I like Fairchild's FSA66 and FSA266 parts. The FSA266 has the same charge injection specs as the FSA66 which is generally 10 times better than most IC switches.
-- Mark
Thanks, Mark!
You just saved my day! I just had a go-round with Analog Devices...
"We don't have a SPICE model that we can share with customers for this device."
I thought I was relegated to Maxim... shudder ;-(
Thanks, again, Mark!!
...Jim Thompson
--
| James E.Thompson, P.E. | mens |
| Analog Innovations, Inc. | et |
| Analog/Mixed-Signal ASIC\'s and Discrete Systems | manus |
| Phoenix, Arizona Voice:(480)460-2350 | |
| E-mail Address at Website Fax:(480)460-2142 | Brass Rat |
| http://www.analog-innovations.com | 1962 |
America: Land of the Free, Because of the Brave
Says 15ohms typical ON resistance at 3V, over signal range, whatever that means.
-- Regards, Joerg http://www.analogconsultants.com/
Doesn't matter to me... I'll be using it at the positive rail.
"Whatever that means": Look at some analog switch ON resistance curves... worst-case is mid-way between rails. (Remember I've redesigned a few :-)
...Jim Thompson
--
| James E.Thompson, P.E. | mens |
| Analog Innovations, Inc. | et |
| Analog/Mixed-Signal ASIC\'s and Discrete Systems | manus |
| Phoenix, Arizona Voice:(480)460-2350 | |
| E-mail Address at Website Fax:(480)460-2142 | Brass Rat |
| http://www.analog-innovations.com | 1962 |
America: Land of the Free, Because of the Brave
You might have to parallel two to get close to 6ohms at 3V. IIUC that would net a capacitance of about 10pF when open. Don't know how they got it this low.
What, AD as well? What is it these days? TI's application "support" has IMHO already all but collapsed.
That sure is a great device Mark recommended.
-- Regards, Joerg http://www.analogconsultants.com/
It's pretty flat unless you drop to really low supply voltages. Well, maybe you luck out and get close to 6ohms with just one.
-- Regards, Joerg http://www.analogconsultants.com/
In all honesty the IC makers seem to have become so accustomed to making highly specialized ICs (for those customers who want some specific function that they'll put in a million cell phones) that I think it's become surprising to them to see any "board level" engineers left trying to do actual *design* rather than just slapping parts together like building blocks the way their "design guides" suggest you should.
The "discrete device" designer is a dying breed, I'm afraid -- those who are pretty good at it are going to end up in IC design, I suspect. I just wish we had the technology so that someone like you could do an analog circuit design and get back packaged parts for the "nickels and dimes" (that you're fond of using) in reasonable quantities (hundreds or thousands?).
The questions I had were about more generic but new chips where datasheets leave a lot to be desired these days. I noticed a drastic deterioration in customer service, the most dramatic so far at TI. They simply do not respond anymore and I've heard that from numerous others as well. All I can say is that they lost the first design-in because I just had it with them and went discrete.
I am not so sure anymore. Had a chat with a new client last week and they were explicitly looking for a non-IC guy. They had a hard time finding one. Same for another client who needed one as an employee and it took us a whopping two years (we finally had to import one ...).
... I just wish we
That's exactly what companies in places like Shenzen do. However, inflation is also hitting those areas now so discrete assemblies might become a little more expensive.
-- Regards, Joerg http://www.analogconsultants.com/ "gmail" domain blocked because of excessive spam. Use another domain or send PM.
HI Joerg,
I mentioned to a TI rep just last Thursday (TI "Tech Day" in Portland) that we were seriously considering changing their battery charge controller IC out for an almost identical Linear Tech part since the TI part would tend to oscillate at times, and while we think we know why and have come up with a workaround, the dead silence coming from TI's tech. support e-mail was not encouraging. His only response was, "Oh, sorry about that." :-(
We do, however, have a couple of local reps who at least have the contact information for people at TI so with a bit of jumping through hoops we can still contact them. The process is annoying, of course.
Fair enough... I suppose then it's better to classify the discrete part analog designer as something of a rarity, then? I imagine the few that survive will end up commanding pretty decent salaries...
Do you know of any that'll run your designs as part of a mixed wafer for somewhat sane prices? (I.e., four-digit dollars?) You may recall that discussion about semi-custom ICs and the tradeoffs of the semi-custom array solution vs. just doing a full-custom approach, but the usual small-scale players for full-custom such as Mosis seemed largely targeted towards digital processes rather than analog or RF.
---Joel
I just had a problem with an altera config EEPROM EPCS16SI16N. For a start, the datasheet is buried in ch.4 of the FPGA handbook. And near the end is a pinout, but the package drawing is in another, huge, PDF file (along with every other package they make). so far so good.
but throughout the documentation the chip is described as a SOIC16. including in the packaging docs. If you read the dimensions, however....
Alas the chip body is 0.3" wide not 0.15" - its a SOIC-W. As we discovered when we went to solder it on the PCB.
So I turned it into a SOJ and soldered it on anyway.
the really frustrating part was when I laid out the PCB I noticed I could only get the SOIC16 from digikey, so I added that in parallel with the SOIC8 (which BTW is 0.15" wide), thinking myself very clever. doh.
Cheers Terry
It's time to rattle the cage there and write to executive management or the B of D. But I get tired of writing although at one mfg it has resulted in a very surprising response from up top. Boy did they wake up fast. I had written to the CEO and he must have exploded.
I am more concerned about the day when most of us will be at some senior's residence. Who is going to follow?
Four-digit is a stretch. IC design and fab costs have gone up a lot, to the point where some designs move back to discrete because the NRE just won't amortize. But it sure looks like you'll need MPW. So here goes:
You can try these folks which is basically the former Thesys GmbH:
Then closer to you guys but recently acquired by ON so I don't know how well they cater to smaller businesses these days:
AMS does analog, including HV processes:
I am sure Jim has a lot of other leads. At the end of the day you'll have to take a look at available processes, price them out and pick one. It's all buffet style in that market unless you have a huge budget.
-- Regards, Joerg http://www.analogconsultants.com/ "gmail" domain blocked because of excessive spam. Use another domain or send PM.
Thanks for the information, Joerg -- as usual, much appreciated.
Yeah... four-digits for us is in the range of, "OK, even if it doesn't work that well, for that sort of money we can just chalk it up to being a learning experience and not a real loss." For five+ digits it would definitely be more in the, "it had better work or heads will roll!" category. :-) My approach to the "it has to work" problem would be to find a really good consultant to "hold our hands..."
---Joel
But then it would never stay in the four digit range. In the same way it doesn't when you do it yourself because da boss will have to pay your salary, health care, numerous scopes, analyzers etc. :-)
If you want to take an existing analog design that works and migrate it to a chip a guy like Jim would probably be the right fit. There will usually be a redesign effort because some stuff isn't possible in chips. Such as resistors with a precision better than 25%, capacitors over a few hundred femtofarads, too much quiescent in a device and so on. OTOH you do get precision resistor ratios of fractions of a percent. When it's hardcore RF it can be the pits.
-- Regards, Joerg http://www.analogconsultants.com/ "gmail" domain blocked because of excessive spam. Use another domain or send PM.
A little more info on the FSA66 & FSA266. Fairchild engineering gave me this data.
From the characterization data charge injection across the -40 to 85C range is as follows:
1.65V Vcc: 0.004pC to 0.018pC 3.3V Vcc: 0.016pC to 0.028pC 5.5V Vcc: 0.036pC to 0.05pCI thought the charge injection specs were in error in the data sheet, but turned out the part was magic!
Seems Fairchild is turning out some pretty nifty parts these days.
-- Mark
Indeed. A few years ago I thought they didn't have much anymore in terms of engineering but they seem to have come around.
-- Regards, Joerg http://www.analogconsultants.com/ "gmail" domain blocked because of excessive spam. Use another domain or send PM.
Yes, I wouldn't expect it to at that point.
I've had mixed experiences with consultants... we have one right now who's very much worth every penny (and he charges us pretty much the exact same amount we bill our time out at, so it's certainly "reasonable"), but I've had a couple in the past where it was a toss up between whether they were actually helping or hurting a project. :-) It's been awhile since I've read it, but I like the bit on your web site about "how to find a good consultant."
Yep, that's the thought I've had at times... hence that thread awhile back about his ballpark estimates on what it could cost to re-do something like the Philips UAA2080, which he opined would be right around $100k... with a little luck.
---Joel
It ain't easy. The best avenue might be what one new client will now do with me. I don't entertain fixed bids unless specs are cast in concrete which is hardly ever possible. So I agreed that they toss me an initial small project, not to exceed xx hours. Along the lines of "Here is a nasty problem we've always had, let us know how we could tackle it". So their risk is limited. If they like what they receive then it'll become a longer term engagement.
If you really need an old style naked radio chip and there is no space for a discrete solution you guys might have to do it. Possibly you could sell them as well in order to offset some of the NRE. Or maybe Lansdale would buy the rights afterwards.
Maybe you could consider one of these, using a non-standard crystal:
Of course, tech support could be a roadblock right now.
Without transmitter but at rock bottom cost there is:
-- Regards, Joerg http://www.analogconsultants.com/ "gmail" domain blocked because of excessive spam. Use another domain or send PM.
Hi Joerg,
Yes, and while $100k is obviously quite a bit of money, there have been a couple of potential contracts where it wasn't necessarily completely out of the question.
Nice part, I hadn't seen that before.
"The tuning concept of the Self Tuned Radio (STR) is based on FUZZY LOGIC."
Good to see someone's still getting mileage out of that term. :-) Their tuning algorithm does appear sound, regardless of the name, at least.
For pager-type receives, I've been meaning to evaluate this guy:
- their idea of "low power," while nowhere near as good as the UAA2080/2082, is still not awful (
I hurrahed too soon. The Fairchild "models" are "protected" and can't be read by PSpice :-(
Anyone know of a crack??
...Jim Thompson
--
| James E.Thompson, P.E. | mens |
| Analog Innovations, Inc. | et |
| Analog/Mixed-Signal ASIC\'s and Discrete Systems | manus |
| Phoenix, Arizona Voice:(480)460-2350 | |
| E-mail Address at Website Fax:(480)460-2142 | Brass Rat |
| http://www.analog-innovations.com | 1962 |
America: Land of the Free, Because of the Brave
Actually one of the last chips I did in 1998 was about $100K. But that was only for low-level design, layout, mask gen and shuttle run. Not counting my time (I did the main schematic and sims).
Yes, but when you need it outside the FM band you might have to rely on their tech support and right now I personally would not count on that. WRT my designs here all I can say is that this sloppy support is already hurting them financially. They just don't know it (yet).
Nice, but kind of pricey for a receiver. Digikey wants over $3 and no stock which is a little bit of a red flag for me.
-- Regards, Joerg http://www.analogconsultants.com/ "gmail" domain blocked because of excessive spam. Use another domain or send PM.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.