replacing 4558s in a Lake-People c76 DA-converter

[hop.sing]

Hello
I am a total Newbie in modding things.
I just saw those 4558s, together with op275s in the outputstage, in my converter, which already sounds pretty good(better than my RME Multiface).
Can I savely replace them with better ones, without changing the circuit?
I would love to see, if I can hear a difference.
Any suggestions which opamps to try?
Here are the schematics: http://people.freenet.de/protein/lakepeople_dac_76.jpg
Thanks Tobias

 

[SonsOfThunder]

Tobias,

I would remove the 4558, put in machine-pin sockets and drop OP275 in that location as well. That's the easiest thing to do, especially if you are not so confident in your modding skills.

Hope that helps!
Charlie

 

[hop.sing]

Hi
There are already sockets installed. Very kind. But is there a reason that a not so cheap device like that converter (650 Euro for a 2 Channel DA) uses such a crappy opamp?
Would something like a opa2132 work also? They are more easily available for me.
Thanks Tobias

 

[gyraf]

..4558 is NOT crappy, if you treat it right and don't run it to it's limits.

For low-gain applications there is a good chance that you won't detect any audible difference with esoteric stuff..

The designers most probably knew what they were doing..

Jakob E.

 

[hop.sing]

Hi Jakob
Thanks for your reply.

..4558 is NOT crappy, if you treat it right and don't run it to it's limits.

"Crappy" is the the impression I got, reading forums like this. Lake people has a very good reputation here in germany and this thing actually sounds really good. I was just wondering if it could even be better....
Tobias

 

[SSLtech]

If they used 275s elsewhere, then they obviously weren't religeously against paying for 275s. They didn't use 4558s just to be cheap then.

unfortunately there's a lot of posting al over the internet by people who really want things reduced to simple terms (TL072=terrorist=bad... OPA274=freedom fighter=good) and it really isn;t that simple. Sadly the reasons why it isn't simple are even less simple, and as a result they're most often ignored, and the prevailing impression which spreads far and wide over the internet is that you can take any piece of gear, swap the "terrorist" chips out for "freedom fighters", and your new, democratically-goverened Behringer will be a sonic delight.

There are a few instances where early gear used things like 741s because they were new on the scene, and just about any chip will make an improvement. This is because the gear is probably from the days when there were only bad chips to be had!

My guess: -If the gear sounds good to begin with, it's not by accident. (I never met anyone who designed a great piece of gear completely by accident) The designer sounds like he (or she) knows that sometimes you have to use a more expensive chip, and is happy to do it. At a guess, I'd venture to suggest that you might not find any improvement by chip-swapping, although the socketed ICs mean that you are at perfect liberty to try, without having to break out the soldering iron.

If you do find no improvement, can you please post back... It'd be so refreshing to hear back from people that the designer actually got it right for a change... so many people register on forums like this and John Klett's forum and their forst post is "What chips sound better than what I have"... they go away and you never hear from them again... It all contributes to the "people out there are swapping chips and finding improvements" theory, when it usually is just not that simple.

By the way, Studer use 4558s in the audio path of their later 2" machines. There's a lot more half-informed opinion and poorly-tested ideas on the interenet than there are good designers! :wink:

Keith

 

[gyraf]

Thanks, Keef - that pretty much sums up my opinion - only I'm not as easy with words as you..

:thumb:

Jakob E.

 

[SSLtech]

...but you're a whole lot cleverer with tubes than I'm ever likely to be! :grin:

:thumb: Much respect :thumb:

Keith

 

[NewYorkDave]

Excellent post, Keef.

 

[Rossi]

I agree, Keith. Yet there are some instances where cheap chips were used not because the designer thought they would do the job nicely but because the sales department had to cut costs. I have three B*ringer Comp*sers, and all three use different opamps. One of them actually had 4558s for makeup gain, and it did sound better when I changed those for 5532s. I did, however, mod one channel first and compare it to the other. That's not to say you can transform a cheapo B*ringer into an SSL by swapping a couple of chips.

 

[SSLtech]

Right...

Most people seem to leap directly to replacing every chip in a console... There are many horror stories! For sanity's sake (and your wallet) just do one channel and leave ons as a control channel. For true unbiased comparison, have a good-eared colleague compare the two (or more) channels without telling them what's been done, or to which channel. Sometimes friends have been unable to hear something that I thought was a gigantic leap forward in quality, other times they've picked the unmodified channel, ...sometimes they even agree with me!!! :wink:

In the case of the Bear-ringer, I bet it didn't sound that good to begin with? Am I way off base there?

Things that affect op-amp choices? well, current drive, voltage noise and current noise considerations, DC offset, GBW product, There must be plenty more factors...

Keith

 

[Rossi]

Yes, that particular unit sounded a bit darkish, especially when the comp was working hard. I didn't have a schemo, so I compared it to the other two and read the THAT notes and data sheets. The other two units had different opamps next to the VCA chip. The oldest one had a BE027 (B*ringer selected or custom manufactured, nobody knows) and the newest one had a NJM4580 - which has become B's standard opamp since. The one I modded was the inbetween one. If I remember correctly THAT recommends the NE5532, so that's what I used. Besides, I like the good old 5532, and it's cheap of course. I don't think it makes sense to put expensive chips in a cheapo unit.

 

[SSLtech]

Nothing wrong with 5532s in the right place... Saint Rupert himself uses them

keith

 

[SonsOfThunder]

[quote author="NewYorkDave"]Excellent post, Keef.[/quote]

Sorry to ask you NYDave, but I am confused by:
What you posted here?

Though I realize that this is not the same application.

I agree that 4558 is not bad when used within its limits, but i did not want to discourage someone from experimenting. If you listen and like what you hear...cool! I certainly have learned a lot by trial and error changing out chips and caps and juggling values. I have only made one or two pieces of working gear into trash.

What is not clear to me is whether the input to this ckt is I or V. Probably V as I suspect that the first 4558 will not do the conversion from I to V. My experience with DACs is mostly with I output, in which case you need a pretty fast OA to do the conversion well.

Trying to be helpful and encourage some experimentation!
Charlie

 

[hop.sing]

Ijust wrote Lake-People a mail about the 4558 at this place of the circuit. They are a small comany, they send the schematics with the manual, maybe they will share their ideas.
Tobias

 

[smallbutfine]

the NJM4580 was mentioned. this one was dedicately designed to be a drop in replacement solution for the NJM4558 with better noise specs. Same for his predecessor NJM4560 (as my memry serves me right...). If B*ringer changed the chipping in their devices, this really doesn't mean much (same for the alesis 3630: mine is stuffed with the new THAT series, some were with the old one. just series differences in that cases and mostly undocumented, so the customer is pretty ***ed up when trying to buy a second unit of a device to have same sound and specs).
Another thing with my old cheap S*P*L vitalizers. They used TL07x which are not bad, but experimenting with opa chips showed a respectible drop in noise and distortion and a much better stereo imaging (ok, that one is my subjective experience). But this devices were relatively cheap ones from the early 90's so replacing with newer and quieter devices is maybe an obvious experimenting field.
The "very nice" processor is a much newer device (i think at least) so it would not be so obvious to go that way to "upgrade".
I would try a change whenever i think that the opamps were "state of the art" when the machine was built, but "trying" does not mean i expect other chips to be "better" in any way. (Except they are definately meant to be a drop-in replacement.)
But sometimes the unit gets another quality of sound (means *different*) and i stay with it cause i like it more than the original. Has nothing to do with noise in the first place.
I have a valley gatex and a dbx MC6 which i never thought of to mod just because i really like the way they sound...nevermind if it could be done.
Mods on the psu, changing for dedicated audio capacitors and sometimes changing electrolytics at the in's of cheaper but interesting line and instrument effects for better headroom is nearly a more interesting field.
I even read an article of someone who "upgraded" a very cheap VA synth with cheap transformers on the outs (i think they're the ones mentioned in the pu-simulation article of geofex / musikding.de?) to get a completely new instrument.
i see this topic more from a musicians point of view, so *better* is more like a matter of "taste" and a hunt for more unique sounds out of things that were mostly crap before, at least for me.
With quality good sounding units i'm very lazy to do any mods anyway...
why should i change something i really like already?
I think i would not try to change something on a unit like the Lake People high end gear... as i wouldn't in my EMT limiter...

 

[clintrubber]

(I never met anyone who designed a great piece of gear completely by accident)

:green: :thumb:

Not sure though if that 'I want a radio, but got this thing'-story is true...

 

[Rossi]

I heard the aphex exciter was invented by accident. Don't know if that counts as a great design, though.

As for the B*ringer comp*sers: the VCA chips vary a lot too. Each of mine has a different ones. But that's mostly due to the transition from DBX to THAT. Some inbetween ones are from µPC. But I don't think they make much of a difference. They're all slightly lower grade THAT2150 compatibles. I wouldn't expect an appreciable difference by upgrading to THAT 2180/2181.

 

[pstamler]

[quote author="Rossi"]I heard the aphex exciter was invented by accident. Don't know if that counts as a great design, though.[/quote]

And the designer's story keeps changing. Sometimes he was building a Dynaco Mark III power amp and wired it up wrong, and sometimes it was a Dynaco PAS3 preamp. There's a bit of difference between those two designs.

Peace,
Paul

 

[Rossi]

I think the greatest invention was not the exciter itself but the marketing hype around it. Given that it's basically just an overdriven hi pass mixed with the original signal. But then again, it came out when treble roll off was everywhere and bright sound wihout hiss was expensive.