Audio Asylum Thread Printer
Get a view of an entire thread on one page
|For Sale Ads|
I don't have a scope :/ but I guess if it sounds good and the opamps don't get too warm to the touch, it means they're happy?... that's how I did it with my Revox. I fiddled with local decoupling until it sounded right. Not hollow, not distant, not overly sharp, just good. Took a while but it worked.
A scope is your friend. Every audiophile that wants to tweak things should have one. I hear friends yakking about how much 'more open' it sounds after swapping op amps with super bandwidth devices (not always a good idea), but after sticking a scope on the outputs, find that this improvement in 'airiness' was due to oscillations. The LM4562 is designed for high quality audio, so I dont expect you will have a problem with it, but if you cant see, you will never know. You can get a good used analog scope for nothing these days, and a decent digital scope new for a few hundred. Get a scope.
Sure, a scope would be sweet for everyone who wants to swap an opamp or two, but it's a pretty big expenditure and learning curve. These opamps are not mysterious or so difficult to implement. With proper bypassing (no excuse for skimping on that), decent circuit layout and non-capacitive load, the LM4562 should work if the OPA2134 already did. To ensure stability, a 100 ohm resistor can be placed in series with the outputs. Done.
As your friends have learned, simply dropping a high speed opamp into any circuit can have bumpy results, but it seems that the road to success is already mostly paved in this case.
If the OP can't sleep at night for fear of oscillation, he could borrow a scope or find a competent tech.
berate is 8 and benign is 9
In the context of an opamp in a crossover circuit, what could a "capacitive load" be? Sorry if I ask stupid questions, I just need practical examples to understand the theory better...
The output opamp could see a capacitive load from IC's. Any opamp "inside" the xover should be fine.
Once again, if the 2134's were properly decoupled (local 0.1uF bypass + and - to ground) and stable, odds are good that the 4562 will also be stable. Its output overshoot is only slightly more pronounced into a capacitive load than the 2134, but both are sensitive to rising capacitance so you should damp with the 100 ohm resistor. Of course, there are myriad other differences between the two, but not many that will affect their stability and performance in a crossover circuit. The 4562 GBP is much higher than the 2134, which might make a marginal 2134 application into an unacceptable 4562 implementation.
I have found that the sound each produces is different, with the 4562 providing far more high freq detail, nearly an excess sometimes tending toward a sterile sound. It seems the bass was more prominent from the 2134, but that may have been just because of more relaxed highs. Perhaps my 4562's were oscillating and I never knew it! But I doubt it because I took all the proper precautions and the system performed suitably for a couple years that way, so it must not have done damage to anything.
There is a ton of info regarding opamp comparisons, most of it bewildering and inconclusive because everyone has their own taste and the parts are used in all kinds of applications. The cost of experimenting is certainly not prohibitive unless you're determined to blow something up. I'm not a technician, and I have managed to try a few opamps in my active xovers without destroying anything.
berate is 8 and benign is 9
Call me a lunatic but my previous experience (related it seems by other users) is that LM4562 needs a good 100hours minimum to start sounding as it is supposed to. The first days it sounds cold and sterile and slightly forward, then it mellows down but remains transparent and open. I know op amp break-in is a controversial subject but I heard it... i guess most people trying to quickly compare don't give enough time to the chips to settle down. My opinion of course. :)
You get no criticism from me regarding break-in of ANY modification to a system. New solder joints and all other components require some indefinite period to somehow settle down. I think the length of time required relates to current flow and complexity. At line level, it can take several days or even weeks of regular playing. At speaker level, I think the process occurs in less time, except for speaker drivers which also involves a mechanical element.
The 4562's were in my system for a couple years. They did improve initially, but only to a certain point. I was never completely satisfied with their overall timbre. Lots of people like them, but I eventually moved on to the much more realistic OPA627, then the even better Sparkos Labs discrete.
I hope your experiments go well. Let us know.
berate is 8 and benign is 9
Could you recommend a couple of oscilloscopes to look at?
Audio only, so 50Mhz~100Mhz should be fast enough, dual channel.
(time to replace my 40's vintage HP).
Is their a decent generator that would do what an HP200C could do?
Any scope from a major manufacturer would fit the bill. I picked up a Tektronix 2215A (60MHz, dual trace) a few years back for $125. You can probably get one cheaper today. Look for Tek, HP, B&K, etc. for used scopes. For new digital ones, Siglent makes some low cost ones that would work well.
Post a Message!
This post is made possible by the generous support of people like you and our sponsors: