Why do companies create their own part numbers for electronic components?

[philby]

I’ve worked at a few Electronics companies and, with the exception of my current company, they have always created their own unique part numbers for components. For Example, a 1k resistor might be assigned the code RES001K0 as opposed to the manufacturers code of XHE0010TJNM862.

What is the logic behind assigning new part numbers that are unique to your company to the electronic parts you buy and use in your designs?  I’m thinking about putting forward a proposal that we change over to using an internal numbering scheme.

My reasons
-Part numbers can be made shorter, logically grouped together depending on component type and easier to remember.
-If a component has supply problems or is made obsolete, the manufacturers part number is not trapped in BOM’s and need to be updated in multiple locations.
-If a component is made obsolete the old part numbers are not trapped in the schematic component parameters.
-Makes specifying multiple sources easier.

Any other reasons anyone can think of? I'm open to criticisms of the reasons I've listed. Can anyone recommend a good (interesting if possible) book on inventory management?

[miguelvp]

-Makes specifying multiple sources easier

This, but usually you can still search by manufacturer's part number, specially nowadays since if they didn't they wouldn't show up on google searches and all those spider crawlers or whatever they are called now.

[Dago]

I’ve worked at a few Electronics companies and, with the exception of my current company, they have always created their own unique part numbers for components. For Example, a 1k resistor might be assigned the code RES001K0 as opposed to the manufacturers code of XHE0010TJNM862.

What is the logic behind assigning new part numbers that are unique to your company to the electronic parts you buy and use in your designs?  I’m thinking about putting forward a proposal that we change over to using an internal numbering scheme.

My reasons
-Part numbers can be made shorter, logically grouped together depending on component type and easier to remember.
-If a component has supply problems or is made obsolete, the manufacturers part number is not trapped in BOM’s and need to be updated in multiple locations.
-If a component is made obsolete the old part numbers are not trapped in the schematic component parameters.
-Makes specifying multiple sources easier.

Any other reasons anyone can think of? I'm open to criticisms of the reasons I've listed. Can anyone recommend a good (interesting if possible) book on inventory management?

What if there are identical manufacturers codes since there is no guarantee they are unique? With your own internal codes you can guarantee the codes are unique.

[philby]

What if there are identical manufacturers codes since there is no guarantee they are unique? With your own internal codes you can guarantee the codes are unique.

That's a good one thanks.

[Tandy]

The reasons we use our own numbers are...

The part numbers help easily identify the category of the product, i.e semiconductors starting with 276 and resistors 271. Back when there were Tandy stores that meant that those products could be found on a particular display stand in the store. We use this same method for locating parts in our stock rooms. Part numbers like PIC16F232C and ATMEGA328P give no indication of where they are stored and to a non electronics engineer/enthusiast gives them no clue what the part is. So they would have to look on the computer to see what bin, tray or shelf the item is in. It is easier to know that (276-006) 276 is the semiconductor section 006 is near the beginning of the rack of semiconductors and 276-9000 somewhere near the end. (Example random made up part numbers not actual parts.).

Partly we have maintained the same scheme for historic reasons as the part numbers being used from as far back as the 1970s means that there are a lot of projects and even books (Forrest Mims for example) that reference Tandy/Radio Shack part numbers so people are using those part numbers to find parts.

A standardised part number makes printing barcode labels easier as the part number has a fixed maximum length so that we can be sure that it fits on a standard sized label.

The large distributors that act as agents for manufacturers for example stocking the same 2N2222A transistor from 6 different semiconductor manufacturers each with their own part numbers for example ON semiconductor P2N2222AG and Fairchild PN2222ATA are functionally the same part. We do the opposite however, in that we stock only 1 2N2222A TO-92 type transistor (276-2009) that could be from a number of different manufacturers. It might sound strange to an experienced person but people are often bamboozled by doing a search and being given 20 variations of the same part, not quite sure if there is some reason that they should choose one of them and what if they choose the wrong one. Instead if someone builds a circuit using our part number 276-2009 they can always build the same circuit even if the manufacturers change part numbers. A classic example is that many manufacturers changed their part numbers to indicate a change to RoHS compliant versions. Obviously that was important to engineers who had to be certain that they were including RoHS compliant parts in their designs but the change in part number just added confusion to most hobbyists. So we just continue to use the same 276-2009 number for the RoHS version so that the hobbyist doesn't need to worry about it.

[Richard Crowley]

Obfuscation is another reason. Many products and circuits are proprietary and protected only as a "trade secret". Clearly, calling an ordinary resistor or capacitor by a different name won't prevent someone from identifying the component, but putting "house numbers" on otherwise-identical integrated circuits will at least make it that much more difficult to reverse-engineer.  We have seen several examples of this in Dave's tear-down videos.  And I have seen several examples where the printing on the IC was obliterated and left completely blank.

[AndyC_772]

A company's internal numbering scheme can carry relevant, pertinent information that just using a manufacturer's number cannot. For example:

- on a given product, it might be known that two specific manufacturers' alternative devices are tested, qualified and approved for use. A single part number's database entry can list these two approved alternatives, and over time, can be modified if and when parts go EoL and new ones become approved.

- alternatively, for non-critical components like standard resistors, there may be no specific manufacturer's part number. Any 0603 1k 1% resistor is pretty much the same as any other, and using the company's part number can allow this

- certain ranges of part number can be defined to have special meanings. For example, you could say that any part in the range 90000-99999 is "special", and has certain requirements associated with it which must be met, and which are noted in the database. For example, "do not substitute", or "safety critical item", or "calibration certificate must be supplied with this item", and so on.

- another field can indicate a revision number, so 10000-02 is the part which supersedes 10000-01. If some drawing calls up the -01 suffix, and there's a -02 on the database, you know straight away that you need to check if the drawing should be updated.

My general approach has always been to try to avoid duplication of effort wherever possible. If, for example, a part goes EoL, it's much quicker to add an identical equivalent part to a database than it is to go through every single schematic that uses it, change those, then change every document that refers to those schematics, then every document that refers to those documents, and so on.

[Galenbo]

For Example, a 1k resistor might be assigned the code RES001K0 as opposed to the manufacturers code of XHE0010TJNM862.

-when selling spare parts for your product to customers, giving another name gives them no direct link to a cheaper option.
-if multiple vendors make the same object, or multiple objects fit, RES001K0 can be a group name for compatible items.
-it's illegal to use the code name of vendor1 and also use it in you company for "similar" items from other vendors. We lost a lawsuit in 1985.

[m100]

-it's illegal to use the code name of vendor1 and also use it in you company for "similar" items from other vendors. We lost a lawsuit in 1985.

Similar thing with two major parts vendors in the UK, RS & Farnell, would be in the 80's or early 90's.   One of them, can't recall who,  produced a printed equivalent list showing their parts numbers alongside that of the competitor, they might have even shown prices.  It lasted a few months, maybe a couple of catalogues,  and then something happened,  maybe legal action or a few phone calls and it was never seen again. 

Also back then, or at least through to the early 80's logic and linear IC's from RS were marked or overmarked with their own generic part numbers.    You knew you were getting a 74LS74 for instance but with no idea who manufactured it.  No traceability meant their bits never got used in a lot of repairs on critical kit.  Some parts you could wipe with isopropanol or toluene and the true part number would be apparant, but one week it would be TI the next NS.   

[Tandy]

... then something happened,  maybe legal action or a few phone calls and it was never seen again. 

Most likely the manufacturers were unhappy with customers being shown possible cheaper alternatives. It would only take one of the large ones to say pull it or we will drop you as our distributor. Unlikely to be anything legal.

[m100]

The dispute was between RS & Farnell over the use of the other parties 'unique' part numbers and not those of their upstream suppliers.   The manufacturers own part numbers were not on the comparison list.  it was like a phone book in text density

Supplier 1 part Number 123-456   Our equivalent  345-7890
Supplier 1 part Number 123-457   Our equivalent  345-7891
Supplier 1 part Number 123-458   Our equivalent  345-7892

etc

You needed both companies catalogues open at the relevant pages to make any use of it, no internet of note back then and this was way before the web. 

[German_EE]

HP were terrible for this. That blown IC in your expensive piece of equipment could be an ASIC or a quad NAND but all you see is the HP part number on the top so they are the only source of spares. Bad news if support for your equipment ended ten years ago.

[philby]

Thanks for the replies everyone. Some good info thanks.

-it's illegal to use the code name of vendor1 and also use it in you company for "similar" items from other vendors. We lost a lawsuit in 1985.

Wow. Would have never thought of that.

[GK]

This reminds me of Dick Smith electronics back in the 90s. They stocked a range of BC series small-signal bipolar transistors, but they ordered them custom labeled DS instead. For example a BC548 was a DS548. They would have ordered in enough quantity so that the custom labeling wasn't a prohibitive cost but the point in the first place was lost on me.

[Galenbo]

-it's illegal to use the code name of vendor1 and also use it in you company for "similar" items from other vendors. We lost a lawsuit in 1985.

Wow. Would have never thought of that.

The strange thing is that it was discovered by the competition because by accident it appeared on the bill of the customer, but we weren't punished for that. We were only punished for using their numbers inside our organisation.

[LabSpokane]

Another aspect of creation of an internal part numbering system is the computer system (MRP/ERP, BOM control, etc.) that is needed to manage 1000's of parts, not just electronic, from manufacturers with totally in-congruent part numbering systems.  Even if every component one used was single source, it would be very difficult to use each manufacturer's native part number and have a BOM be comprehensible.  It also made doing reporting easy because it's simple to do a report on a particular type of inventory, work in progress, or finished product (see the coding system below).

A lot of this was driven by the needs of the computer systems in use during the 1960's - 90's, where the rule was mainframes running custom software written in COBOL with fixed field databases and reports generated on continuous, pinfeed printouts.  On a product with 100's of components, a uniform part numbering system made reading reports and indented bills of materials *infinitely* easier by humans, because everything fell into neat columns - something that would be impossible using a mismash of vendor part numbers.

Where I'm at, a lot of the part numbering systems used locally actually originated with General Instrument, so the part numbering schemes look like:

XXX-YYYY-ZZ

XXX: Category of component (i.e. IC, PCB, mechanical, plastics, sub-assembly, etc.)
YYYY:  Specific part in category
ZZ:  Variations of that part (color, length, or similar characteristics)

A company-specific part numbering system really is worth the effort, because a well-implemented one also makes life easy for the humans that have work with it.  A system like the one above is a lot like giving each part its own telephone number, and humans are surprisingly adept at remembering and collating many hundreds of parts in such a system.  I am constantly amazed by the ability of stockroom and production workers to remember components based on such a part numbering system. 
 

[vk6zgo]

Automotive parts were organised in a similar manner for many years before it started to be used in Electronics,so that GM may have used,say, 740-145-098 for a part,whereas BMC would have used another series of numbers for the identical part,Lucas may have called it something else again.

The car trade never had the saturation of standard part numbers from manufacturers that happens in Electronics.
On the Electronics maintenance side,Techs & EEs all use  the standard part numbers.

If I'm fixing something which uses,say a LM380,I don't give a rodent's exhaust port what the equipment manufacturer calls it,I'll grab an LM380 & get the thing working!

[Tinkerer]

My company does it because its easier to specify approved parts under one number that are essentially identical but otherwise made by different manufacturers. If you take the manufacturers number you end up with 10 different versions of a 100ohm resistor without any reason to.

[free_electron]

HP were terrible for this. That blown IC in your expensive piece of equipment could be an ASIC or a quad NAND but all you see is the HP part number on the top so they are the only source of spares. Bad news if support for your equipment ended ten years ago.

Nonsense. Hp had the decency of publishing the cross reference list. You can trace every hp number.

[MOSIUR540]

By making the unique part number, is it possible to reduce the counterfeit IC?

[MOSIUR540]

By making the unique part number, is it possible to reduce the counterfeit IC?

[jh15]

are the hp xrefs available now? As in BAMA, etc?

[srb1954]

HP were terrible for this. That blown IC in your expensive piece of equipment could be an ASIC or a quad NAND but all you see is the HP part number on the top so they are the only source of spares. Bad news if support for your equipment ended ten years ago.

Nonsense. Hp had the decency of publishing the cross reference list. You can trace every hp number.

Not every HP part number. On several occasions I have searched for HP part numbers and have not found them on any of the published cross reference lists.

[thermistor-guy]

...  I’m thinking about putting forward a proposal that we change over to using an internal numbering scheme.
...Any other reasons anyone can think of? I'm open to criticisms of the reasons I've listed. ...

Restricted characteristics: You may need a part to have a certain characteristic that only some manufacturers meet. Your internal p/n system can accommodate that. Examples that I've encountered:

** a telephony line interface circuit uses a common signal diode, but the diode's leakage current at Vr=100V must meet a certain limit. Only some OEMs guarantee it on the data sheet.
** a protection circuit relies on resistors that are rated for a certain overload condition. Only some resistor OEMs guarantee it.
** a product's connectors must have a certain gold plating spec. Only some connector OEMs offer that option.
** the PCB holes for a certain pin header were made too small, by mistake; luckily there is one OEM who mades headers that will just fit; you store those headers in your warehouse under a separate part number, for production and for servicing spares.

Traceability: You have a successful long-running product. Older units supposedly use the sames parts as current units, but the actual component OEMs used have changed over time. You're now getting strange field faults on some current units that don't appear on older units, and you're trying to relate the field failures to changes in component vendors. The history of changes to your part number system can help.

In-house component screening: you want to screen some components for a certain characteristic that no OEM guarantees. So you screen components in-house; the ones that pass get stored under a special part number.

[AndyC_772]

Note to all: this thread is six years old...

By making the unique part number, is it possible to reduce the counterfeit IC?

No. How would it?