opa134's x NE5534
J.P. Ambrogi
I do have some Burr-brown OPA 134's laying around and I'm thinking
about
swapping out the NE5534's on the Buss driver section of my AMEK ANGELA
modules and putting the BB's on. The idea here is getting a cleaner
signal since the distortion specs are much better for the 134's. Would
I have to remove the Pin 5/8 compensation capacitors? Would the
current consume be an issue (I don't know how much current the 5534's
do consume!)?
Thanks,
J.P. Ambrogi
BRAZIL
Arny Krueger
"J.P. Ambrogi" <jpam...@ig.com.br> wrote in message
news:3254de3c.02041...@posting.google.com...
Are you sure that OPA 134's have lower distoriton, or that their
distortion is actually audible?
Please read http://www.dself.demon.co.uk/webbop/opamp.htm#select
Please check out http://www.dself.demon.co.uk/webbop/2134.htm Please
notice the sharp rise in distortion above 10 KHz.
Please check out http://www.dself.demon.co.uk/webbop/5532.htm .
Monte P McGuire
J.P. Ambrogi <jpam...@ig.com.br> wrote:
>I do have some Burr-brown OPA 134's laying around and I'm thinking
>about swapping out the NE5534's on the Buss driver section of my AMEK
>ANGELA modules and putting the BB's on. The idea here is getting a
>cleaner signal since the distortion specs are much better for the
>134's.
Well, the problem with the specs is that the 5534 is pretty old. Back
then, its performance was basically at the test equipment's residual.
Measure it with gear from '76 and it looks pretty shabby. Measure it
today with better test gear and you'll find that a 5534 is far better
than a 134 when driving low impedance loads at high frequencies. The
134 might be more euphonic, but it certainly isn't cleaner.
What's better than a 5534 is an LT 1468. It actually has similar
noise properties too (the 134 has a lot more voltage noise) if this is
important. It also has better DC specs, less distortion, better
bandwidth etc. It's also unity gain compensated, so you can ditch the
pin 5-8 compensation caps.
Or... you could leave your Angela stock. There really isn't much
wrong with a 5534 used in a competently designed circuit.
Have fun,
Monte McGuire
mcg...@theworld.com
Scott Dorsey
>
>Please check out http://www.dself.demon.co.uk/webbop/2134.htm Please
>notice the sharp rise in distortion above 10 KHz.
Right, but who cares about distortion above 10 KHz? The second harmonic
of 10 KHz is audible, but beyond that it's all in the gods.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."
Arny Krueger
"Scott Dorsey" <klu...@panix.com> wrote in message
news:a9f5o7$1n4$1...@panix2.panix.com...
> Arny Krueger <ar...@hotpop.com> wrote:
> >
> >Please check out http://www.dself.demon.co.uk/webbop/2134.htm
Please
> >notice the sharp rise in distortion above 10 KHz.
> Right, but who cares about distortion above 10 KHz?
Everybody who believes that intermodulation distortion can be a
significant issue.
>The second harmonic of 10 KHz is audible, but beyond that it's all
in the gods.
Yes, but if we are talking about amplitude modulation distortion (
and in this case that's exactly what we are talking about) IM and THD
come together, inextricably connected. I've got some graphics that
show this in a new web page I'm working on. You can see it (under
construction) at http://www.cdabx.com/technical/nonlinear/ . Note:
the pictures are posted, but I'm still working on the musical samples
so they are all broken links at this time.
MS
sounding IC op-amp I've heard in the various pre circuits I've popped
it in (some garbage, some great). You might try some A/B recording
tests with your 134's, 5534's and some 227's, I think you'll be
pleasantly surprised when you get to the OPA227. BTW - I found it to
sound better than it's un-comp'd OPA228 sibling.
Peter Larsen
[someone asked]
> > Would I have to remove the Pin 5/8 compensation capacitors?
> > Would the current consume be an issue (I don't know how much
> > current the 5534's do consume!)?
One option would be to fit sockets and leave pin 8 out, literally that
is .... or simply
physically remove it, and leave the pcb unmodidied. It is the smallest
change.
> Are you sure that OPA 134's have lower distoriton, or that their
> distortion is actually audible?
I put a handful in a cross-over to test them. Sounds different than with
176's, more "up front". Turn-on and turn-off transients are however
worse.
> Please read http://www.dself.demon.co.uk/webbop/opamp.htm#select
I have read it with a sense of deep wonder. Just how many seconds of
audio does one need for a comparison with the ABX thingie?
> Please check out http://www.dself.demon.co.uk/webbop/2134.htm Please
> notice the sharp rise in distortion above 10 KHz.
>
> Please check out http://www.dself.demon.co.uk/webbop/5532.htm .
--
*************************************************************
* This posting handcrafted by Peter Larsen, MCSE *
* My site is at: http://www.muyiovatki.dk *
* I subscribe to http://www.spamcop.net *
*************************************************************
Arny Krueger
"Peter Larsen" <pla...@mail.tele.dk> wrote in message
news:3CBD0E0C...@mail.tele.dk...
> Arny Krueger wrote:
> I put a handful in a cross-over to test them. Sounds different than
with
> 176's, more "up front". Turn-on and turn-off transients are however
> worse.
> > Please read http://www.dself.demon.co.uk/webbop/opamp.htm#select
> I have read it with a sense of deep wonder. Just how many seconds
of
> audio does one need for a comparison with the ABX thingie?
I've never experimented to see what the lower bound is, but it has
to be on the order of a few milliseconds.
The practical lower bound for file size is probably not set by the
software, but rather comes from the "punch-in" time of the sound
card. What I'm referring to here is the fact that the PCABX
Comparator works by stopping one sound file from playing, changing
(or not changing) the name of the sound file being played, and then
starting the next (or the same) sound file from the same place given
in milliseconds, that the sound was playing when it was stopped. I've
found that with many sound cards it takes from a fraction of a
millisecond to a few dozen milliseconds for them to produce a
non-zero signal, after the playing of a file commences. I have some
samples at www.pcabx.com that are just over half a second long that
seem to play just fine.
The upper bound for files that the PCABX Comparator can play was set
by the design of the sound player in Windows. Windows accepts a
starting location in milliseconds, but the starting location is
expressed as a 16 bit integer. So, the longest file that can be
effectively played with the PCABX Comparator is about 65 seconds
long.
If this was a music listening test, then 65 seconds might be too
short. But it isn't a music appreciation class, it's a sound quality
listening test. It is presumed that a musical passage that is hard to
reproduce correctly has been selected by other means, and that this
musical selection is all that the listener needs to listen to for one
test. We all have audio editing software and sound file libraries,
don't we?
If there are more than a few seconds between two sounds, the human
ability to hear small differences between them goes to hell in a
handbasket. Therefore, the sounds can't be more than a few seconds
long if maximum sensitivity is desired.
J.P. Ambrogi
Thanks everybody for the fine posts!
I think I've learned something with you guys on this group: what makes or
breaks a circuit design is not only the quality of components but, mainly,
the interaction between the components, right?
I'll accept the advice to leave my board stock! I guess I've been a little bit
presumptuous thinking that I could do it better than AMEK designers, especially
considering my modest technical knowledge!
J.P. Ambrogi
BRAZIL - land of Gisele Bundchen
Ulysses
jpam...@ig.com.br (J.P. Ambrogi) wrote:
> Thanks everybody for the fine posts!
> I think I've learned something with you guys on this group: what makes or
> breaks a circuit design is not only the quality of components but, mainly,
> the interaction between the components, right?
> I'll accept the advice to leave my board stock! I guess I've been a little bit
> presumptuous thinking that I could do it better than AMEK designers,
especially
> considering my modest technical knowledge!
Yes. I swapped the 5534s for 1468s in the monitor section of my Amek BCII
and also jumpered some coupling caps. Now my level pot is noisy and
there's some RF oscillations. I haven't found the time to put it back the
way it was or hunt down the dilemma yet. I did this to one of the two
monitor paths ans I can't hear a difference between the two so it was
probably not worthwhile.
--
Justin Ulysses Morse
Roll Music Studios
Minneapolis, MN
www.rollmusic.com War is terror.
Scott Dorsey
>
>
>Yes. I swapped the 5534s for 1468s in the monitor section of my Amek BCII
>and also jumpered some coupling caps. Now my level pot is noisy and
>there's some RF oscillations. I haven't found the time to put it back the
>way it was or hunt down the dilemma yet. I did this to one of the two
>monitor paths ans I can't hear a difference between the two so it was
>probably not worthwhile.
Put the cap immediately before the level pot back into place.
Apply 22 pf compensation caps on each of the chips.
Geoff Wood
geoff
"Ulysses" <uly...@rollmusic.com> wrote in message
news:ulysses-1904...@192.168.0.2...
Monte P McGuire
...across the feedback network, i.e. from pin 6 to pin 2, _not_ across
pin 5 and 8. Install this cap as close to pin 2 and 6 as possible,
perhaps on the solder side of the PC board, under the chip.
For the folks at home, the 1468 has no provision for a true
compensation cap, which would be attached across pins 5 and 8, so what
Scott really means is to provide a HF feedback path so that the
feedback loop closes without odd phase shifts from a possibly large
and inductive feedback network.
This is the price of fast op amps - they can become unstable in a
layout that was designed for a different chip. Some circuits however
are hand tuned to the quirks of an op amp, so do not expect this or
any other "do this, do that" remedy to work. You need to study the
circuit and look at the layout to see where the problem is.
Also make sure that the power supply bypassing is good at HF. A small
(.01uF) cap across pins 4 and 7 may help, even if there is no good
local ground.
Regards,
Monte McGuire
mcg...@theworld.com
Peter Larsen
> Yes. I swapped the 5534s for 1468s in the monitor section
> of my Amek BCII
There are caveats, one of those is that one should aim for understanding
the circuit modified. Problems in sensible IC upgradings are few and far
between, but they can and do happen. The OPAx134 range was suggested to
me for use in mixing desks EQ section by someone over in r.a.t rather
than the OPAx604 range because of stability concerns, so it is probably
a benign choice. The AD 275/176's are also "user friendly", note that
the 176 is not totally identical because it is a later version. IMO it
could be slightly cleaner based on long time listening assesments of
different devices - long term = a couple of weeks of use with all kinds
of programme.
> and also jumpered some coupling caps.
You MUST understand the circuit prior to so doing, caps that protect a
pot or a fader may NOT - or "hardly ever" be bypassed. The BB OPAx604
happens to be almost the only known usable replacement of a certain
stoneage Mitsubishi opamp occasionally used by Sony to be able to skip a
coupling cap in their input design on one of their DAT's. Generally it
may be possible to replace some coupling caps with a polypropylene
component of smaller component value, but beware of creeping loss of
bandwidth. Same component value will usually NOT be usable because it
will be physically larger. Wise PCB design use back to back
electrolytics with the appropiate leg of the supply voltage to their
center, but this simple cost-cutting and quality maximizing meausure is
hardly ever used in real life.
> Now my level pot is noisy
This because a "dc" current flows through it. You MUST address this
because the potentiometer will become permanently noisy if you do not.
What happens is micro-arcing over when it is moved which results in
physical damage to the conductive surface.
> and there's some RF oscillations.
Perhaps you also bypassed a decoupling cap or two ... ? ... into the
mixer you go, and do the snippety snip - mind you: seems like superceded
by another poster ....
> I haven't found the time to put it back the
> way it was or hunt down the dilemma yet.
Removing the cap bypassing does not take long time. This is urgent.
> I did this to one of the two monitor paths ans I can't hear
> a difference between the two so it was probably not worthwhile.
If you want to listen for IC differences then is somewhat easier to do
it earlier in the signal chain, but that too is not - for other reasons
- always easy. I still do not understand the results on Self's site, I
can not make them correlate well with aural subjective perception.
> Justin Ulysses Morse
Ulysses
Dorsey) wrote:
I love you, Scott.
P Stamler
>>
>> Apply 22 pf compensation caps on each of the chips.
>> --scott
Also put back the cap between the level pot's slider and the *next* stage; the
input current from this stage is a likely cause of a noisy pot.
Peace,
Paul