njm4558s op amp replacement?

[Synthetech]

EDIT-

Believe it or not, I found out the source of the K3 clipping was in fact my mixer aux send was overdriven for the channel the K3 was on. Doh!




I have a Kawai K3 that begins to clip the audio output when volume is past half way.

There are two 4558s's in the audiopath right before the line output jacks.




I suspect after some searching that these are not the best opamps put in gear.

Is it possible these opamps are the source of my clipping issues?



Anyone have any ideas if I might be able to simply swap in njm2068's that are SIP's like the ones in my K3?

[madtheory]

Actually 4558 are perfectly fine, although there is a lot of c**p spouted about them on the tinternots. An error of causation. If you want proof, see Douglas Self's book "Small Signal Audio Design". Or you could just count the amount of audio gear you own that uses them. Or look inside an Audio Precision analyzer- lots of 4558 in there.

If the thing is clipping then that's a fault, nothing to do with the 4558 (because they're part of the design, and AFAIK the K3 is not know for clipping). Personally I would clean/check the relays first, after that maybe look for open/ short circuits caps in the feedback networks.

[Synthetech]

Thanks for the reply.

Yea.. I took a look at a Korg DW8000 filter board I have laying around and noticed it has a 4558s next to each filter chip.


I only see the one relay. I doubt that is the source of my clipping.

What are "feedback networks"?
Am I looking for a bad electrolytic cap or caps?

Any other suggestions as to where to look for the part causing my clipped audio?
It's tolerable to work with, I just feel it should be a bit louder before it clips out in me.
Seems my VU monitor in my DAW shows it peaked out when it clips, but I don't even have the K3's volume up past 3/4 full.

[madtheory]

Why do you doubt the relay? These can and do cause clipping. It's the first thing I would check, but if you think otherwise I'd like to hear your views on it.

Do you have the service manual? Could just be a calibration issue.

Feedback- fundamental to amplifier design. Here's a good article:
http://www.edn.com/design/consumer/4418 ... s-too-much
So once you've ascertained where the feedback part of the amp circuit is, check the components there.

[Steve Jones]

Have you checked that the analog power supply is symmetrical?

[Synthetech]

Thanks for your advice.

What would I be looking for to indicate the relay is cause of clipping? I thought it was just a simple switch that passes audio to the outputs when the unit powers on.. according to the schematic clip I attached above.

Yes, I have the service manual.

It is here
http://www.soundprogramming.net/manuals ... Manual.pdf

Any idea where calibration info is located in it?

What is symmetrical in the power supply that I am checking for?

[madtheory]

Dirt/ dust in the relay makes it switch intermittently. Sounds just like clipping.

But listen to Steve first

Cal info? Looking at the contents, my guess would be "Check Points for repair" and the part about Test Tones.

Power supply? Here you go:
http://lmgtfy.com/?q=symetrical+power+supply

[Synthetech]

Dirt/ dust in the relay makes it switch intermittently. Sounds just like clipping.

if the relay were faulty, then wouldnt it distort/clip at any volume setting?
it only does it when I push the gain up past halfway.

I'd assume if I bypass the relay and the relay was the fault, the clipping would cease, right?

Cal info? Looking at the contents, my guess would be "Check Points for repair" and the part about Test Tones.

Yea, I'll have to check into it. I do recall a trip pot nearby the opamps right at the line outputs.

Power supply? Here you go:
http://lmgtfy.com/?q=symetrical+power+supply

cute google trick. I been googlowned!
I think I was just looking for a simple answer like, check the +/- voltages for both 5 and 15v supplies to be as close to the same as possible. IE, if the 5v side is 4.98v on the + , it should be very close to that on the neg 5v supply too, around -4.98v.
Is that what was meant by "symmetrical"?

[sam]

It's handy to buy a signal tracer as you can follow the signal back to the clipping...it's not a matter of replacing the op amps.
You did check the voltages on the schematics. .?

[Steve Jones]

Keeping things simple, it is the first thing to do with any repair - Check the DC power supply rails with an oscilloscope. Preferably not a multimeter unless that is all you have. If the supply voltages are incorrect or there is ripple then this needs to be fixed first.

[rschnier]

Like others have said, at least check the power supply voltages with a voltmeter to make sure they're not reading low. If you absolutely cannot get an oscilloscope to check the power supply rail waveforms, then your best bet is to replace the filter capacitors and/or the bridge rectifier in the power supply. After years of use and heat, the caps tend to dry out and lose their filtering ability, effectively lowering the DC voltage on the rail. Rectifiers also can start to lose their "one-way" current carrying ability with age and heat, resulting in the same kind of effect. Those are fairly simple things to replace (or have replaced) that can't do anything but help the overall health of the unit in any case.

I've encountered so many units where the filter caps have dried out, and the bridge rect has gone soft...it's surprising how often this is a main problem. (The rectifier going soft and allowing bi-directional current flow also hastens the demise of the filter caps.)

[synthparts]

You can check for ripple on most multimeters by setting it to measure AC volts. I'd say anything over 100mV is too much.

[Synthetech]

Thanks for all the suggestions.

Yes, I have a Tektronix 466 scope to check voltage ripples and signals.

I need to diagnose it the next few days when I can take the K3 to my shop with a proper test bench. Doing it at home on the kitchen table drives the wife nuts.
Fortunately, the way Kawai built this keyboard makes for easy troubleshooting.

[Steve Jones]

Another really useful piece of test gear for situations like this is a capacitor decoupled audio signal probe. Quite easy to build, plenty of schematics on the web.

[sam]

Another really useful piece of test gear for situations like this is a capacitor decoupled audio signal probe. Quite easy to build, plenty of schematics on the web.

I found one on a car boot sale for 2 euros....best investment ever