soundcraft ghost console modification

[angelothewolf]

Im "upgrading'' a souncraft ghost console. Caps and opamps. the stock opamps are TL072 and NE5532, im replacing those with OPA2604 and LME49860.
The idea is to replace that components on the entire console. Now i have done half (16 ch) and its everything ok except that im getting some little oscilations i think, it happens with high gain (gain trim on 50-60db) its like a copy of the input signal but faint and distorted on the background, almost imperceptible. that issue leaks through channels even muted.

The other think is power supply, there is not too much supply current difference between ne5532 and lme49860 (10-13mA) but there is a significant difference between tl072 (2,5mA) and opa2604 (10,5mA), so in theory i need like a 1,7 extra amps for this chips, is that right?
Now im running the console with 16 ch upgraded and 16ch stock with no power problems, but i think that will be an issue when i do the other channels. Any tips?



btw, the PSU claims 5.25A per rail +/-17V

[Mark Black]

Most of the Soundcraft consoles I have worked on have suffered from some form of problems related to the IDC ribbon cables and/or header pins. In some cases it has been oxidation or looseness at the point where the wire is pressed into the connector and in others I have actually found the header pins to have developed a problem with the plating at point of connector contact. Or both. I chock this up to thermal cycling and/or environment on boards that haven't seen road or club use, possibly aggravated by using tin rather than gold-plated connectors & pins.

The kind of mod you are doing is one I would not take lightly given the possibilities for new interactions in a complex environment (as you may be finding). Best of luck to you though!

[The Dude]

I've found the same ribbon cable problems. To add: I've also seen several of the larger consoles with burnt ribbon cables on the wires for the +-15V lines. There were no shorted components on any of these. It's my belief that the ribbon cable wires are not large enough for the amount of current drawn by the larger consoles. In these cases, rather than replace the ribbon cable with another one that will likely develop the same problems, I run jumpers across the break AND to both ends of the console. This achieves 2 things. 1) it takes some of the current away from the tiny ribbon cable. 2) If the cable should burn up again in the middle somewhere, the +-15 has another path to get to all of the channel cards so that parts of the board are not rendered inoperable. There's not much joy in loosing half of your console mid-show.

[km6xz]

That is a pretty big mod, in time and money not to have done some research first. Why are you suspecting the console needed high current opamps. Did you prototype a channel and measure the differences due to opamp changes? 5532 opamps are not a problem, and perform very well in a wide variety of applications much higher end than that board. Neve uses them in their well regarded 8816 mixer and it will run circles around your modded Ghost. Yes, expect oscillations and diminished performance when inserting the 49860 chips into a circuit expecting different operating characteristics, so one is going to be able to help you in that. What are you trying to accomplish and why were you so certain you could violate the engineering truism of "there is no free lunch" which expresses the general axiom that there is not modification that yields old positive results. All design is a trade off of various compromises, so what are you willing to give up in order to get what?
The stages that use the 5532 are stages designed around them and and use their characteristics well, like high slew rate needed in high gain stages and eq. The lowly TL072 is used in stages that do not require the low noise that high gain stages do.
The Ghost is not a terrible board, in fact some well done and well regarded albums have been done on them. It is certainly better than 99.9% of the users and their songs put through it. What is the goal and what exactly are you intending this mod to result in? Did you take the baseline measurements needed to compare before and after? What parameter were you unhappy with where you measured deficient performance? The faint signal leakage is called "Crosstalk", what measured level did it have before adding the high current amps? Cross talk is one of of the downsides to high current chips...as per the "no free lunch" rule.

[MarkusBass]

Now i have done half (16 ch) and its everything ok except that I'm getting some little oscillations i think, it happens with high gain (gain trim on 50-60db) its like a copy of the input signal but faint and distorted on the background, almost imperceptible. that issue leaks through channels even muted.

This is exactly what was expected. You changed the opamp and you ignored the frequency response of the gain stage. You cannot change opamps without tackling the problem of greatly extended bandwidth. Sometimes the change is from 1MHz to 20MHz. This causes that the mixer is prone to oscillations and in fact it sounds worse than before the upgrade. I would change only the most critical opamps and I would also check the frequency response (whether it may cause oscillations).

The other think is power supply, there is not too much supply current difference between ne5532 and lme49860 (10-13mA) but there is a significant difference between tl072 (2,5mA) and opa2604 (10,5mA), so in theory i need like a 1,7 extra amps for this chips, is that right?
Now I'm running the console with 16 ch upgraded and 16ch stock with no power problems, but i think that will be an issue when i do the other channels. Any tips?
btw, the PSU claims 5.25A per rail +/-17V

This is another problem. You may need 4 times more current than before. I doubt whether the power supply can deliver it. I've seen cases where the transformer was not able to deliver enough current/voltage. In extreme case there was less than +/-15V before 7815/7915 ICs. So the stablisators didn't work at all. The final result in such cases may be much worse than with plain TL072.
So you may do the upgrade but along with the upgrade you have to perform a lot of tests. Otherwise, you just loose money.

Mark

[dai h.]

might be a good idea with this sort of thing to do one channel only, then confirm it's working as it should and is an actual improvement before proceeding. I'm no expert, but my (vague) understanding is that the quality gained from an op amp depends on things such as how much gain is asked from it (less gain asked = more likely to be linear), layout, grounding, power supply bypassing, suitability (in terms of noise, drive capability, power consumption, plus whatever else). As I understand it, things do not automatically improve just because an older, less trendier op amp is replaced with something newer, more fashionable etc.

I've done a limited amount of "messing about", and my impression (from my experimentation and reading) is that sometimes there could be a slight bit (or noticable if the mixer is not something very good in the first place) of improvement from matching components better at mic inputs (for increased common mode rejection), bit cleaner EQ by changing to low distortion types (polypropylene, NP0 ceramic, polyphenylene sulfide, or whatever) if the ones there are mylar, maybe a bit of improvement by judiciously decreasing ground impedance in places(just adding a bit more copper--presuming the foil is normally thin 35um thickness), increasing size of alu electros so there is less voltage drop across them (= less distortion). Possibly relocating Rs and Cs closer to the op amp to make loops more compact (might be pointless if the board is already a surface mount type).

And (maybe) a sense of perspective could be useful as well, since doing (presumably) successful mods don't make a poor song or performance better. I suppose it might be more about helping to make sure the gear doesn't "get in the way" by having too much noise or not being able to provide enough gain for a low output mic or whatever.

[angelothewolf]

Ok, thanks everyone for sharing your knowledge, im socketing all the opamps and yes i try diferent combinations and experimeted with the opamps on one channel before. Replacing the mic pre with the opa2604 makes a perceptible and good difference in sound, now i dont hear any improvement replacing the opamps on the EQ and the rest of the channel strip other than oscilation issues. im replacing all the electrolitics with ELNA SILMIC and this seems to add an improvement on noise and low freq response.
So now i decided to just replace the preamp chip and caps, this gives me a good upgrade to my ears and this will need only about 400mA extra on the PSU.
This is a job for a client and i have little time to do it,

So the only issue now is the oscillation due to the extended bandwidth, would be very helfull to know how can i mod this stage to solve this problem

[Jazz P Bass]

One question that came to mind:
Is this 'upgrade' at the customers request?
If so, I would contact them & explain why it may not be such a good idea.
As Stan pointed out, the mixer was designed around specific components with very good results.
No amount of IC cork sniffing will improve it without a redisign.

If the 'upgrade' was your idea I would rethink it as you honestly are not providing the customer with any real benefits.

Power supply, coupling capacitor rework I could see.

Aside: in my humble opinion the sockets are a bad idea.
You went from a soldered connection to a compression connection.

[g1]

Keep in mind, a few posters in this thread have mentioned problems with the connectors and interconnect cabling. Asking those connectors to carry extra current can only amplify the problem.

[angelothewolf]

well this was a costumer idea, he search on forums and buy the specific opamps (opa2604 & lme49860) after this experiments i get a to a good result just replacing the preamp opamp with the opa2604 and all electrolitic caps... this was the only upgrade i could hear without compromising too much the circuit design, just replacing the pre opamp i got no oscillations at all, and the crosstalk is within specs. i dont think this would be too bad for the PSU too, im thinking to test this opamps on master section and see what happens.
i place sockets on a few channels just to experiment more easily,now im soldering all the chips...
there's a lot of threads on the net about mods on this console but no one with details and conclusions. i dont hear any difference between ne5532 and the LME49860. So now, for the time i have to do this job, the only honest upgrade i can do is the preamp chip and caps...

Now im curious about the creation audio labs mod for this console wich is like a $6000>

Creation Audio Labs

[angelothewolf]

Keep in mind, a few posters in this thread have mentioned problems with the connectors and interconnect cabling. Asking those connectors to carry extra current can only amplify the problem.

Yes, but the board is in brand new condition!

[The Dude]

Yes, but the board is in brand new condition!

That matters little. These cables have inherent problems not exclusive to Soundcraft. Mackie and others have had their share of ribbon cable problems, too. Funny how things work. I just (today) got a Soundcraft FX16ii in for repair. Different and smaller board, but similar problems. It, too, is in "brand new condition". Symptoms were: intermittent noise at the outputs, channel 3 intermittently dead, channels 12 & 16 = intermittent loud static noise, banging on the top of board will make symptoms go/stop. All symptoms were traced to ribbon cable connection issues.

What I believe g-one is saying is that ribbon cables have proven to be at least on the low side of capable as far as handling the current drawn by the mixer as it is stock. Installing op amps that draw more current may stress the cables even further. I'd have to agree based on past experience.

[angelothewolf]

That matters little. These cables have inherent problems not exclusive to Soundcraft. Mackie and others have had their share of ribbon cable problems, too. Funny how things work. I just (today) got a Soundcraft FX16ii in for repair. It, too, is in "brand new condition". Symptoms were: intermittent noise at the outputs, channel 3 intermittently dead, channels 12 & 16 = intermittent loud static noise, banging on the top of board will make symptoms go/stop. All symptoms were traced to ribbon cable connection issues.

What I believe g-one is saying is that ribbon cables have proven to be at least on the low side of capable as far as handling the current drawn by the mixer as it is stock. Installing op amps that draw more current may stress the cables even further. I'd have to agree based on past experience.

im not having any problems with ribbon cables, im adding less than 10mA extra per channel without issues. if i had change all the opamps in the channel strip as it was the first idea, it would had been a problem!

[g1]

Sorry, I didn't realize you had revised your upgrade. So you are leaving the TL072's alone? Or 5532's?

[Jazz P Bass]

well this was a costumer idea, he search on forums and buy the specific opamps (opa2604 & lme49860) after this experiments i get a to a good result just replacing the preamp opamp with the opa2604 and all electrolitic caps... this was the only upgrade i could hear without compromising too much the circuit design, just replacing the pre opamp i got no oscillations at all, and the crosstalk is within specs. i dont think this would be too bad for the PSU too, im thinking to test this opamps on master section and see what happens.
i place sockets on a few channels just to experiment more easily,now im soldering all the chips...
there's a lot of threads on the net about mods on this console but no one with details and conclusions. i dont hear any difference between ne5532 and the LME49860. So now, for the time i have to do this job, the only honest upgrade i can do is the preamp chip and caps...

Now im curious about the creation audio labs mod for this console wich is like a $6000>

Creation Audio Labs

I knew that Creation Audio Labs was involved somehow.