Touchstone Semi?

Despite my being in the custom I/C design business, I wouldn't design ANYthing in that didn't have at least two sources. Single source is just begging to be screwed. ...Jim Thompson

Unless it is from Linear Technology. That is one of the few companies I trust in that regard. Have to, because it would otherwise be next to impossible to design cutting edge switch mode supplies without an often timeline-prohibitive IC design costing six digits.

You can do that with something made of opamps and transistors and gates. But try to make a modern design with a fast CPU and good ADC, say You will have to make an awful lot of compromises doing that.

A decade or two ago I designed in one of the few "two sourced" CPUs, a C167. You could get ST ones and Infineon ones.

When the supply dried up, turned out they were all made by ST.

Having said that, never had a supply problem since with semis. It's always things like LCDs or printers.

Yep. Everyone, in the name of miniaturization, has lost all ability to do their own device-level design... except for me... but, then, I've been designing switchers since the early '70's ;-) ...Jim Thompson

Analog Devices is good too, as is TI.

For performance, sole-source is usually mandatory.

Looks like most of their products are performance improvements of existing products by other mainstream manufacturers. I'm being hit by these excrucia ting marketing videos of theirs where the product design engineer spends a full five minutes explaining and demonstrating something he could have put across in about 30 seconds of verbal. Looks like they are going after Digik ey customers.

There's nothing man can conceive that can't be done with a TI part, and no worries about discontinuing parts, face it, they're still making products pushing 45 years of legacy!

Let's just say I know who the A-team players were at Maxim, and I know who I would never hire let alone work under.

Uh, good luck with that part.

Also ditto the other posters on ADI and LTC being fine for sole source. Both are good companies. Owning your fab is a big deal in analog.

I have nothing but praise for LTC, though you have to wonder what the future holds since many there have retired or passed on. It is the home of the gurus.

worries about discontinuing parts, face it, they're still making products pushing 45 years of legacy!

I'll take that challenge! Touchstone has the TS3002 which is a very low power oscillator which can be used as a VCO at 240 kHz. I believe the current at that frequency and 1.8 volts Vdd is around 12 uA.

What does TI have that can match those specs?

ADI for switching supplies? I use a lot of TI's switchers (and FETs now, too). One boost regulator didn't work out so well, though. It's PSRR is simply terrible. The local FAE said it must be my layout, so gave me one of the EDKs. It's just as bad, if not worse. I changed to another one of their parts and everything is peachy. ...another reason first-pass boards never work for me. Invariably, something doesn't work as advertised.

Yep, but people here still argue that point. You have to go pretty far down the food chain to get parts that have multiple sources. Once you decide to cross that bridge, in for a penny...

worries about discontinuing parts, face it, they're still making products pushing 45 years of legacy!

TI used to have the charming habit of doing their market research by publishing data sheets for non-existent parts and accepting orders for them.

About twenty years ago I got burned by a TI op amp that had an unusually high input capacitance which they hadn't bothered to specify on the data sheet.

I tend to stay away from their parts if I can.

I don't think we use any ADI switchers. Lots of opamps, vrefs, ECL comparators, diffamps, ADCs, DACs, isolators.

I use a lot of TI's switchers (and FETs

It hasn't stung us much so far. It's amazing that most people keep thousands of parts in production.

ADI has bit us once or twice. AD1862, and some parallel DAC I forget.

Try this with TI:

TI is a pretty good company but there are areas where others took the lead. Even in switchers at times.

But my main gripe with them is that they have now managed to thoroughly screw up their web site. To me it has almost become useless. So even if they come up with a super-ueber-gizmo chip, it's become harder to find out about it.

worries about discontinuing parts, face it, they're still making products pushing 45 years of legacy!

That's a one-of-a-kind exception /for now/. If a market develops for it, TI will have one too, rest assured.

worries about discontinuing parts, face it, they're still making products pushing 45 years of legacy!

They have a lot of LT equivalents in linear at about 1/5 the cost, usually the specs are wider range than LT of course. But any unknown linear part needs to be beat up in prototyping before you commit to it, even to this day, even from LT.

That's outside the realm of what's considered "linear", it's an RF part, looks like something that should be in M/A Com's lineup.

worries about discontinuing parts, face it, they're still making products pushing 45 years of legacy!

specs are wider range than LT of course.

Why "of course"? TI's habit of making nominally industry standard parts with a sloppier-than-industry-standard data sheet is a feature of the company that I first ran into around 1975, and it's another reason to avoid them.

commit to it, even to this day, even from LT.

But you'd better evaluate the TI parts a lot more carefully than you'd evaluate parts from a more reputable supplier.

I have had conversations with multiple company reps about their latest website update. So many of them are so bad. I always tell them to go study the mini circuits site. It is (or at least used to be) like the drudge report of websites... very crude but very effective. It always struck me tht the mini-circuit guy knew exactly how to do it right, yet a web committee would have probably rejected his site as juvenile. It is a little more professional looking now, but they did not lose the good properties in doing so.

I especially pleaded with omni-spectra to build a proper website. They used to be the absolute dominant player in RF connectors. Could hardly find them on the web and then the site did not mimic their wonderful old catalog , but made you answer 10 questions about what connector you wanted. Everyone wanted a picture and yet the company could not figure this out. I once tried to actually pry my way into the org to plead with them to build a proper website.

ttdesign.com

I do not think we have had any product introduction for the past ten years which did not have an obsolete part in it when we finally got it certified and to market. We have a guy in our office whose only job is to manage obsolete parts.

Your decision needs to factor in the life expectancy of the product and the anticipated number of units the company expects to sell and how difficult it is to buy 10 years worth of the parts if it does go obsolete. will ten years worth of part be $1000 qorth of stock or $10,000,000 worth of stock. How important is it to get the product to market quickly?

I think the answer to your concerns can cut either way.