74LS90, 74LS92 and other TTL are OBSOLETE???

On 13 Mar 2006 09:35:21 -0800 in sci.electronics.components, "pdmtr" wrote,

Sure they are. Any two or more such chips are replaced with a microprocessor. Maybe one of them.

What logic is NOT obsolete, other than microprocessors, FPGAs, ASIC chipsets?

Yes of course it can be replaced by a uCU... or a CPLD, FPGA, ASICs, etc! The question is, WHY to replace a simple 1$ (+ cost to make the code) one to do the job??? if you just need to do only the ting the TTL logic does? Do they going nut (IC makers) or ... ?

NTE still carries replacements for the 74LS90 and 74LS92 as well as many other otherwise-obsolete parts. Their parts are a decent solution if you need to repair some older equipment, but are generally too expensive to consider using in new projects.

These sorts of small logic and small-scale-integration chips are very much an endangered species, and are widely listed by those manufacturers who make them as "not intended for new design".

The trends these days seem to be:

- Low-voltage operation. 5-volt logic is passe.

- Very-small-footprint packages.

- Large, complex logical networks are generally implemented via microcontrollers, PLDs of one sort or another, ASICs, and FPGAs. Building complex logic out of simple DIP or large-surface-mount logic ICs is only rarely done these days.

- A lot of single logic gate types are available in tiny little packages (5- or 6-pin SOT-like) and are intended for adding small amounts of logic "around the edges" of a more complex chipset. What it boils down to, I think, is that if you need to repair some older equipment or a project which used the old DIP/SO TTL parts, you can probably do so but may need to substitute, scrounge, etc.

If you want to build something new, you'll probably be better off designing it to use modern parts and techniques. You might find that a simple PLD, or even a PIC microcontroller, could absorb the functions for which dozens of TTL-logic parts were originally used. You'll end up with a simpler, easier-to-build circuit.

--
Dave Platt                                    AE6EO
Hosting the Jade Warrior home page: http://www.radagast.org/jade-warrior
  I do _not_ wish to receive unsolicited commercial email, and I will
     boycott any company which has the gall to send me such ads!

No, they're not nuts.

It's simply a matter of economics. Decades ago, most digital electronics were build around this sort of part - the manufacturing volumes were extremely high and they were churned out like jellybeans.

These days, most consumer electronics devices no longer use such parts... because it's possible to make TVs, VCRs, computers, etc. with more modern parts that are smaller, faster, use less power, and are easier to handle on modern high-speed manufacturing lines.

In short, the actual demand for small-scale-integration TTL parts in large packages has died. The total world demand for such parts is probably a small fraction of a percentage of what it was, decades ago.

At this point, there's just no longer an incentive for manufacturers to keep making and stocking them. It's more trouble and money than it's worth to them. I'd guess that the cost of maintaining the older wafer-fabrication lines used for these old designs becomes prohibitive.

For modern designs, nobody replaces a single $.10 IC with a $1 PIC or CPLD unless they absolutely have to.

Instead, they design a single board which uses $5 worth of micros and ASICs and FPGAs, but which does the same job (or a better job) than a rack of large boards which used hundreds or thousands of $.10 ICs.

The large-volume parts buyers no longer want 74LSxxx DIP parts.

--
Dave Platt                                    AE6EO
Hosting the Jade Warrior home page: http://www.radagast.org/jade-warrior
  I do _not_ wish to receive unsolicited commercial email, and I will
     boycott any company which has the gall to send me such ads!

Direct Components Inc. 4828 W. Gandy Blvd. Tampa, Fl 33611 Phone: 1-888-723-7279 or 813-835-3883 Fax: 813-831-0295 Email: snipped-for-privacy@directics.com

Has them in stock

Hi Dave That I'm saying is that you can keep the things simple enough and always use proven & reliable parts for something that it doesn't need all the "design power" a newer part provides, while that costs >10 times the former solution!!! For instance, say you have a TCXO based clock of 12MHz and you just want to get three subclock of it, a 6MHz, a

4MHz and also a 2.4MHz... A single 7490 will do the job on the fly for only 10cents maximum cost... and it could be also SO package SMT... not DIP! And I'm not saying on building something that will had a rack sized board full of TTL's on it... For that simple case I think the TTL logic solution is the best for value... I'm sure nobody (at least until today, as I start to think that I'll also see this in my life) replace a 2 transistor circuit with an ASIC for example ;) What is starting to worries me, is that we have lost the sense of "keep it simple enough but not simpler than it sould be" and we try to build everything based on the "but this goes up to eleven" way... (from a film...)

However I'm understanding exactly what you're saying... :(

Hello Sotiris,

Ask TI, they show the LS90 as active. But prices are already above $1, not a good sign. There are other counters though.

LS probably, HC won't go obsolete for a long time. In fact, I am still designing commercial products with CD4000 and 74HC series devices. They even migrated them to TSSOP which certainly would not have happened if obsolescence was on the horizon.

There are applications where a simple logic chip beats out a uC big time. Always will be. Logic chips are alive and kicking but they are migrating to lower voltage families.

Regards, Joerg

On 13 Mar 2006 11:08:29 -0800 in sci.electronics.components, "pdmtr" wrote,

But you don't need just one TTL IC. What are you building? What are you building that doesn't need a few dozen TTL ICs = or one micro to do the same job?

Of course there are plenty of 74LS still on shelves at

etc., but can you guess what your boss would say if you designed a few of them in to a product that was going to be made in 100k quantities? How big do you want your cell phone to be?

Well, the truth is that there is a need for ONLY 2 TTLs in the design!!! Those are just used to give 6 diferrent subclocks from a TCXO source. The design has also uCU, RF ASICs, and lot of other stuff like LCD, Keyboard, etc... It's sure not a "one chip" show :)) As I see it got to use a CPLD for replacement of the 2 TTL... the cheaper I have found until now is ~1.70$ so that's far more than ~20cents!!! I'm sure not only mine but nobodies "boss" wouldn't like to spend 150K more than it could :) Of course a 2U 19" rack mounted isn't so portable as a cell phone, so why we have to shrink everything we are doing in life even if there is plenty of room to use? It's not look weird for you to open a 29" TV set and find out that all it is behind that large screen is a chip with say 3 drivers a HT T/R all in a board that could fit easy at 4.5" monitor? Yes, I know thats the way it goes... And it's sure more economical for a production line of >1M units / year to build your "all in one" ASIC! But at least we must admitte that is a bit weird :)) to shrink everything, even when there isn't a real need to do so!!! I have done designs with FPGAs but there always a need to do it with. Either large complex circuitry, or even dynamic reconfiguration of the circuit needed and either space limits and total cost, the most cases was for portable devices. But this is not always the case...

Nice to hear your opinions on this subject..

PS Thanks for Jameco... .

Hi Joerg Checked TI... Yes the 1$/pc is the final trick!!! Raising the price for something that was 10cents/pc gives a good push away from it!!!

I'm not so sure now if the package migration meaning the continue existance of a chip :( But hope I'm wrong ;)

Yes, I'm sure that this last paragraph is true... Just hoping somebody out there (IC makers) realize it... Have an example of what simplicity is... Years ago have done a design where I use a single transistor to open a uCU bus instead of a 14DIP IC that competitors use to do it! At some cases basic electronic circuitry does the job EXCELLENT!!! But of course if we have forget what makes a gate and see it always as a 'block" then we will always use overkilled ICs to do something simple, and then we will try to shrink it so much that we will finally make it small enough as the starting basic simple circuit!!! Yes, of course we will have alot of "extra power" available for future, but that's not always usefull... cheers Sotiris

Oh oh, I better stock up on the 7490/7492 family as I like to build digital clocks with them like this nixie clock here.

Regards, Sal Bris>Hi Joerg

No,

The large manufacteers stop making them. But some smaller ones are still making them, and will keep making them as long as there is demand. And there are huge lots still available on the market.

Regards, Pieter email: snipped-for-privacy@hoeben.com without the NOSPAM of course

Hello Sotiris,

Well, they would not go through the whole ECO release procedure if the market for a chip was declining.

The smart ones do but not all IC manufacturers have smart management ;-)

It's like with PCs. People become lazy, want everything automated and now you need a PC with 256MB of RAM to do the same things that an old

8086 DOS version could perform in under 500kB. Or look at cars. My first one had 16 horses and worked just fine. Today you can't buy any under 50 horses.

Keep up your skills to design with discretes. Most young grads can't figure it out anymore, meaning that guys like us will have a pretty good job security. When I presented my first switch mode supply that did not have a PWM controller people stared at the white board. "How on earth does this thing work?"

Regards, Joerg

A couple of years back on sci.electronics.design, I asked about clever ideas for a one-out-of-ten LED sequencer built of discretes, and Spehro came up with this five-stage twisted ring counter that was self-decoding when driving LED's. I then laid out a circuit board using

0201 surface-mount discretes and the tiniest LED's I could find, and it turned out to all fit inside a 0.4" by 0.8" rectangle (roughly the outline of a DIP).

Tim.

Hi,

I just found this thread through a search. I'm looking for a 74LS92 for a divide-by-twelve stage in a one-off project. Does anyone know if a replacement for the 74LS92 is stocked by any of the major UK electronics suppliers like Farnell and RS? If so, what is the product number? So far I can only find old stock 74LS92s with small mail order suppliers.

Many thanks,

Chris

Both Cricklewood Electronics (020 8452 0161) and Dalbani (8393 7774) list these.

If you can't get sufficient quantity of LS chips, I believe that the HCT version will work equally well.

--

rgds
LAurence

...Enter any 12-digit prime number to continue.
---*TagZilla 0.059* http://tagzilla.mozdev.org