Op amp selection using DK's parametric search


#1

This post discusses the use of Digi-Key’s parametric search for selecting op amps. It presumes that you’ve got a grip on op amp theory and know what you’re looking for, and offers pointers on how to use the parametric filters effectively to come up with a short(er) list of candidate parts. If you’re looking for help with op amp theory and application, this is not the resource you’re looking for–this other one would probably be better.

Task number one in shopping for op amps is just finding the things. One technique is simply to search for “op amp” to get the primary product family of interest near the top of the results screen, where it’s more conveniently clicked.
image

The vast majority of the op-ampish parts a person might commonly look for are filed there, though a few other product families do exist that may contain product of interest for more specialized use cases. If you look through the Linear – Amplifiers – Instrumentation, OP Amps, Buffer Amps family with no luck, hitting the “Integrated Circuits” link in the breadcrumbs at the top of the results screen brings up a listing of all families in that product category. The other families containing op amp-ish products are adjacently filed, prefixed with a “Linear - Amplifiers - ____” product family name.
image

image

After finding the main op amp family, one is confronted with a wide array of parametric filters to choose from.


Their individual usage is outlined below in left-right order as presented, and people often ask “which should I pick first?” The All-Purpose Engineering Answer (“It depends…”) applies here. Most applications have a few basic requirements that are all but non-negotiable, supply voltage being one of them, and package type (SMT vs. through hole) often being another. Start with those sort of mundane must-haves. Beyond this, there’s usually only one or two performance parameters that are of particular import, the trick of course being to identify which among them is key in any given application. That’s a theory and application question, and outside the scope of this post.

Manufacturer:
If you’re a brand loyalist or have sworn never to use brand X again, you can make selections in this column to match those preferences. Save for any confusion caused by the recent flurry of electronics companies buying and selling each other, this is pretty straightforward.

Packaging:
This deals with the manner in which products are contained, and has some purchasing implications—if you want to buy a total of 3 op amps for example, you don’t need to see results for packaging styles that have a 5000-piece minimum order quantity, and probably don’t care to pay to have your three parts neatly put on a Digi-Reel. In that case, you’d select everything save for “Digi-Reel” and “Tape and Reel,” though for op amps the “Cut tape” and “Tube” listings cover the vast bulk of available manufacturer’s part numbers.

Series:
Manufacturers’ brand or product line names are listed and selectable here. Though not as helpful in context of op amps as for some other product families, listing it is a canonical DK business practice, so there it is…

Part Status:
Here, one can filter product listings based on their lifecycle stage. “Active” (in current production with no impending discontinuance declared) parts are preferred for new design activity for reasons that should be obvious, though a combination of the “In Stock” and “Obsolete” filters can be amusing on those days when you’re feeling mischievous.

Up to this point, the parametric columns are all pretty much boilerplate, and the concepts can be transferred and applied throughout most of the DK product portfolio. Now, we start getting into the actual op-ampy stuff…

Amplifier type:
This parameter reflects the dominant adjective used by the supplier’s marketing team to describe a part. Many of the values listed represent significant variations on the generic op amp architecture (Instrumentation, Isolation, and Logarithmic for example) and thus are quite meaningful as well as distinctive enough to be properly identified as the dominant adjective for the part. There are also other values in this column that describe relatively minor implementation variants on a basic op amp (JFET, CMOS), reflect non-defined marketing terms (Audio), reflect common architecture classes that are not mutually exclusive with the other terms (Voltage Feedback, Current Feedback), and terms that aren’t particularly helpful, despite a refreshing air of candor (General Purpose).

Due to the mixed context of what this parameter represents and the non-orthogonal nature of the selectable values, this parameter should be used sparingly and with great care. While it’s mostly safe for selecting distinctive variants on the basic op amp (InAmps, isolation amps, etc.) it’s likely to cause more trouble than it’s worth if what you’re looking for is represented schematically by a standard op amp symbol.

Number of Circuits:
If you have a preference for the number of distinct amplifier circuits combined in a package, this is where you can select for that. The broadest selection of amplifier characteristics is found among the 1 and 2-circuit devices, with the triples tending to be products developed with an eye toward video applications.

Output type:
Characterizations of a device’s output structure are reflected here. This parameter is an unfinished work, and while the dataset that feeds it is reasonably accurate in what it does contain, it’s nowhere close to being complete—roughly half of the parts in the family have no value assigned. As a result, if making a selection from this column and you want to be thorough, include the dashed (null) value to avoid missing out on viable candidate parts. You’ll have to sift through a lot of false positives, but you won’t be summarily eliminating half the product portfolio.

Slew Rate:
Characterizations of amplifier slew rates are influenced by a number of factors, and not everybody who makes op amps does it exactly the same way. Consequently, the data shown here should be considered an approximation, and understood to come with a lot of qualifications that aren’t exposed in the parametric search. When filtering using this parameter, it’s recommended to include values a bit on the “wrong” side of your target value along with those exceeding the target by a wide margin, along with the dash value. So doing will minimize the risk of eliminating valid candidate devices for the application.

Gain Bandwidth Product and -3dB bandwidth:
GBP and -3dB BW are common metrics for measuring an amplifier’s overall “speed” and have a direct correlation with slew rate. Being generous with one’s parametric selections here also is advised—pick the continent that the amp you want lives on, not its street address.

DO NOT MAKE SELECTIONS IN BOTH COLUMNS!!!
USE ONE -OR- THE OTHER!!!

It is uncommon (<10% of current listings) for an amplifier to be characterized using both parameters. Voltage-feedback devices tend to quote a GBP, whereas current-feedback devices are usually characterized using -3dB BW; the figure of merit that’s more application-independent differs between the two architectures.
At the time of writing, recent developments on the DK site have led to the lack of an ability to include products with dashed (not specified/waiting to be populated/not applicable) values in search results for parameters where range searching is enabled. Because the two parameters are rarely both defined for a part, making selections from both of these parametric columns will eliminate over 90% of the parts in the product family, even if every available value in both columns is selected.

Current-Input bias:
The entries in this column generally represent a manufacturer’s “typical” characterization value, which might be an order of magnitude less than a guaranteed maximum, if a manufacturer even offers one. In situations where bias current is a driving concern, making selections here that are somewhat more restrictive than allowed by one’s error budget can take up some of the slack between a “typical” and “maximum” value, and help narrow down the list of results to the best-qualified candidates a bit more quickly.

Voltage-Input Offset:
Values in this parametric column may reflect either a “typical” or “maximum” value, and the difference between the two again may involve moving a decimal point. Offset voltage seems more likely than bias current to come specified with a guaranteed limit, however the test conditions under which the quoted values apply can vary rather significantly; some may apply only at a stable 25°C for example, others may apply over a part’s rated temperature range.

Given the variability in qualifications that attach to values listed in this parametric column, some strategy is needed when an application cites offset voltage as a driving criterion. Making a more restrictive selection than called for by the application will reduce the number of false-positives one has to sift through, while likely eliminating some viable candidates. On the other hand, an initially broad selection can help limit the field for application of other important selection criteria, after which a person can return to the offset voltage as a guide for making final selections.

Current-Supply:
Values in this column may reflect either a manufacturer’s typical or guaranteed maximum spec, and factors such as the supply voltage used and device temperature will have an influence on actual observed behaviors. Often as not, it’s a parameter that will inform decisions among competing parts chosen on the basis of other criteria rather than being a chief concern in itself, though it can be useful for narrowing the initial candidate field in applications where a definite power budget has been established.

Current-Output/Channel:
Values in this column reflect manufacturer’s indicated maximums. Should an application require a device to supply significant output current, selecting all devices with a listed value greater than the design requirement is a good place to start. If significant output current over an electrically long time period is anticipated, make a selection that offers some margin between the application’s requirement and the listed value; riding the limit on this parameter generally doesn’t yield good results.

Voltage – Supply, Single/Dual (+/-)
Many applications have available supply voltages pre-established, and in such cases this is likely the first parameter for which one will want to make a selection.

Values shown in this column commonly take on two forms:

  • Those formatted ±X~Y indicate ‘dual-supply’ devices which require signals to be biased somewhere near the midpoint of the span between supplies

  • Those of a form 2U~2W, ±U~W indicate devices capable of ‘single-supply’ operation, which implies a common-mode input range that includes the more-negative supply rail. Though very common, an output capable of swinging near the more-negative supply rail is not necessarily implied for a “single-supply” device; if that’s a needed capability, verify with the datasheet prior to making a final selection.

When selecting devices on the basis of available input voltages, first identify whether the external circuit requires a “single supply” amplifier, i.e. if the common-mode input will closely approach the more-negative supply rail. If so, one need not select any parametric values of the form ±X~Y. If not, thinking in terms of the total span between the supply rails can be helpful, as this allows looking toward the left of the column when making selections among the single-supply capable devices.

On a first pass, select all values incorporating the intended supply voltage. Better results are often obtained when the intended supplies are nearer the middle or upper end of a device’s allowable range, as the minimum supply values are often little more than the sum of a device’s headroom requirements; yes, a device might technically operate within spec on as little as 5v worth of supply, but if the inputs and outputs are only allowed to wiggle within a 0.25v window in that case, it’s not going to be very useful for many applications…

Operating temperature:
To a degree (sorry…) this parameter describes other information offered about a part, as much as it describes the part itself. When a manufacturer offers a guaranteed limit for some non-ideal part behavior such as bias current or offset voltage, that limit carries some temperature qualification. Quite often, it’s stated as the rated operating temperature range of the part.

Choose parts with a rated temperature range inclusive of projected operating conditions if at all possible; when things go amok, one doesn’t want to be in a position of defending a choice to operate parts outside their stated limits. While the physics police do offer some leniency in the enforcement of temperature ratings, you’re on your own if you get caught violating them…

Mounting type, Package/Case, and Supplier Device Package:
The mounting type parameter lets a person choose among several common classes of physical device packaging, and can be viewed as a sort of macro filter for the other packaging parameters. The Package/Case parameter calls out physical device packaging using a standardized nomenclature, while the Supplier Device Package reflects the same thing using whatever nomenclature the manufacturer chooses. Pick whatever package you want. As a loose rule of thumb, if you want small parts, package designators with lots of letters from the back half of the alphabet are a good place to look…