Jellybean op amps

[Phil Hobbs]

We've been going back and forth on the LM324/358 etc. recently. One of
my lowish-noise, low cost faves has long been the bipolar MC33078/9,
whose decent combination of noise, bandwidth, cost, DC performance, and
output swing is fairly stellar at about 20 cents in reels, just a bit
more than a knockoff TL084. (*)
It isn't a true single-supply op amp like the 324, but on the other hand
it actually works at AC.

TI recently introduced the OPA1678/9, which are some sort of weird CMOS
riff off the old Motorola chips, and are sort of interesting. They're
twice the price, but a 40-cent quad won't break the bank for most of us.
The specs are quite similar, except for the improved input bias
current and lower distortion. (The 33079's 800-nA maximum bias current
and 500 fA/sqrt(Hz) current noise are pretty annoying in some applications.)

The OPA1678/9's bias and offset currents are ~10 pA at low temperature,
but go nuts above 25C, getting up into the nanoamps at 85C. That's not
horrible for a JFET, but since it's a CMOS part, I gather that something
funky is happening with the input protection circuitry.

Anyway, one for the tool kit. Anybody used it?

Cheers

Phil Hobbs



(*) Avnet claims to be selling the ST Micro TL084 for 5.6 cents in
reels--cheaper than an LM324A!

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com

 

[John Larkin]

On Fri, 3 May 2019 05:26:57 -0400, Phil Hobbs
<pcdhSpamMeSenseless@electrooptical.net> wrote:

We've been going back and forth on the LM324/358 etc. recently. One of
my lowish-noise, low cost faves has long been the bipolar MC33078/9,
whose decent combination of noise, bandwidth, cost, DC performance, and
output swing is fairly stellar at about 20 cents in reels, just a bit
more than a knockoff TL084. (*)
It isn't a true single-supply op amp like the 324, but on the other hand
it actually works at AC.

TI recently introduced the OPA1678/9, which are some sort of weird CMOS
riff off the old Motorola chips, and are sort of interesting. They're
twice the price, but a 40-cent quad won't break the bank for most of us.
The specs are quite similar, except for the improved input bias
current and lower distortion. (The 33079's 800-nA maximum bias current
and 500 fA/sqrt(Hz) current noise are pretty annoying in some applications.)

The OPA1678/9's bias and offset currents are ~10 pA at low temperature,
but go nuts above 25C, getting up into the nanoamps at 85C. That's not
horrible for a JFET, but since it's a CMOS part, I gather that something
funky is happening with the input protection circuitry.

Anyway, one for the tool kit. Anybody used it?

We use some OPA171s which look very similar, same bias current effect
at temperature. Nice little EMI hard part. Well behaved.

Cheers

Phil Hobbs



(*) Avnet claims to be selling the ST Micro TL084 for 5.6 cents in
reels--cheaper than an LM324A!

1.4 cents per opamp. That's insane. These used to cost over $100 each,
when $100 was worth a lot.

https://www.dropbox.com/s/9idlgfug24rqxyb/Philbricks.jpg?dl=0


--

John Larkin Highland Technology, Inc
picosecond timing precision measurement

jlarkin att highlandtechnology dott com
http://www.highlandtechnology.com

 

[George Herold]

On Friday, May 3, 2019 at 5:27:08 AM UTC-4, Phil Hobbs wrote:

We've been going back and forth on the LM324/358 etc. recently. One of
my lowish-noise, low cost faves has long been the bipolar MC33078/9,
whose decent combination of noise, bandwidth, cost, DC performance, and
output swing is fairly stellar at about 20 cents in reels, just a bit
more than a knockoff TL084. (*)
It isn't a true single-supply op amp like the 324, but on the other hand
it actually works at AC.

TI recently introduced the OPA1678/9, which are some sort of weird CMOS
riff off the old Motorola chips, and are sort of interesting. They're
twice the price, but a 40-cent quad won't break the bank for most of us.
The specs are quite similar, except for the improved input bias
current and lower distortion. (The 33079's 800-nA maximum bias current
and 500 fA/sqrt(Hz) current noise are pretty annoying in some applications.)

The OPA1678/9's bias and offset currents are ~10 pA at low temperature,
but go nuts above 25C, getting up into the nanoamps at 85C. That's not
horrible for a JFET, but since it's a CMOS part, I gather that something
funky is happening with the input protection circuitry.

Anyway, one for the tool kit. Anybody used it?

Cheers

Phil Hobbs



(*) Avnet claims to be selling the ST Micro TL084 for 5.6 cents in
reels--cheaper than an LM324A!

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com

The closest I use to a jelly bean is the LT1013, my goto
single supply opamp.

George H.

 

[Winfield Hill]

George Herold wrote...

The closest I use to a jelly bean is the
LT1013, my goto single supply opamp.

I'm also a fan of LT1013, only 150uV offset.
But it had the usual high LTC price, until
TI offered a 2nd source. But now TI's price
is as high as the LTC / AD's price, maybe
even higher, $3.43 qty 100.


--
Thanks,
- Win

 

[bitrex]

On 5/4/19 1:48 AM, bitrex wrote:

On 5/3/19 5:26 AM, Phil Hobbs wrote:
We've been going back and forth on the LM324/358 etc. recently.  One
of my lowish-noise, low cost faves has long been the bipolar
MC33078/9, whose decent combination of noise, bandwidth, cost, DC
performance, and output swing is fairly stellar at about 20 cents in
reels, just a bit more than a knockoff TL084. (*)
It isn't a true single-supply op amp like the 324, but on the other
hand it actually works at AC.

TI recently introduced the OPA1678/9, which are some sort of weird
CMOS riff off the old Motorola chips, and are sort of interesting.
They're twice the price, but a 40-cent quad won't break the bank for
most of us.   The specs are quite similar, except for the improved
input bias current and lower distortion.  (The 33079's 800-nA maximum
bias current and 500 fA/sqrt(Hz) current noise are pretty annoying in
some applications.)

The OPA1678/9's bias and offset currents are ~10 pA at low
temperature, but go nuts above 25C, getting up into the nanoamps at
85C.  That's not horrible for a JFET, but since it's a CMOS part, I
gather that something funky is happening with the input protection
circuitry.

Anyway, one for the tool kit.  Anybody used it?

Cheers

Phil Hobbs



(*) Avnet claims to be selling the ST Micro TL084 for 5.6 cents in
reels--cheaper than an LM324A!


The step-up bipolar jellybean from the LM324 I like is the STMicro
TS462. Lower max supply voltage, GBW (12 vs 16), and slew rate (4 vs 7)
than the MC33078. Worse max offset voltage than the MC33078 5mV vs. 2.

Lower input bias current, lower noise, operates down to +/- 1.25 volts,
and RRIO. Approximately the same price in quantity

https://www.mouser.com/datasheet/2/389/ts461-957349.pdf

correction they're RRO not RRIO

 

[bitrex]

On 5/3/19 5:26 AM, Phil Hobbs wrote:

We've been going back and forth on the LM324/358 etc. recently.  One of
my lowish-noise, low cost faves has long been the bipolar MC33078/9,
whose decent combination of noise, bandwidth, cost, DC performance, and
output swing is fairly stellar at about 20 cents in reels, just a bit
more than a knockoff TL084. (*)
It isn't a true single-supply op amp like the 324, but on the other hand
it actually works at AC.

TI recently introduced the OPA1678/9, which are some sort of weird CMOS
riff off the old Motorola chips, and are sort of interesting.  They're
twice the price, but a 40-cent quad won't break the bank for most of us.
 The specs are quite similar, except for the improved input bias
current and lower distortion.  (The 33079's 800-nA maximum bias current
and 500 fA/sqrt(Hz) current noise are pretty annoying in some
applications.)

The OPA1678/9's bias and offset currents are ~10 pA at low temperature,
but go nuts above 25C, getting up into the nanoamps at 85C.  That's not
horrible for a JFET, but since it's a CMOS part, I gather that something
funky is happening with the input protection circuitry.

Anyway, one for the tool kit.  Anybody used it?

Cheers

Phil Hobbs



(*) Avnet claims to be selling the ST Micro TL084 for 5.6 cents in
reels--cheaper than an LM324A!

The step-up bipolar jellybean from the LM324 I like is the STMicro
TS462. Lower max supply voltage, GBW (12 vs 16), and slew rate (4 vs 7)
than the MC33078. Worse max offset voltage than the MC33078 5mV vs. 2.

Lower input bias current, lower noise, operates down to +/- 1.25 volts,
and RRIO. Approximately the same price in quantity

<https://www.mouser.com/datasheet/2/389/ts461-957349.pdf>

 

On Friday, 3 May 2019 19:40:14 UTC+1, John Larkin wrote:

On Fri, 3 May 2019 05:26:57 -0400, Phil Hobbs

(*) Avnet claims to be selling the ST Micro TL084 for 5.6 cents in
reels--cheaper than an LM324A!

1.4 cents per opamp. That's insane.

5.6c a quad is too much

These used to cost over $100 each,
when $100 was worth a lot.

https://www.dropbox.com/s/9idlgfug24rqxyb/Philbricks.jpg?dl=0

Prices have come down since Babbage's machine too. Didn't he need a national government to bankroll him?


NT

 

[bitrex]

On 5/3/19 2:39 PM, John Larkin wrote:

On Fri, 3 May 2019 05:26:57 -0400, Phil Hobbs
pcdhSpamMeSenseless@electrooptical.net> wrote:

We've been going back and forth on the LM324/358 etc. recently. One of
my lowish-noise, low cost faves has long been the bipolar MC33078/9,
whose decent combination of noise, bandwidth, cost, DC performance, and
output swing is fairly stellar at about 20 cents in reels, just a bit
more than a knockoff TL084. (*)
It isn't a true single-supply op amp like the 324, but on the other hand
it actually works at AC.

TI recently introduced the OPA1678/9, which are some sort of weird CMOS
riff off the old Motorola chips, and are sort of interesting. They're
twice the price, but a 40-cent quad won't break the bank for most of us.
The specs are quite similar, except for the improved input bias
current and lower distortion. (The 33079's 800-nA maximum bias current
and 500 fA/sqrt(Hz) current noise are pretty annoying in some applications.)

The OPA1678/9's bias and offset currents are ~10 pA at low temperature,
but go nuts above 25C, getting up into the nanoamps at 85C. That's not
horrible for a JFET, but since it's a CMOS part, I gather that something
funky is happening with the input protection circuitry.

Anyway, one for the tool kit. Anybody used it?

We use some OPA171s which look very similar, same bias current effect
at temperature. Nice little EMI hard part. Well behaved.


Cheers

Phil Hobbs



(*) Avnet claims to be selling the ST Micro TL084 for 5.6 cents in
reels--cheaper than an LM324A!

1.4 cents per opamp. That's insane. These used to cost over $100 each,
when $100 was worth a lot.

https://www.dropbox.com/s/9idlgfug24rqxyb/Philbricks.jpg?dl=0

Bob Pease gave its price over much of its mid-century life (1950-1970?)
as $22, about $100 in 2019 dollars.

If the same essential design was on sale at that price for 20 years, at
that time (before my time...) I don't know, but by about 5 years in it
was already overpriced and under-performing. Ahh,
backwards-compatibility market.

Adding one or two germanium PNPs available by the mid-late-50s to the
circuit making it a hybrid I think could've probably lowered costs and
boosted gain/bandwidth performance significantly while keeping the other
compatibility specs the same, but hindsight is 20/20 I guess

 

[bitrex]

On 5/4/19 12:20 PM, bitrex wrote:

On 5/3/19 2:39 PM, John Larkin wrote:
On Fri, 3 May 2019 05:26:57 -0400, Phil Hobbs
pcdhSpamMeSenseless@electrooptical.net> wrote:

We've been going back and forth on the LM324/358 etc. recently.  One of
my lowish-noise, low cost faves has long been the bipolar MC33078/9,
whose decent combination of noise, bandwidth, cost, DC performance, and
output swing is fairly stellar at about 20 cents in reels, just a bit
more than a knockoff TL084. (*)
It isn't a true single-supply op amp like the 324, but on the other hand
it actually works at AC.

TI recently introduced the OPA1678/9, which are some sort of weird CMOS
riff off the old Motorola chips, and are sort of interesting.  They're
twice the price, but a 40-cent quad won't break the bank for most of us.
  The specs are quite similar, except for the improved input bias
current and lower distortion.  (The 33079's 800-nA maximum bias current
and 500 fA/sqrt(Hz) current noise are pretty annoying in some
applications.)

The OPA1678/9's bias and offset currents are ~10 pA at low temperature,
but go nuts above 25C, getting up into the nanoamps at 85C.  That's not
horrible for a JFET, but since it's a CMOS part, I gather that something
funky is happening with the input protection circuitry.

Anyway, one for the tool kit.  Anybody used it?

We use some OPA171s which look very similar, same bias current effect
at temperature. Nice little EMI hard part. Well behaved.


Cheers

Phil Hobbs



(*) Avnet claims to be selling the ST Micro TL084 for 5.6 cents in
reels--cheaper than an LM324A!

1.4 cents per opamp. That's insane. These used to cost over $100 each,
when $100 was worth a lot.

https://www.dropbox.com/s/9idlgfug24rqxyb/Philbricks.jpg?dl=0



Bob Pease gave its price over much of its mid-century life (1950-1970?)
as $22, about $100 in 2019 dollars.

Reference:
<https://www.electronicdesign.com/analog/whats-all-k2-w-stuff-anyhow>

 

[Winfield Hill]

bitrex wrote...

On 5/4/19 12:20 PM, bitrex wrote:
On 5/3/19 2:39 PM, John Larkin wrote:
On Fri, 3 May 2019 05:26:57 -0400, Phil Hobbs

1.4 cents per opamp. That's insane. These used to cost over $100 each,
when $100 was worth a lot.

https://www.dropbox.com/s/9idlgfug24rqxyb/Philbricks.jpg?dl=0

Bob Pease gave its price over much of its mid-century life
(1950-1970?) as $22, about $100 in 2019 dollars.

Reference:
https://www.electronicdesign.com/analog/whats-all-k2-w-stuff-anyhow

I used a few of those K2-W tube op-amps, when I started at Harvard
in Chemical Physics, in the mid 60s. Their high output voltage was
very convenient. Later we used semi op-amp modules. Then came the
uA702 (arggh, all-NPN, +12 -6V supplies, +/-3.5 volt swing into 10k)
and the uA709. Finally things started to settle down with the LM301
and uA741. I made a slow ramp integrator using uA741, IIRC, but it
suffered from crazy offset-voltage jumps, soon enough we learned
about popcorn noise!


--
Thanks,
- Win

 

[Winfield Hill]

bitrex wrote...

https://www.mouser.com/datasheet/2/389/ts461-957349.pdf

correction they're RRO not RRIO.

Yes, it's not even single supply, that's pretty limiting
for a low-V op-amp.


--
Thanks,
- Win