History of bandwidth/feature upgradable scopes via license code

[Maik]

I need your swarm intelligence:
For a research project in the field of media archaeology I'm looking at scopes (and other gear as well) that are locked down by software to be later unlocked via a payed license code. Searching for those teams I mostly found either buying or hacking information on rather current models and for me it seems surprisingly hard to find infos on which manufacturer started this practice in the measuring tech industry.
Information about the same practice in other industries could also come in handy handy (for example similar to Intel’s "upgrade service" form a decade ago letting you unlock extra cache and features in some first and second generation core i processors).

I would be very glad if you could give me some hints in the right direction! Thank you very much!

[tom66]

We had a Tek scope from around 2005 that had serial EEPROMs to lock its options, on removable keys.

I remember that people had figured out they just contained plain text strings for the option values, so they got reverse engineered, though I'm not sure how common that was.

Also, I had a really old HP digital scope (mid 90's) that advertised options that could be installed, though it wasn't clear if these were hardware or software, and it seems no way to add them if not added in at the factory, without modding hardware or software.

[jwet]

There are really kind of two things here to understand I think.  One is selling a model that is "crippled" as a lower model, the Rigol 100 Mhz software hack and actually offering to sell users licenses to "unlock" things in their scope.

In the first instance, this is just manufacturing efficiency, economics and marketing.  It is simpler to manufacturer one assembly and rebadge it than to maintain two assemblies for what costs the manufacturer very little.  There may also be real savings in final test and calibration time but it isn't necessary.  Products exists in "tiers" in a marketing sense and there is a $300 tier that is distinct from a $500 tier.  There are different competitors, different customers and different needs.  The easiest way for a manufacturer to enter this market is to offer a "crippled" version.  This can be just to test the waters.  If the cheaper model does very well, the manufacturer can go back and actually cost reduce the system to increase margin though this seldom happens.  All manufacturers do this- its kind of trade secret which models share a common platform.

The other "after purchase upgrade" goes back a long way too.  If I was to guess, it probably started with IBM in the mainframe business.  Software has no direct cost of manufacturing to a first order.  Selling a base model and offering upgrades at time of sale or later is a good way to increase profits while keeping the base model price competitive.  If first saw this model in high end HP Test Gear in the 70's.  It was sort of a win win- I didn't have to buy capabilities that I didn't need and it gave HP a way to recoup their development expenses on those who did need them.

The other piece of this that's a bit subtle is that in business to business transactions, the initial price is often decoupled from the final price.  I sold chips to industrial equipment makers like Honeywell and Allen-Bradley.  If Keebler wants to make a cookie factory- they come up with a spec and shop it out to A&E firms that design and build turn key factories.  These factories will be built by the low bidder.  So the industrial suppliers try to make their base models cheap to help their customers provide low bids.  This comes out of a large capital pool for new construction in Keebler.  Once the system in installed, there is another budget, operations, manufacturing, etc that can be used to tailor the equipment with upgrades.  It was difficult to sell to these industrial guys because they wanted their base model to be very low cost but they could appreciate the performance you could add.  We had to find strategies to combat this.  Generally, its done with option cards, etc. if its hardware or with software switches.

I spent a lot of time thinking about these things as a Business Development Exec at a chip company.  There are a lot of creative ways to enhance value or perceived value of products.

[SmallCog]

The automotive industry is perhaps another example

Stereo systems in some vehicles only have digital radio in upper models. The hardware is there in the lower models but the feature is disabled.

Power outputs/tunes are another. In some product ranges lower model vehicles have the same engine just de-tuned to a lower power output.

[Berni]

Hard to say when it began, but it makes the most sense for midrange scopes.

The low end scopes are optimized for cost so they can't afford to drag on extra hardware(but is changing due to how easy scopes are becoming to make), the high end scopes are pushing what is possible so there is no room for artificial performance reduction. But the mid range scopes are many times more expensive than the low end stuff and are produced in lower volumes. So it makes sense for those models to share common hardware to get the benefit of the economy of scale. They have to develop hardware for the high bandwidth model, then save design time and manufacturing complexity by simply making that same hardware have less bandwidth.

It is not always done by software tho. Sometimes the different between a higher and lower bandwidth model is just a few component values in the analog front end. This lets them use the same PCB design and everything, then just feed the pick and place machine a different file. But this means once boards are built they are more difficult to swap. So if you have the change be in software you could make a production run of 10 000 scopes, keep them in a warehouse, then set it for the correct bandwidth for what the costumer ordered before it goes out the door. This way you don't get the problem of say the 300MHz model going out of stock while you still have lots of 500MHz models sitting there in the warehouse.

Due to the advancements in technology it is not that much more difficult to make a 500MHz scope compared to a 100MHz scope these days. So this is becoming more popular. A good modern example of this is the Keysight S series scopes, they come from 500MHz to 8GHz models. All of them being software upgradable to 8GHz

I also noticed that sometimes they will also use the lower models to use up badly performing ASICs. The Keysight 9000 series has the sample rate of 20GS/s, but the lowest bandwidth model only does 10GS/s. This is likely a way of using up the ASICs that couldn't pass quality control at full speed, but still worked at reduced speed. Much like what CPU/GPU manufacturers do with binning the same piece of silicon into different products depending on the test results.

[Someone]

Hard to say when it began, but it makes the most sense for midrange scopes.

Not really hard to say when it began, these big companies love their press releases which makes for accurate dates of "milestones".

[Berni]

Hard to say when it began, but it makes the most sense for midrange scopes.

Not really hard to say when it began, these big companies love their press releases which makes for accurate dates of "milestones".

They do not always make this public.

There are cases where the manufacturer offers bandwidth upgrades in the form of shipping the scope to a service center. Sometimes this means swapping out the board, it might mean swapping some passives, other times this means just software config. Why not offer this to buy in the form of a license key and new sticker? Might be that they don't want people to know they are this easy to upgrade, or they need calibration steps done to it after the upgrade for the scope to meet spec.

[Someone]

Hard to say when it began, but it makes the most sense for midrange scopes.

Not really hard to say when it began, these big companies love their press releases which makes for accurate dates of "milestones".

They do not always make this public.

There are cases where the manufacturer offers bandwidth upgrades in the form of shipping the scope to a service center. Sometimes this means swapping out the board, it might mean swapping some passives, other times this means just software config. Why not offer this to buy in the form of a license key and new sticker? Might be that they don't want people to know they are this easy to upgrade, or they need calibration steps done to it after the upgrade for the scope to meet spec.

Even when it was possible to board swap or update hard configs but was hidden from the user, those processes included recalibration (in the correction/adjustment/alignment sense, not the metrology sense). The step change was being able to seamlessly change the bandwidth without requiring re-calibration, which of course was exposed to the customers as quickly/easily as possible (for profit!).

[tautech]

The step change was being able to seamlessly change the bandwidth without requiring re-calibration, which of course was exposed to the customers as quickly/easily as possible (for profit!).

Hardly.
For decades cost effective production methods have used SW upgradable PCB's and some examples are TDS200, TDS1000 and 2000 models from as early as the 90's and likely others well before.
Owners of more recent equipment have spent countless hours sniffing data buses in efforts to understand the SW  BW limiting techniques and how they can be circumvented with just a SW patch or license code.

Today most models in a product series all share the same HW and are SW configured at the factory and despite the best efforts of manufacturers the motivated will go to some trouble to discover the model licensing/configuration methods used.

Manufacturers certainly don't disclose this sort of information however the collective efforts of communities such as EEVblog work tirelessly to discover it and share so everyone can benefit.....us sellers too ! 

[Someone]

The step change was being able to seamlessly change the bandwidth without requiring re-calibration, which of course was exposed to the customers as quickly/easily as possible (for profit!).

Hardly.
For decades cost effective production methods have used SW upgradable PCB's and some examples are TDS200, TDS1000 and 2000 models from as early as the 90's and likely others well before.
Owners of more recent equipment have spent countless hours sniffing data buses in efforts to understand the SW  BW limiting techniques and how they can be circumvented with just a SW patch or license code.

Today most models in a product series all share the same HW and are SW configured at the factory and despite the best efforts of manufacturers the motivated will go to some trouble to discover the model licensing/configuration methods used.

Manufacturers certainly don't disclose this sort of information however the collective efforts of communities such as EEVblog work tirelessly to discover it and share so everyone can benefit.....us sellers too ! 

Factory config is very different from field/customer change/upgrade, of course economy of scales had them all made from (in many cases exactly, some times almost exactly) the same underlying hardware. SmallCog mentions this above with the example of the Automotive industry, same engine, same parts, different map/tune (often one or two key parts like the clutch/gearbox/axle might be different) and a span of performance/price models. No one is disputing that.

Were the TDS200 modules calibrated identically and able to be changed in the field without readjustment? The service manual says, no, they were calibrated and adjusted differently. That someone can change their unit over is no different to retuning a car engine, not supported by the factory, and generally needs adjustment specific to the unit.

The OP very specifically asked about the field upgrade being exposed to the customer:

I'm looking at scopes (and other gear as well) that are locked down by software to be later unlocked via a payed license code