If B-52s Can Keep Flying, So Can Your “Obsolete” Product

Let’s face it: the typical lifetime of many consumer-oriented mass-market electronic products is on the order of five years, give or take a few. After that, a product usually disappears (stored away, trashed, or recycled) and is replaced with a newer, shinier model.

In other applications, however, products are expected to have a longer viable life. Industrial, automotive, military/aerospace, medical, and test and measurement products can be in use for at least a decade or more. For example, the average car in the US is 12 years old, and that average is increasing due to improved reliability, coupled with the higher cost of new vehicles.

Then there are the mil/aero extreme cases. The long-range B-52 Stratofortress bomber — an eight-engine behemoth — was introduced in 1952 and has undergone eight major updates, with the last one rolling off the assembly line in 1962 (Figure 1). Although “common sense” says they should all have been scrapped by now, about 70 are still in active use from the approximately 750 that were built.^{[1] [2]}

Figure 1: Hard to believe but true: the B-52 Stratofortress long-range bomber was in production from 1952 to 1962, and about 70 of them are still in use. (Image source: Boeing)

Over its long operational life, the B-52 has undergone an extensive series of retrofits and modernizations as its role changed from a high-altitude bomber to a low-level penetrator to a stand-off weapons launch platform. The planes have had major upgrades in engines, avionics, weapons systems, and radar, just to mention a few subsystems. There are, however, still many original or nearly original parts in each plane, in addition to the airframe.

Keeping these aircraft flying involves scavenging parts from retired ones; if none are available, it requires getting form, fit, and functionally identical replacement parts and circuit boards made. The current plans are to keep upgrading and flying them until at least 2040, perhaps longer.^[3][4]

In many ways, replacing the mechanical parts of a system is easier than doing the same for electronics. For example, for an older car, it’s possible — admittedly at a cost in time and effort — to recreate an identical replacement part using CAD/CAM software, suitable materials, machining, and even additive manufacturing (3D printing).

Nor does this attribute of mechanical parts apply just to cars. Recently, a team was contracted by the US Air Force to reverse engineer and recreate a 300-piece “throttle quadrant” from the E-3 Airborne Early Warning and Control System (AWACS) — a heavily modified Boeing 707 with a huge flat radome mounted above the fuselage — by dissembling an existing unit piece-by-piece (Figure 2). ^[5]

Figure 2: This throttle quadrant from an E-3 AWACS radar aircraft was recently recreated using precise measurement and fabrication of each of its 300 pieces. (Image source: Verisurf)

They used a combination of tools, including basic calipers, advanced metrology systems, CAD/CAM software, close-up photographs, and other techniques to capture and then recreate this control unit to tolerances of better than 0.005 inches (in.).

But it’s a very different situation with electronics. When replacement parts are no longer available, the choices are limited and often painful. For example, the U.S. Air Force contracted to redesign and remanufacture the AN/ASK-7 data-transfer system for the B-52 bomber aircraft. The existing unit cannot simply be upgraded or modernized because the system has serious obsolescence issues.^[6]

The original system has 10 obsolete parts, with no known replacements across the circuit cards. The contracted vendor will therefore have to design and produce a form, fit, and functionally equivalent drop-in replacement for the legacy unit.

What’s the problem?

Long, useful life expectancy is not just for cars and aerospace situations; medical electronics may be in service for a decade or more. After all, you don’t just throw out a multimillion-dollar MRI or CAT scanner because the power devices that switch currents have failed and are no longer available from their original vendor. Even in cases where a redesign of a board seems feasible, it is usually much more than just another engineering project, as the replacement board design must be requalified to medical, safety, and operational mandates.

One solution is to consider a second source or alternate-source part. In principle, a second source part is the same but only from a different vendor. It may be reverse engineered by that second vendor to be the same (good), it may be made under license using the same tooling (better), or it may even be fabricated on the same production line (best).

In contrast, an alternate source is a pretty good attempt to duplicate the specifications of another vendor’s part, but maybe with some ­— hold on here — improvements, and so is dubbed an improved second source or ISS.

Except that’s where the trouble can begin. You never know what slight change in design details or specifications an ISS may have, whether it was designed to be slightly better and capture more sockets from that first source or for production reasons.

I learned that the hard way: Many years ago, I was involved in a project which used some basic 74XX TTL logic ICs (yes, it was that ancient). All the production boards were working fine — until one batch wasn’t. It took a week of long days and nights to figure out what happened.

It turns out that the purchasing agent bought some ISS versions of those logic ICs; they were touted as better, and they were — except that they had a slightly different output structure. Our lead designer, a clever circuit genius who devoured datasheets, had counted on the output structure of the original IC to eliminate the need for a few external resistors.

With a different structure, the circuit performance was erratic. Luckily, we were able to save the situation by doing some manual rework and adding the resistors, but that is a debugging experience I never want to repeat.

A much better option

One obvious solution to finding otherwise unavailable parts is to search the web and various “grey market” sites for the needed part. As attractive as that may sound, it is a very risky proposition. These parts may be used, mislabeled, out of spec, abused, or outright counterfeits that were designed to just barely get by functionally, at least at first. This scenario doesn’t end well in most cases. Even if these parts work, they are often marginal concerning performance, including temperature-related behavior.

There is a much better approach: going with a certified, legitimate source for parts that have been discontinued by their original vendor for various reasons. That’s where a vendor such as Rochester Electronics plays an important role. They obtain the end-of-run parts, as well as any unpackaged wafers and die, along with original tooling, mask sets, test procedures, and even equipment (if needed), and process details from the original source, all under a formal licensing agreement. Everything is stored under optimum conditions so there is no long-term degradation.

If you need a part and Rochester has it ready, you can purchase it immediately. If they don’t have any to ship, they can fab new ones that are identical to the originals. Given the present lack of available fab capacity for devices made using older processes, coupled with the shortage of used older fab equipment that could be put to use, having these multiple resources is a major advantage.^[7][8][9]

Getting an original of the discontinued part is often worth the effort. Consider a basic part, such as the CA3130AT 15 MHz, BiCMOS op amp with MOSFET input and CMOS output, a circa-1996 IC originally from Harris Semiconductor, available in the classic 8-lead TO-99 can (Figure 3).

Figure 3: The CA3130 op amp from the mid-1990s came in an 8-lead metal can, a package that is now largely as obsolete as the IC itself. (Image source: Rochester Electronics)

[Historical note: Harris Semiconductor was spun off by its parent company in 1999 and renamed Intersil Corp.; in turn, Intersil was acquired by Renesas Electronics Corp. in 2017.]

While there are undoubtedly very similar parts available, you never know what subtle second or third-order parameters and specifications make the circuit in which it is used function properly. These newer alternatives certainly don’t duplicate the original part’s schematic or fab process (Figure 4).

Figure 4: Imitated somewhat but likely not duplicated, the schematic diagram of the CA3130 op amp tells only part of its story as there are layout, process, and test issues to duplicate as well. (Image source: Rochester Electronics)

Further, even a minor circuit board re-spin to accommodate a new package type (who supplies op amps in 8-pin cans anymore?) can cause problems and unforeseen consequences, especially in sensitive, high-precision situations.

The reality is that the circuit designer who selected this part and the associated design-review notes (if any) are long gone; further, the designer may not even have been consciously aware at the time of the criticality of a subtle parameter. After all, the part worked in the circuit and system, and the design was tested, approved, and released to production.

Therefore, the use of a “pretty close” part as a drop-in replacement may bring immediate problems, or some that only appear later, or both.

Rochester-branded components are manufactured using either die or wafers purchased from the original suppliers, or Rochester wafers recreated from the original intellectual property. All re-creations are done with the approval of the original component manufacturer (OCM). Parts are tested using original factory test programs or test solutions developed by Rochester to guarantee that the product meets or exceeds the OCM’s datasheet.

Conclusion

In the consumer world, products with major electronic content are expected to be replaced with upgraded versions after just a few years of use. At the same time, OEM IC vendors may not have the resources to support their parts for ten, twenty, or even more years, but don’t want to breach the long-term relationship with customers. By formally licensing these parts and their tooling to an authorized vendor, problems associated with unavailable or alternate replacement parts can be avoided.

Related Content

“Rochester Electronics Provides a Unique Purchasing Experience Through DigiKey”

https://www.digikey.com/en/blog/rochester-electronics-provides-a-unique-purchasing-experience

References

1. Boeing, “Historical Snapshot: B-52 Stratofortress”

2. Boeing B-52 Stratofortress Association, “The Story of the Boeing B-52 Stratofortress”

3. The Wall Street Journal, “U.S. Pushes to Keep B-52 Bombers Going as Pressure From China Grows” (Oct. 17, 2022)

4. The Wall Street Journal, “For Wars of the Future, Pentagon Looks to Distant Past: The B-52” (Jan. 24, 2021)

5. Verisurf, “Reverse Engineering the Boeing E-3 Sentry’s Secondary Flight Controls”

6. Military & Aerospace Electronics, “Parts obsolescence forces redesign and remanufacture of AN/ASK-7 data-transfer avionics for B-52 bomber”

7. The Wall Street Journal, “Semiconductor Industry Isn’t Spending Big on Scarce Old-Tech Chips” (Nov. 9, 2021)

8. The Wall Street Journal, “Chip Shortage Has Manufacturers Turning to Lower-Tech Models” (Nov. 14, 2021)

9. The Wall Street Journal, “What’s Harder to Find Than Microchips? The Equipment That Makes Them” (Nov. 6, 2021)