line and the positive rail, I get a stable 0.05 ohms .  Seems  like almost a dead short somewhere...

I am now going to pull a bunch of channels out to access where the ribbon harness meets the incoming cannon multipin connector under the centre of the desk,  to investigate the integrity of the connections.

nEon
 
-7.5 on the negative rail wont be right. It should probably be between 15 and 17 volts. So suspect your power supply.
First, get the full manual. Then, study the power supply. Unplug the power supply from the console and check output voltages with no load. Then apply a load.
 
Dr nEon said:
If I do a resistance check on the ribbon harness, with the power off,  between 0v line and the positive rail, I get a stable 0.05 ohms .  Seems  like almost a dead short somewhere...

Yes. It's good thing you located the fault at entrance. Was the power cable disconnected from the PSU during the measurement?
 
radardoug said:
-7.5 on the negative rail wont be right. It should probably be between 15 and 17 volts. So suspect your power supply.
First, get the full manual. Then, study the power supply. Unplug the power supply from the console and check output voltages with no load. Then apply a load.

Please, reread reply 6.
 
OK guys, thanks for the continued replies...

Yes moamps...The ribbon harnesses are totally free of the channels, and the psu connector is disconnected from the psu, at the desk end. They still read a short on the +ve 7.5v line.

So I've taken out the masterbuss  channel, and a few channels either side to access the distribution boards which are mounted down in the chassis . These two ribbon cables both terminate at the same board..it can be seen in the attached pic...

Ahaa... there's a pair of tantalums in there...!

The board isn't marked, but near the tantalums are red and black leads.  Using resistance check, these correspond to my +/- 7.5 points on the ribbons.  So on the terminal block I've checked between earth and the +/- terminals , and the same short is there between earth and the red terminal.  Next, I removed the red wire.. In the pic I have already disconnected the feed in, and checked earth to red lead, and now it shows a nice large resistance and no short! So the short is most likely on this board, and since there is a tantalum across that point, I'm hopeful that this may be the cause...in the same manner as my Racal story earlier.

I'm now looking at carefully removing that board , to get a closer look, and perhaps desolder a leg on that tant to check it..

Fingers crossed,

nEon.


P.S.  Sorry for the stupidly large pic file size!  A mistake!
 

Attachments

  • Soundcraft2400 distribution.jpg
    Soundcraft2400 distribution.jpg
    1.8 MB · Views: 17
Bingo!...

...Gently wiggled the distribution card from it's adhesive standoffs,  to allow a closer inspection, and seeing the +ve rail tantalum cap from the side reveals the tell tale sign!

pucho812 said:
tants usually do not fail, lytics fail as they dry out over time.  but usually is not always so it's possible the ans need replacement

Pucho812 I agree, they are generally regarded as reliable, nowadays, and not necessarily the first suspect in fault-finding.  However, my experience with rennovating a RACAL RA1792,  which I mentioned in a reply above,  involved the replacement of most of the Tantalums in the set.  This was Racals first microprocessor controlled radio in the late seventies, and was full of light blue coloured Thomson-csf  Tantalums.  The set was dead due to numerous tants which had failed short.  Many had black scorch mark or holes in the side, exactly like this one.  Apparently, RACAL service engineers just take a shotgun approach, and  replace them all on sight....so untrustworthy are they!    I guess this is an old desk, only a few years younger than the RACAL, and these tants are in quite a demanding position, so it's fair enough.. 

Funnily enough, not long after I fixed the Racal set, I was mixing an album at Miloco in London, and when we came to switch on the half-inch ampex, to line up, one channel was dead, with no metering...we just called in the tech , and he quickly found the fault on the Rch card to be a failed tant...    Maybe certain brands from back then are less reliable, with age... Thomson CSF sure seem to be one of them, anyway..

Cheers and thanks,
nEon
 

Attachments

  • Soundcraft Distribution Board Tantalum.jpg
    Soundcraft Distribution Board Tantalum.jpg
    427.3 KB · Views: 14
OK guys...

This tantalum capacitor is definitely the culprit..It is s/c!  With it removed, a resistance check on the 7.5v rail now shows no short.  Good vibes!

Interestingly, looking at the pic, that's a Thomson logo on the cap...!

This cap is 4u7 35v.  I plan to replace both caps , even though the one on the neg rail is OK.

I don't have that specific value in stock...  As sods law dictates, I have loads of values around that figure, except the one which I need.    I will order replacements from Farnell tonight, which will be here by Tuesday.

However in the meantime,  I do have some brand new 6u8 35v tants in stock, and I wonder if it would be safe to temporarily sub with those, in order that I can move on and power up  the rest of the desk... that is check to see that no other damage or faults have been induced by the shorted rail.  It would be nice to be able to check this out as soon as possible,  but I'll happily wait for the 4u7s to arrive, before powering up,  if y'all think a 6u8 sub is a bad idea......???

Many thanks in advance,

nEon.
 

Attachments

  • Soundcraft 2400 Failed Tantalum.jpg
    Soundcraft 2400 Failed Tantalum.jpg
    922.8 KB · Views: 2
JohnRoberts said:
Modern low ESR aluminum is probably an OK substitute for tantalum, with less risk of future melt down. 

JR


Thanks John!    I have these 4u7 50v Dubilier 'Low impedance' caps in stock.  They all measure around 2.6ohms esr (see attached pic)... Do you think that's adequately low enough to replace the tants?    .... Or should I be buying something like a Panasonic FR, say, instead?

Cheers

nEon 
 

Attachments

  • Dubilier Possible Tantalum Replacement.jpg
    Dubilier Possible Tantalum Replacement.jpg
    505.2 KB · Views: 9
Dr nEon said:
Thanks John!    I have these 4u7 50v Dubilier Low impedance caps in stock.  They all measure around 2.6ohms esr (see attached pic)... Do you think that's adequately low enough to replace the tants?   

Cheers

nEon
I am not intimately familiar with why they specified the tantalum caps in the first place. Perhaps it was just considered good practice at the time, and not addressing a specific problem.

I would say just try them and see what happens , worst case you can add ceramic disc in parallel, but lets see what happens. (In case it isn't obvious I am not a fan of tantalum caps.)

It's your console so do what makes you most comfortable.

JR
 

Latest posts

Back
Top