Mixer is dying. Time for a mod.

[Mylithra]

Got the NE5532s and some new brown dog adapters. After everything I went through the first time to get the chip on/off/on/off the board, I decided to use DIP scokets on the brown dogs. Soldered everything up and installed the board.
It doesnt have the RF noise in out the control room anymore. I havent buttoned it all back together yet though. Im going to let the mixer run for 24hrs and check back on it later, but as of now, everything looks to be working as I expect it to.

I do hear something different between the modified and unmodified channels at the output but Ill need to do more extensive listening test after Im certain I've licked this problem.

Thanks for the great suggestion Martin.
Jeff

 

[smallbutfine]

:razz: Fine!
Almost always where I've seen 4560/4580 opamps they were the cheap NE5534/2 supplement to cut cost in otherwise OK circuits. Often things are designed with NE5532 and the final version is a simple financial compromise to cut a few cent per unit in mass production.
I guess the designers are not involved in the final decisions that are made from management - please note that 4580 looks fine from datasheet but sounds less than great in most applications....


:guinness: :thumb: I'm very glad it works for you without that big investigation that 2227's would mean....!

Thanks for the kind words and best regards,

Martin

 

[Mylithra]

I finsihed the project.
I ended up replacing all the OPA2228s with NE5532s.
fired everything up and it works good now.
For an initial listening, I couldnt hear much difference if any between the Modified and unmodified channels but one thing I did notice,
The mixer was TONS quieter. Stock the mixer alway has an annoying hiss to it, albeit a low hiss, still annoying.
Now I dont hear that anymore.

I dont know that Ill be able to do the whole thing as I dont think there will be enough power supply to handle it. There may be a power supply mod in the future and I might do the rest of the mixer then. I still have a bit of learning to do in figuring out current draws based on how the circuit is designed. Any recommendations for good reading matierial on this subject would be great.

Once I get some time, Ill post a couple of clips of some Acoustic guitar through this thing and let you hear the before and after clips.
Once again, thanks everyone for your help.
Jeff

 

[smallbutfine]

Hi again,
I just checked old datasheets and assume now that 5532 nearly doubles the slew rate from 5 V/us to 9 V/us, which is quite a big difference/upgrade.
While nearly maintaining sonic character, you may not onlx notice lower hiss but also better transient response.
(this is not that obvious to hear in the first tests - but a huge improvement the more channels you might use!)
Power dissipation may be increased nearly the same factor from my datasheets, BUT i found that it really depends closely from the package type!
SSOP 4580 dissipates only half of the DIP type - and I assume it's the same with 5532 where I only have an old njm datasheet at hand...
- I would recommend to change all 4580 to 5532 smd form factor without any fear....:guinness: :thumb:
Nevertheless, upgrading psu is always a food idea on budget gear....but you might get away well without it in this case!
Have much fun with your quiter desk.:guinness: :thumb:
Always welcome, I'm glad my recommendation was of help.

Kind regards,
Martin :guinness: :thumb:

 

[smallbutfine]

P.S.: A quick thumb rule for current draw calculations:
current draw increase x chip count = current draw increase at the psu !

Good read on this topic (and many others) is Eddie Cilettis webpage tangible-technology!!! (A classic for beginners of pro-audio diy!)
Timeless and Highly recommended! :thumb: :thumb: :thumb:

Kind regards,
Martin

 

[Mylithra]

Sorry for the long delay. Here are the promised Clips.
The Setup:
The only Mics I have in a matched pair are my pair of NT5s so those were used. I placed the mics about 14" away from the guitar, one on top of the other, 1 angled the bottom one slightly up, the other slightly down. (The idea to get the same spot on the instrument.) The mic goes straight into the mixer. I have no Compression, EQ, or effects on. This is a dry as it gets. It goes Mixer out via Light Pip into the computer via a Fronteir Design Dakota. I record with Reaper.

This is the mixer stock.
http://goose-tracks.com/thegeekgoose/clips/njm4580.wav

This is with the channels modified by replacing the Mic pre and the opamp that drives the A/D converter with an NE5532
http://goose-tracks.com/thegeekgoose/clips/NE5532.wav

 

[[email protected]]

Sounds really good. Good work.
I may have to take on this venture. I wonder if there is a better opamp then the 5532 I can try
I wonder if the PSU could handle 8 new opamps.
Maybe I'm not understaning something here but, I thought it was better to have a faster slew rate. Isn't slew rate how fast the output can react to the input signal?
Meaning the faster it can follow the curves, the more accurate it can reproduce the output?
I might have to ask some questions when I am ready.

Thanks
A

 

[gswan]

OK.. I started getting into the whole rewire of the studio.. Got the basics hooked up. (1 output from the Computer in via Light pipe, the Word Clock and the control room out. At random, it will start having this high frequency noise.. My guess would be something akin to RF. I thought it might be because the new opamps were oscillating though Ive never heard an oscillating opamp so Im not certain what one sounds like. I attempted to put in some decoupling caps, .1uf 50v ceramic caps between the the V + in and the V- in but then it sounded very weak and distorted. Im guessing I got the decoupling wrong.

Did I use the wrong values, did I just get the decoupling wrong or could it be something else like one of the amps is bad?

You got it completely wrong.
Decoupling of the supply rail does not go in series with the supply rail, it is placed between the supply rail and GND as close as possible to the component requiring the decoupling.

 

[gswan]

I have a Behringer DDX3216 that Ive had for a few years now. Today during my band's practice, the entire mix started cutting out. Being that the alternative is buying a new mixer (which is $$$ I dont have at the moment) I decided to attempt to fix it and while Im at it, Mod the thing.

Since it was affecting the overall mix and none of the meters showed a signal dropout, I have to assume the problem is on the Analog Out board.

It could be any number of things but I havent dug deeper into it yet. At this point.. I think it might just be better to see if I can replace the entire Analog output board. I also thought it could just simply be one of these opamps has gone bad in which case, the next bit might solve that problem for me.

It sounds like a switch or connector problem from your symptoms. I would be chasing electro mechanical problems before changing components.

As to the modding,
Assuming I can fix the output problem, The only real mods I thought about doing to this was getting rid of all the NJM4580s in favor of the OPA 2228 and replacing all signal path caps with Panasonic. The A/D Converters are AKM5383s. I use the BLA Microclock as an external clock source so I wont be screwing the clock or the converters on this.

I doubt that replacing the signal path caps will make any difference whatsoever. large value PSU caps may if they are getting old.

Op amp replacement may help in certain places, depending on the gain and use of the op amp in the circuit.

Unless you really need to use an external clock to sync with other gear, the internal clock will almost always be more accurate and stable than an external clock.

 

[s2udio]

On the subject of noise in this mixer,just cracked mine open for a service (now 7yrs old)
Making some strange noises when powering up ,hissing noise from the psu and
audible digital and analogue garbage on the outputs with all channels muted,this subsides to
normal operation after a 15 min warm up.
I suspect the SM PSU caps are ageing ?
On opening up the psu and extracting the pcb,at first glance yup,lots of bulged and shrinking plastic
covered electros in the +- 17v section,6 x 1000mf 25v caps which all tested at about 400-600mf !!
Time for a complete recap,although most tested fine in the two other sections
will post some pics when i've finished.
All in all this seems a very robust unit,and very serviceable for a SMPSU

 

[gswan]

On the subject of noise in this mixer,just cracked mine open for a service (now 7yrs old)
Making some strange noises when powering up ,hissing noise from the psu and
audible digital and analogue garbage on the outputs with all channels muted,this subsides to
normal operation after a 15 min warm up.
I suspect the SM PSU caps are ageing ?
On opening up the psu and extracting the pcb,at first glance yup,lots of bulged and shrinking plastic
covered electros in the +- 17v section,6 x 1000mf 25v caps which all tested at about 400-600mf !!

The capacitance value is not a valid test for ageing/dry electrolytics. ESR is what you need to measure. The ESR (equivalent series resistance) determines the capcitor's effectiveness to charge.

Many times have I seen electro's whose capcitance is within tolerance of the value stated and way out of tolerance for ESR.

 

[s2udio]

The capacitance value is not a valid test for ageing/dry electrolytics. ESR is what you need to measure. The ESR (equivalent series resistance) determines the capcitor's effectiveness to charge.
Many times have I seen electro's whose capcitance is within tolerance of the value stated and way out of tolerance for ESR.

Yes i agree its not a totally valued test,(I do understand the ESR point)however these caps are definately near to popping there tops
and my cap meter has never told a lie yet,i of course tested the suspects against known good caps
of the same value to check for errors.
Also i tested the caps by heating (very briefly)with a shot of hot air,the value climed by 100mf very quickly
from 350mf,where a good new one hardly changed at all ,these caps are FAULTY!
A 1000mf cap should not read 350mf..........Thats a tolerance of 70%!  no disrespect of course
will report on the outcome

 

[Mylithra]

I thought I had made a reply here some time back but I guess It didnt make it here. I ended up replacing all the opamps in the mixer with NE5532s. My cutout problem wasnt simply a bad connector. If that was the problem, It would have only done it on the main outs but it was doing this across all out simultainuously. Since the output board was replaced and I did the mods, I havent had a problem since. Since I did the mods, Instruments stack much better, Transients have improved and the mixer is overall quieter.

 

[s2udio]

Seems like a success then will be moding mine soon but probably only the A outs as I use my outbord pres
as direct inserts.
However if the PSU starts whistleing from cold a PSU recap is advised

 

[Mylithra]

Yeah.. I thought about just recapping the PS just to be on the safe side. Got to be careful when modding this.. Theres only 1A on the +-17v rails so its easy to hit the ceiling of what the PS is capable of.

 

[antone]

Hi
Did you ever think about modding/ upgrading the preamps on the board??
Just curious

Thanks
A

 

[s2udio]

Hi
Did you ever think about modding/ upgrading the preamps on the board??
Just curious

Thanks
A

Have you read the whole thread ?

BTW ,PSU recap went fine will update with pics soon
Mixer now runs a lot better with no noisy digital garbage on main outs or psu squeeling on startup !

 

[antone]

My apologies must have missed it. I though he only did the Analog out board opamps.
A

 

[Mylithra]

heh.. No.. I replaced ever NJM JRC4580  OPamp in the thing except the one that runs the headphones.

 

[antone]

Did you end up changing any of the caps??
Some one mentioned they recapped the power supply.

Thanks
A