I am looking for an alternative to the ST micro. BUV48AFI npn bjt. It is a 450V Vceo device in a isolated package (ST proprietry ISOWATT-218 package). ST are discontinuing it and they have no suitable alternative.
It is not being used in a switcher so these related characteristics (ton etc) are irrelevant - it is being used in linear mode. What is important is the package which must be a similar physical size (eg TO-3P,TO247 ....),it must be isolated (fullpack,FP or whatever) and the beta must be 30+ at an Ic of 200mA. Max circuit Ic is 0.5A
Unfortunately a T220 will not do.
Required ~2kper month ongoing for next few years (min 3 yrs)
Bulk buying stock for the next 3 years is certainly an option that has been considered. Somehow I dont like the idea of knowingly using a part which is no longer available. Somebody's law says that many more, or zero, will be required and that a bulk purchase will all be right at the bottom of the spec (devices with a particularly low gain cannot be used).
Win - I am sure you must have come across such an animal as is needed here. I live in hope.
What does this part do, what's it used in? How much current and voltage are you switching, what's the flyback voltage?
It's interesting that the BUV48 is mentioned as a high-voltage power transistor in ST's old an656 app-note on reading power- transistor datasheets. Hmm, I wonder if that had anything to do with its being chosen?
Let's talk about power-transistor packages. As you know, I'm sure, the TO-220 and its cousins were designed to be able to replace TO-3 metal-can packages, in that the hole in the tab and the base and emitter leads are located to drop right into an unmodified PCB. Now, as for whether a TO-220 or some other larger package is used, the parameter to consider is the part's thermal resistance. And, as it happens, this is determined primarily by the size of the die inside the package, and not by other aspects of the package, except of course, an insulating layer, like you have in your BUV48AFI.
BTW, the BUV48 has a rather large die, no doubt associated with its capability to switch 15 amps, although one has to say, not very well! That's because at 15A it has a high 5V saturation voltage, despite being driven by 3A of base current, for a low saturated-drive beta of only 5. Anyway, I'd say the large die puts this part in an small group. (This part's low gain, BTW, comes from its high 1000-volt Vces rating. Do you need that capability?)
I doubt you're using the BUV48 at such high currents, but as stated, besides saturation voltage at a given base drive, an important parameter is the part's thermal resistance. One transistor you could consider is ON Semi's MJF18008, which comes in a "fullpack" insulated package.
Vce(sat) Vces at Ic RthJC BUV48AFI 1kV 1.5V 8A 2.2C/W MJF18008 1kV 0.7 4.5 2.78
This part does have a smaller die, hence the lower Ic value used for the saturation spec, and the higher Rth specs, but it's not a wimp; it's saturation-voltage plot shows about Vce 1.5V at 8A, like the BUV48. Anyway, any replacement decision would depend on how you're using the transistor.
There are larger parts, of course, but not many made with the fully-insulated feature. Fairchild has an insulated high-current part, but with a slightly lower 800V Vcbo.
Vce(sat) Vces at Ic RthJC BUV48AFI 1kV 1.5V 8A 2.2C/W FJPF13009 800V 1.5V 8A 2.4C/W
That one looks pretty similar to yours, and shows that the TO-3P, etc., vs. TO-220 package isn't responsible for the high-current performance.
The device is used in linear mode with maximum collector current of ~70mA with a Vce of ~ 100V max in one app. and as a 300mA switch on a
170V supply in another.The important features are Vceo (Vces is not applicable !!) of 400V or more (noisy environment when transistors off ), the Beta which would be nice to be 30 plus, and the package which must be isolated and mechanically compatible with the ISOWATT-218 in a complex (ish) mechanical assembly that has been specifically designed around this package. The TO-3P package is similar enough to fit as probably would a (FP) TO247 if anyone made such an animal. The TO220 will not.
No I didnt design it - why people design non industry standard parts into small scale production items that they assume (hope) will have some production longeveity bemuses me. When a change of a BJT type requires a major mechanical re-design is unfortunate. When a device is sourced that does not give the prime requirements on its datasheet(s) is *&^%ing madness. (Note the ST datasheet for the BUV48AFI offers neither the gain of the device at anywhere near the currents used, nor the isolation voltage). Whilst several manufacturers offer/offered BUV48A's only ST provided it in the ISOWATT-218 package. ST hace several parts in this package with appropriate Vceo's but their current gains are quite low - 'typically' in the low 20's at the currents of interest which will probably mean some fair number being supplied with a gain in the teens. These would have to be rejected.
The application crys out for MOSFETs, and no doubt one that fits mechanically is available, although the 4 to 6-volt gate voltage- drive would probably be an issue on the existing PCB circuit. Certainly the 300mA switch would benefit from being a MOSFET.
If you look at the product list of ST devices in the same function group there are a heck of a lot marked NRND and I would be fairly confident thay aren't about to chop all of them in the very near future. If they do it doesn't leave them with much.
My reading of the ST product status is that they still have stock and as yet there is no dicontinued notice. Have you actually contacted ST to ascertain a date for discontinuance or are you just going on the NRND recommendation?
I was not able to locate too many devices which would fit your requirements but there are a few in the Sanken range which will do it
Thanks for the response - the 2SC4577 will certainly be investigated.
ST have informed that date that the last orders for the BUV48AFI (and quite a number of other horiz. deflection/switcher orientated transistors) will be accepted is sept 2007. Stocks will obviously remain for a while and the obsolete parts purveyors will no doubt already have been allocated provision.
Yesterday we added thousands of new pdf data book scans to DatasheetArchive.com. We also improved our cross reference tool to try and match parametric data from our database. Run some of the part numbers through our site and see if its any help.