8/9/2023 0 Comments Hp transistor cross reference![]() ![]() Tek published their selected parts parameters (1982 was the last publication) - so you could sort a lot of cheapie JEDECs for Hfe or fT to meet the Tek spec, if you had the time. ![]() Now for certain Tek suffixed parts, you NEED the Tek part - a good example is some of the amp transistors in the 465/475 and variants - I've seen ringing/oscillation occur with off-the-shelf stuff.ĩ0-95% of the selected parts were selected to meet the extremes of Tek/HP specs. +1 on what Dale says - use the generic Jedec device where possible - I worked in 3 Cal Labs, and we never ordered the HP or Tek number, unless it was a suffixed Tek Part, or there was no cross in the Bench Briefs. Devices that I have replaced in situations where the device would almost certainly be selected (such as my recent replacement of the VCO JFET in a 675A) are not found in the Bench Briefs cross refs or in the Sphere cross ref compiled from them, and must be replaced with the devices that HP originally supplied. As I noted upstream they appear to be only for non-selected devices. HP was always designing right to the limits of technology and the way they did that was a lot of selection on the production line or on the vendor lines.Īre you saying that the cross refs that HP published in Bench Briefs were bogus? I have never had a problem replacing an HP part with a JEDEC part when it was listed in a Bench Briefs cross-ref. HP did wonderful designs that pushed the state of the art and the way they did in many places was to select the hell out of parts. The Sphere list will get you close in most cases and may get something to work. A lot of this old gear is getting fixed to function but not to meet the original specs. I have repaired quite a few late 60's HP transistorized test equipment trying to use the Sphere list and ended up having to select from a few dozen parts to get one that would work in the circuit and meet specs. HP was always designing right to the limits of technology and the way they did that was a lot of selection on the production line or on the vendor lines. JEDEC numbers were only good for really simple non critical things like a low frequency switch or simple amplifier. My Siliconix book lists hundreds of JFETS that came from a dozen or so Siliconix processes.Ī common 2N3904 from five different companies would be quite different so many times I had to spec a 2N3904 from Fairchild with a set of needed specs far more restricted than the JEDEC. Trying to design a VHF oscillator with a 2N Jfet is next to impossible because the critical parameters will vary too much from process to process. I used a lot of FET's from Rich's Siliconix and I always went by process number and a list of selected parameters to prevent purchasing from bringing in stuff from other vendors that would vary too much. National supplied hundreds of JEDEC numbers from 40 transistor processes and so did Motorola and many other companies. For a 2N3904 we had 13 part numbers all selected for some set of critical parameters.Īs a design engineer for 50 years the JEDEC numbers were next to useless. ![]() We had 7 internal part numbers for an MPSA-64 and that was a Motorola process. At my company it was pretty much the same. I worked for many years with engineers from both Tek and HP and they all told me that over 50 % of the semiconductors were selected in some way. Many of the parts on that list are listed like MPSa 64-5 the Motorola part is MPSA-64 the dash 5 is a selected part. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |