[[email protected]]

Hi,

As I am looking at older consoles, I understand that some have the
dreaded 709 chips. I'm curious, is there a "modern" drop-in substitute
with better specs that won't need an extensive PSU upgrade?

Thanks,
Chris

[Scott Dorsey]

wrote:

As I am looking at older consoles, I understand that some have the
dreaded 709 chips. I'm curious, is there a "modern" drop-in substitute
with better specs that won't need an extensive PSU upgrade?

You could probably use just about anything, even an OPA602, with a little cut
and paste work. Problem is that a lot of those designs were on the bare
edge of stability with the 709 and if you put a faster chip in there, the
parasitics will kill you unless you spend some time fixing the ground layouts
and adding decoupling. (These things will ALSO improve the sound if you keep
the 709 in place, though).

On the other hand, for unity gain buffers the 709 is probably just fine.

What consoles are these?
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[[email protected]]

Scott Dorsey wrote:

What consoles are these?

Hi Scott,

Auditronics and Quantum.

[Scott Dorsey]

wrote:
Scott Dorsey wrote:

What consoles are these?

Auditronics and Quantum.

Never worked on a Quantum. On the Auditronics you have lots of other more
important things to worry about than the 709s.
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[[email protected]]

Scott Dorsey wrote:

Never worked on a Quantum. On the Auditronics you have lots of other more
important things to worry about than the 709s.

Auditronics, what are these other concerns?

I'm sure you must have heard what the Quantum or the Auditronics
consoles sound like, are they decent-sounding?

[Paul Stamler]

"Scott Dorsey" wrote in message
...

On the other hand, for unity gain buffers the 709 is probably just fine.

No it ain't. For unity gain buffers the 709 need maximum compensation, which
makes the GBW and SR pitifully poor. (The slew rate is, in fact, about
0.33V/usec). According to the manual, at unity gain the full-power bandwidth
is about 3kHz. Try putting a cymbal crash through *that*. No, on second
thought, don't.

The 709 draws 2.5mA per amplifier, so that's what you'd have to work with,
but Scott's right about the stability issues.

Peace,
Paul

[Paul Stamler]

wrote in message
ps.com...

Scott Dorsey wrote:

Never worked on a Quantum. On the Auditronics you have lots of other
more
important things to worry about than the 709s.

Auditronics, what are these other concerns?

I'm sure you must have heard what the Quantum or the Auditronics
consoles sound like, are they decent-sounding?

There are many Quantum consoles from several generations; some had 709s and
301s in 'em, some had LM318s in the inputs (yes, really) and sounded quite
decent once you replaced the 4136's in the EQ circuits, and some had 5534s
all over, and they sounded very good once you replaced the 4136s (or if you
kept the EQ switched out). I've worked with the latter two series, as you
might guess; they also had Jensen transformers in them.

Peace,
Paul

[[email protected]]

Paul Stamler wrote:

decent once you replaced the 4136's in the EQ circuits, and some had 5534s
all over, and they sounded very good once you replaced the 4136s (or if you

Hi Paul,

I understand that the 4136 are gyrators and use an uncommon pinout not
shared by most other quad opamps. So what do you swap the 4136s with
then?

Best,
Chris

[Paul Stamler]

wrote in message
oups.com...
Paul Stamler wrote:

decent once you replaced the 4136's in the EQ circuits, and some had
5534s
all over, and they sounded very good once you replaced the 4136s (or if
you

Hi Paul,

I understand that the 4136 are gyrators and use an uncommon pinout not
shared by most other quad opamps. So what do you swap the 4136s with
then?

I used TL075s, now discontinued. They're part of the same family as TL072s,
but share the same weird pinout as the 4136.

Peace,
Paul

[Scott Dorsey]

wrote:
Scott Dorsey wrote:

Never worked on a Quantum. On the Auditronics you have lots of other more
important things to worry about than the 709s.

Auditronics, what are these other concerns?

Just lots of layout problems. The EQ is an old-style inductive network,
but the preamps are pretty noisy by modern standards and there is a lot
of crosstalk as well. The summing networks are all passive but aren't
laid out so well... first thing I'd do if I had one was update the summing
amps.

I'm sure you must have heard what the Quantum or the Auditronics
consoles sound like, are they decent-sounding?

Not in years... these were broadcast gear for the most part.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Scott Dorsey]

wrote:

I understand that the 4136 are gyrators and use an uncommon pinout not
shared by most other quad opamps. So what do you swap the 4136s with
then?

The 4136 has a weird layout, so you will have to make adaptors. They can
be used as gyrators, but then so can any other op-amp.
--scott


--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[[email protected]]

wrote:
Hi,

As I am looking at older consoles, I understand that some have the
dreaded 709 chips. I'm curious, is there a "modern" drop-in substitute
with better specs that won't need an extensive PSU upgrade?

Thanks,
Chris

Get a wirewrap socket and bend the leads into the standard 8 pin dip
layout. For 4136, Brown Dog has made adaptors by my request to convert
the obsolite 4136 pinout to 2 surface mount duals. Phase compensation
or feedback correction may be needed and add some .1 uf mono ceramic
psu bypass caps and some speed demon parts will work great.

Remember, they don't work if you let the smoke out.

Jim Williams
Audio Upgrades

[Paul Stamler]

"Chel van Gennip" wrote in message
...
On Tue, 17 Oct 2006 06:23:56 +0200, Paul Stamler wrote:

"Scott Dorsey" wrote in message
...

On the other hand, for unity gain buffers the 709 is probably just
fine.

No it ain't. For unity gain buffers the 709 need maximum compensation,

Not exactly needed, you should choose compensation according to your
specific wishes. The choice may depend on noise, output swing, sensitivity
for circuit board design (stability) etc. requirements. If you want full
swing audio, C1=500pF, R1=1K5, C2=20pf often is a good choice. With the
709 you sometimes need to think about the design a bit.

which makes the GBW and SR pitifully poor. (The slew rate is, in fact,
about 0.33V/usec). According to the manual, at unity gain the full-power
bandwidth is about 3kHz. Try putting a cymbal crash through *that*. No,
on second thought, don't.

Depending on your needs slew rates up to 20V/us and 60dB amplification at
1MHz bandwidth are possible.

Please note the comment to which I was replying: "For unity-gain buffers the
709 is probably just fine." That is, in fact, the worst case for a 709. At
higher gains they can, with very careful design, provide okay performance,
provided they're not asked to drive low-impedance loads. But unity-gain, on
a 709, is a recipe for audio disaster.

Peace,
Paul

[[email protected]]

wrote:
For 4136, Brown Dog has made adaptors by my request to convert
the obsolite 4136 pinout to 2 surface mount duals.

Ok, I see them available he

http://cimarrontechnology.com/index....PROD&ProdID=35

but what chip can I use that is a SO8 sub? I was thinking about trying
the 8-pin DIP OP275. Is there an equivalent SO8 for the OP275?

[[email protected]]

wrote:
wrote:
For 4136, Brown Dog has made adaptors by my request to convert
the obsolite 4136 pinout to 2 surface mount duals.

Ok, I see them available he

http://cimarrontechnology.com/index....PROD&ProdID=35

but what chip can I use that is a SO8 sub? I was thinking about trying
the 8-pin DIP OP275. Is there an equivalent SO8 for the OP275?

Forget that obsolite opamp. If you want to use audio grade opamps I
would use the new National LM4562. Otherwise I'd shove some National
LM6172's or THS duals in there, I have a need for speed...

Jim Williams
Audio Upgrades

[GregS]

In article . com, wrote:

wrote:
wrote:
For 4136, Brown Dog has made adaptors by my request to convert
the obsolite 4136 pinout to 2 surface mount duals.

Ok, I see them available he

http://cimarrontechnology.com/index....PROD&ProdID=35

but what chip can I use that is a SO8 sub? I was thinking about trying
the 8-pin DIP OP275. Is there an equivalent SO8 for the OP275?

Forget that obsolite opamp. If you want to use audio grade opamps I
would use the new National LM4562. Otherwise I'd shove some National
LM6172's or THS duals in there, I have a need for speed...

I would think the application or previous circuit design would dictate type, speed,
drive, CMR, gain, and power requirments, for optimization and relibility without modifications.

greg

[[email protected]]

GregS wrote:
In article . com, wrote:
but what chip can I use that is a SO8 sub? I was thinking about trying
the 8-pin DIP OP275. Is there an equivalent SO8 for the OP275?

Forget that obsolite opamp. If you want to use audio grade opamps I
would use the new National LM4562. Otherwise I'd shove some National
LM6172's or THS duals in there, I have a need for speed...

I would think the application or previous circuit design would dictate type, speed,
drive, CMR, gain, and power requirments, for optimization and relibility without modifications.

Yes, I want to improve the board without having to potentially
"redesign" it too much (i.e., just plug and pray =). Surely the OP275
would be significantly "better" than the original 4136 type? The slew
rate of the OP275 is comparable to the LM4562...

[Paul Stamler]

wrote in message
ups.com...

GregS wrote:
In article . com,
wrote:
but what chip can I use that is a SO8 sub? I was thinking about
trying
the 8-pin DIP OP275. Is there an equivalent SO8 for the OP275?

Forget that obsolite opamp. If you want to use audio grade opamps I
would use the new National LM4562. Otherwise I'd shove some National
LM6172's or THS duals in there, I have a need for speed...

I would think the application or previous circuit design would dictate
type, speed,
drive, CMR, gain, and power requirments, for optimization and relibility
without modifications.

Yes, I want to improve the board without having to potentially
"redesign" it too much (i.e., just plug and pray =). Surely the OP275
would be significantly "better" than the original 4136 type? The slew
rate of the OP275 is comparable to the LM4562...

You need to keep an eye on current consumption. An RC4136 quad package
typically uses 5mA quiescent current at +/-15V. A TL075, which is what I
used to replace the 4136s, is about the same. A pair of OP275s would be 8mA,
while a pair of LM4562s would be 20mA. Sixteen of the latter (one for each
channel EQ strip) would pull 320mA, nearly 1/3 of an amp, dissipating 9.6W.
That ain't hay. (The 4136s dissipated 2.4W total.) Even the OP275s would
pull 128mA and dissipate 3.84W, or about 60% more than the 4136s. And the
Quantum boards normally run on the warm side already.

I didn't have to add any additional decoupling to the Quantum when I
switched from 4136s to TL075s; everything stayed stable. Yes, the OP275 is a
better chip, and it doesn't pull *that* much extra current, so it might be
worth a try, but if it were me, I'd stick with the TL07x series. The adapter
that turns two 8-pin surface-mounted packages to a 4136 pinout could be used
with a pair of TL072s and you'd be home free. You could buy an adapter,
solder in two OP275s and see if they work; if so, fine; if not, then TL072s
should do the job. I have to say that the board sounded very nice with the
TL075 chips in it; I suspect the circuit was designed around the limitations
of the 4136, some of which are shared by the TL07x series..

Peace,
Paul

[[email protected]]

Paul Stamler wrote:
You need to keep an eye on current consumption.

Yeah, that was what I was thinking. However, I never realized that the
TL072 would have the same draw as the 4136. So it makes sense to give
that a try first (sorry, I'm still a newbie when it comes to ICs).

I didn't have to add any additional decoupling to the Quantum when I
switched from 4136s to TL075s; everything stayed stable.

Yes, this is the kind of thing I am trying to avoid for now. Maybe when
I get more time (that and maybe an upgrade to an Acopian PS =), I can
try other chips.

I have to say that the board sounded very nice with the
TL075 chips in it; I suspect the circuit was designed around the limitations
of the 4136, some of which are shared by the TL07x series..

OK, I will give the TL072s a shot.

Thank you very much, Paul! You've been very informative!

[GregS]

In article om, wrote:
Paul Stamler wrote:
You need to keep an eye on current consumption.

Yeah, that was what I was thinking. However, I never realized that the
TL072 would have the same draw as the 4136. So it makes sense to give
that a try first (sorry, I'm still a newbie when it comes to ICs).

This chip used to be used to replace the 4136 because they did make
a compatable package TLO75. It had resonable frequency response
and had resonable low noise with high Z circuitry. With low Z inputs
it would have had higher noise. If the board does not have adequate
supply bypassing, the newer faster chips can oscillate or create some
kind of high frequency hash.

greg

[[email protected]]

wrote:
GregS wrote:
In article . com, wrote:
but what chip can I use that is a SO8 sub? I was thinking about trying
the 8-pin DIP OP275. Is there an equivalent SO8 for the OP275?

Forget that obsolite opamp. If you want to use audio grade opamps I
would use the new National LM4562. Otherwise I'd shove some National
LM6172's or THS duals in there, I have a need for speed...

I would think the application or previous circuit design would dictate type, speed,
drive, CMR, gain, and power requirments, for optimization and relibility without modifications.

Yes, I want to improve the board without having to potentially
"redesign" it too much (i.e., just plug and pray =). Surely the OP275
would be significantly "better" than the original 4136 type? The slew
rate of the OP275 is comparable to the LM4562...

It's not just slew rate, it's also noise, the 4136 is 10 nv /hz sq, the
4562 is 2.3 nv/ hz sq. OP-275 is either 6 or 8 nv/hz sq depending on
which data sheet you believe. The most important spec is open loop
gain. All the audio grade opamps are limited to 60 db OLG at 10k hz, to
which distortion prevails in the mids from this. The 4562 and some
other wideband speed demon opamps are 80 to 90 db OLG at 10k hz, 20 db
cleaner. This is why the LM4562 claims .00003% THD. The National LM6172
with it's high open loop gain and 3000 v/us slew rate only draws 2.3 ma
current vs 1.8 ma for a TL07X, the linear tech LT1358 is similar to the
6172's draw. Now if you want some really quiet audio opamps that suck
power, check out the new AD 8599 with it's 1 nv/hz sq noise spec.

If you put TLO7X opamps in there, you will re-create the situation that
drove up the prices for discrete transistor consoles, they sound like
crap. Go buy a $135 power one HDD-15-5A power supply and feed it. BTW,
if you don't have a scope with some troubleshooting and opamp stability
skills, leave it to those who do.

Jim Williams
Audio Upgrades

[Paul Stamler]

wrote in message
ps.com...

It's not just slew rate, it's also noise, the 4136 is 10 nv /hz sq, the
4562 is 2.3 nv/ hz sq. OP-275 is either 6 or 8 nv/hz sq depending on
which data sheet you believe. The most important spec is open loop
gain. All the audio grade opamps are limited to 60 db OLG at 10k hz, to
which distortion prevails in the mids from this. The 4562 and some
other wideband speed demon opamps are 80 to 90 db OLG at 10k hz, 20 db
cleaner. This is why the LM4562 claims .00003% THD. The National LM6172
with it's high open loop gain and 3000 v/us slew rate only draws 2.3 ma
current vs 1.8 ma for a TL07X, the linear tech LT1358 is similar to the
6172's draw. Now if you want some really quiet audio opamps that suck
power, check out the new AD 8599 with it's 1 nv/hz sq noise spec.

None of the noise specs are relevant in this case; the 4136 was used in the
Quantum board in the EQ circuit, which operated at high enough levels that
the chips' noise level's not an issue.

If you put TLO7X opamps in there, you will re-create the situation that
drove up the prices for discrete transistor consoles, they sound like
crap. Go buy a $135 power one HDD-15-5A power supply and feed it. BTW,
if you don't have a scope with some troubleshooting and opamp stability
skills, leave it to those who do.

Thing is, the TL07x series don't *always* sound like crap. Their big
weaknesses are a weak output stage, 3MHz GBW, and susceptibility to
common-mode distortion from high source impedances in the voltage-follower
configuration. (Most FET-input opamps share the last-named problem.) The GBW
primarily limits their usefulness in stages with gain, but as I recall
they're used at vey low gain in the Quantum's EQ circuit, and can operate
with good linearity. I *think* they're used as inverters (it's been a
quarter-century since I looked at the circuit diagram), so common-mode
distortion's not an issue. That leaves the output stage. The 4136 was
notorious for having a weak output stage, and I suspect Quantum designed
around that when they built this EQ, because in practice the TL075 not only
made a major improvement over the 4136, it also rendered the EQ stage pretty
close to transparent when you switched in and out of bypass (assuming, of
course, that the knobs were set to flat!) Far from sounding like crap, the
Quantum boards actually sounded sweet and clean, with or without the EQ
switched in.

I'd be the last to say that TL07x chips are the be-all and end-all of audio.
In a lot of applications they really do suck, and I've replaced them with
better chips often enough. But in some places they're pretty harmless, and a
safer upgrade than some of the more modern chips with poorer stability. In
my experience, the Quantum boards are one of the places where they work
well.

Peace,
Paul

[[email protected]]

Paul Stamler wrote:
wrote in message
ps.com...

It's not just slew rate, it's also noise, the 4136 is 10 nv /hz sq, the
4562 is 2.3 nv/ hz sq. OP-275 is either 6 or 8 nv/hz sq depending on
which data sheet you believe. The most important spec is open loop
gain. All the audio grade opamps are limited to 60 db OLG at 10k hz, to
which distortion prevails in the mids from this. The 4562 and some
other wideband speed demon opamps are 80 to 90 db OLG at 10k hz, 20 db
cleaner. This is why the LM4562 claims .00003% THD. The National LM6172
with it's high open loop gain and 3000 v/us slew rate only draws 2.3 ma
current vs 1.8 ma for a TL07X, the linear tech LT1358 is similar to the
6172's draw. Now if you want some really quiet audio opamps that suck
power, check out the new AD 8599 with it's 1 nv/hz sq noise spec.

None of the noise specs are relevant in this case; the 4136 was used in the
Quantum board in the EQ circuit, which operated at high enough levels that
the chips' noise level's not an issue.

If you put TLO7X opamps in there, you will re-create the situation that
drove up the prices for discrete transistor consoles, they sound like
crap. Go buy a $135 power one HDD-15-5A power supply and feed it. BTW,
if you don't have a scope with some troubleshooting and opamp stability
skills, leave it to those who do.

Thing is, the TL07x series don't *always* sound like crap. Their big
weaknesses are a weak output stage, 3MHz GBW, and susceptibility to
common-mode distortion from high source impedances in the voltage-follower
configuration. (Most FET-input opamps share the last-named problem.) The GBW
primarily limits their usefulness in stages with gain, but as I recall
they're used at vey low gain in the Quantum's EQ circuit, and can operate
with good linearity. I *think* they're used as inverters (it's been a
quarter-century since I looked at the circuit diagram), so common-mode
distortion's not an issue. That leaves the output stage. The 4136 was
notorious for having a weak output stage, and I suspect Quantum designed
around that when they built this EQ, because in practice the TL075 not only
made a major improvement over the 4136, it also rendered the EQ stage pretty
close to transparent when you switched in and out of bypass (assuming, of
course, that the knobs were set to flat!) Far from sounding like crap, the
Quantum boards actually sounded sweet and clean, with or without the EQ
switched in.

I'd be the last to say that TL07x chips are the be-all and end-all of audio.
In a lot of applications they really do suck, and I've replaced them with
better chips often enough. But in some places they're pretty harmless, and a
safer upgrade than some of the more modern chips with poorer stability. In
my experience, the Quantum boards are one of the places where they work
well.

Peace,
Paul

Sorry, Paul, they do suck. Any FFT will prove this. I don't enjoy a
"picket fence" worth of upper level harmonics with my music. 40db loop
gain at 10k hz puts these into LM741 distortion really fast. Pump up
+15 db hi EQ boost and you only have 25 db of loop gain for correction,
not too good. 10 nv/hz sq parts do show up as hiss in top EQ circuits
when boosted heavily. I still have a bunch of 10 nv/Hz sq National
LM6172's in my Delta console. To be fair, some of the noise is due to
poor impedance ratios and large resistors contributing to thermal
noise.
I use TL07X parts, in my metering! They are good enough to look at. I
also use a TL071 in my custom fuzz tone because of all those rich
harmonics. I also use one to buffer my Audio Precision when measuring
guitar pickups frequency response, but never for a noise or distortion
test.

Remember, everything sound great until you hear something better.
I have ditched AD811's, AD 744's and other "audiophile" opamps praised
in Audio Express because better choices are now available.

Jim Williams
Audio Upgrades

[[email protected]]

I have a Quantum QM-128 with QM-8EX expansion (so...20x8 for those who
are counting)

Opened it up and yup there's TL075s in the EQ path. EQ to me sounds
decent - fairly smooth. Even at its highest settings. The trannys
aren't clearly labeled. I'm guessing they're cinemag or reichenbach


wrote:
Hi,

As I am looking at older consoles, I understand that some have the
dreaded 709 chips. I'm curious, is there a "modern" drop-in substitute
with better specs that won't need an extensive PSU upgrade?

Thanks,
Chris