Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > sci.electronics.design > #431324 > unrolled thread

nice little opamp

Back to article view | Back to sci.electronics.design

Contents

  nice little opamp John Larkin <jjlarkin@highlandtechnology.com> - 2016-10-15 12:19 -0700
    nice little opamp Phil Hobbs <pcdhobbs@gmail.com> - 2016-10-15 13:24 -0700
      Re: nice little opamp John Larkin <jjlarkin@highlandtechnology.com> - 2016-10-15 14:03 -0700
      Re: nice little opamp Winfield Hill <hill@rowland.harvard.edu> - 2016-10-15 14:32 -0700
        Re: nice little opamp Phil Hobbs <pcdhobbs@gmail.com> - 2016-10-15 14:45 -0700
        Re: nice little opamp John Larkin <jjlarkin@highlandtechnology.com> - 2016-10-15 15:24 -0700
        Re: nice little opamp krw <krw@nowhere.com> - 2016-10-15 20:08 -0400
    Re: nice little opamp Frank Miles <fpm@u.washington.edu> - 2016-10-16 04:36 +0000
      Re: nice little opamp John Larkin <jjlarkin@highlandtechnology.com> - 2016-10-16 09:49 -0700
        Re: nice little opamp Frank Miles <fpm@u.washington.edu> - 2016-10-16 23:42 +0000

#431324 — nice little opamp

http://www.onsemi.com/PowerSolutions/product.do?id=NCS2005

This is rrio, 8 MHz, 32 volts max supply.

It seems to be c-load stable at 1 uF and up, ceramic or tantalum,
which means it can drive bypassed reference rails and low-current
power rails.

It rails clean, so will work as a comparator, but above 20 volt supply
the positive clamp gets a little strange.

40 cents each by the reel.


-- 

John Larkin         Highland Technology, Inc

lunatic fringe electronics 

[toc] | [next] | [standalone]

#431331

45 nV typical 1-Hz noise, though, yikes!

A fave of mine in the low cost/good performance regime is the MC33078, which is a dual with vaguely similar specs. It doesn't work well below 10V total supply, and has a lot of input bias current, but its 1-Hz noise is an order of magnitude less and it's cheaper: under 30 cents for the 8-VSSOP in reels from Digikey.

Cheers

Phil Hobbs

[toc] | [prev] | [next] | [standalone]

#431334

On Sat, 15 Oct 2016 13:24:58 -0700 (PDT), Phil Hobbs
<pcdhobbs@gmail.com> wrote:

>45 nV typical 1-Hz noise, though, yikes!

Noise doesn't matter much in the thing I'm doing now. 47 uF from the
output to ground helps, too.

>
>A fave of mine in the low cost/good performance regime is the MC33078, which is a dual with vaguely similar specs. It doesn't work well below 10V total supply, and has a lot of input bias current, but its 1-Hz noise is an order of magnitude less and it's cheaper: under 30 cents for the 8-VSSOP in reels from Digikey.

I'll regularly be using these from 3.3 to 24 volts. It's a faster,
c-loadable, cheaper version of LM7301, which we have used 46,000 of so
far. Threw away maybe $25K.

After I finish the design, maybe 34 sheets of schematic, we can review
it and maybe replace some single SOT23's with cheap duals. Gotta
consider pcb placement and pick-and-place feeders as well as the
electricals. Using "universal" c-load high voltage single rrio's is
easy.

Looks like the cheapest opamp that we have in stock is an LM358, 10
cents for the dual. Horrible amp.






-- 

John Larkin         Highland Technology, Inc

lunatic fringe electronics 

[toc] | [prev] | [next] | [standalone]

#431339

Phil Hobbs wrote...
>
> 45 nV typical 1-Hz noise, though, yikes!
>
> A fave of mine in the low cost/good performance
> regime is the MC33078 ... its 1-Hz noise is an
> order of magnitude less and it's cheaper: under
> 30 cents for the 8-VSSOP in reels from Digikey.

 I'm a big fan of SOT-23 op-amps.  It's too bad
 nobody ever adds that package for older dies.


-- 
 Thanks,
    - Win

[toc] | [prev] | [next] | [standalone]

#431340

>I'm a big fan of SOT-23 op-amps.  It's too bad 
 nobody ever adds 
>that package for older dies. 

ISTM that the popular parts among old cheap designs are usually duals or quads, which took advantage of all the real estate DIP packages afford. (What's the single version of the 324/358?)

Cheers

Phil Hobbs

[toc] | [prev] | [next] | [standalone]

#431346

On 15 Oct 2016 14:32:25 -0700, Winfield Hill
<hill@rowland.harvard.edu> wrote:

>Phil Hobbs wrote...
>>
>> 45 nV typical 1-Hz noise, though, yikes!
>>
>> A fave of mine in the low cost/good performance
>> regime is the MC33078 ... its 1-Hz noise is an
>> order of magnitude less and it's cheaper: under
>> 30 cents for the 8-VSSOP in reels from Digikey.
>
> I'm a big fan of SOT-23 op-amps.  It's too bad
> nobody ever adds that package for older dies.

The cheapest single opamp that Mouser shows is 16 cents at 1K.

The cheapest dual is 10 cents!


-- 

John Larkin         Highland Technology, Inc

lunatic fringe electronics 

[toc] | [prev] | [next] | [standalone]

#431360

On 15 Oct 2016 14:32:25 -0700, Winfield Hill
<hill@rowland.harvard.edu> wrote:

>Phil Hobbs wrote...
>>
>> 45 nV typical 1-Hz noise, though, yikes!
>>
>> A fave of mine in the low cost/good performance
>> regime is the MC33078 ... its 1-Hz noise is an
>> order of magnitude less and it's cheaper: under
>> 30 cents for the 8-VSSOP in reels from Digikey.
>
> I'm a big fan of SOT-23 op-amps.  It's too bad
> nobody ever adds that package for older dies.

I've been using the MicroChip MCP629x (x=1-4), which comes in a
SOT-23, SO-8, SO-14, and TSSOP-14.  Good price, if you can use a
million a year.  ;-)

[toc] | [prev] | [next] | [standalone]

#431378

On Sat, 15 Oct 2016 12:19:02 -0700, John Larkin wrote:

> http://www.onsemi.com/PowerSolutions/product.do?id=NCS2005
> 
> This is rrio, 8 MHz, 32 volts max supply.
> 
> It seems to be c-load stable at 1 uF and up, ceramic or tantalum, which
> means it can drive bypassed reference rails and low-current power rails.
> 
> It rails clean, so will work as a comparator, but above 20 volt supply
> the positive clamp gets a little strange.
> 
> 40 cents each by the reel.

Maybe my PDF reader is altering things, but I see only stability at 1nF 
load (in both data sheet and cover - feature list).  Transient responses 
are shown with only ~25pF load.  The voltage drift is rendering in uV/C, 
so the "micro" character seems ok.  {This is with both xpdf and evince on 
a linux system).

Perhaps you have some information from a source other than your link?

[toc] | [prev] | [next] | [standalone]

#431439

On Sun, 16 Oct 2016 04:36:08 -0000 (UTC), Frank Miles
<fpm@u.washington.edu> wrote:

>On Sat, 15 Oct 2016 12:19:02 -0700, John Larkin wrote:
>
>> http://www.onsemi.com/PowerSolutions/product.do?id=NCS2005
>> 
>> This is rrio, 8 MHz, 32 volts max supply.
>> 
>> It seems to be c-load stable at 1 uF and up, ceramic or tantalum, which
>> means it can drive bypassed reference rails and low-current power rails.
>> 
>> It rails clean, so will work as a comparator, but above 20 volt supply
>> the positive clamp gets a little strange.
>> 
>> 40 cents each by the reel.
>
>Maybe my PDF reader is altering things, but I see only stability at 1nF 
>load (in both data sheet and cover - feature list).  Transient responses 
>are shown with only ~25pF load.  The voltage drift is rendering in uV/C, 
>so the "micro" character seems ok.  {This is with both xpdf and evince on 
>a linux system).
>
>Perhaps you have some information from a source other than your link?
>

I experimented with one. It gets a little fuzzy with 10 nF loads, but
stabilizes nicely with microfarads.

Opamps are fairly unpredictable as regards their c-load behavior. RRO
parts are often capacitance tolerant, but it depends on how they do
the internal rolloff. Data sheets used to have internal schematics,
which is getting rare these days.

My ideal opamp is single SOT23, rrio, low noise, low offset, low bias
current, fairly fast, c-load tolerant, 3 to 30 volt supply, and cheap.

OPA197 is a candidate, but I haven't tested it yet. It's 67 cents in
reel quantity. The EMI filtering is nice. It works down to 4.5 volts
and can output +-40 mA, so it might be worth another 30 cents.












-- 

John Larkin         Highland Technology, Inc

lunatic fringe electronics 

[toc] | [prev] | [next] | [standalone]

#431479

On Sun, 16 Oct 2016 09:49:26 -0700, John Larkin wrote:

> On Sun, 16 Oct 2016 04:36:08 -0000 (UTC), Frank Miles
> <fpm@u.washington.edu> wrote:
> 
>>On Sat, 15 Oct 2016 12:19:02 -0700, John Larkin wrote:
>>
>>> http://www.onsemi.com/PowerSolutions/product.do?id=NCS2005
>>> 
>>> This is rrio, 8 MHz, 32 volts max supply.
>>> 
>>> It seems to be c-load stable at 1 uF and up, ceramic or tantalum,
>>> which means it can drive bypassed reference rails and low-current
>>> power rails.
>>> 
>>> It rails clean, so will work as a comparator, but above 20 volt supply
>>> the positive clamp gets a little strange.
>>> 
>>> 40 cents each by the reel.
>>
>>Maybe my PDF reader is altering things, but I see only stability at 1nF
>>load (in both data sheet and cover - feature list).  Transient responses
>>are shown with only ~25pF load.  The voltage drift is rendering in uV/C,
>>so the "micro" character seems ok.  {This is with both xpdf and evince
>>on a linux system).
>>
>>Perhaps you have some information from a source other than your link?
>>
>>
> I experimented with one. It gets a little fuzzy with 10 nF loads, but
> stabilizes nicely with microfarads.

Ah, from testing.  Seems similar to some of the TLx431 family loading 
zones.

Hopefully OnSemi won't change their process on you mid-production, 
continuing to meet their specs but not your expectations.

> Opamps are fairly unpredictable as regards their c-load behavior. RRO
> parts are often capacitance tolerant, but it depends on how they do the
> internal rolloff. Data sheets used to have internal schematics, which is
> getting rare these days.

Sad.  But as long as we buy the devices, the manufacturers will continue
to do this (and probably continue making data sheets less comprehensive).

> My ideal opamp is single SOT23, rrio, low noise, low offset, low bias
> current, fairly fast, c-load tolerant, 3 to 30 volt supply, and cheap.

And low power consumption, and guaranteed availability for ____ time.  
Please let us all know when you find this part!

> OPA197 is a candidate, but I haven't tested it yet. It's 67 cents in
> reel quantity. The EMI filtering is nice. It works down to 4.5 volts and
> can output +-40 mA, so it might be worth another 30 cents.

At least this data sheet shows more information regarding C-load behavior.

   Thanks,
   Frank

[toc] | [prev] | [standalone]

Back to top | Article view | sci.electronics.design

csiph-web