New 555 timer

[MikeK]

I just read about this.  It's a new micro-power programmable 555 timer chip, the CSS555.  Has anyone used one?

[Hydrawerk]

No. http://www.customsiliconsolutions.com/downloads/Revised%20Standard%20products/CSS555_App_Note1_Serial_Interface.pdf

[MikeK]

Wouldn't they have to be cheaper than the cheapest MCU to be worth using?

[w2aew]

Wouldn't they have to be cheaper than the cheapest MCU to be worth using?

Of course not.  There's a lot of value/merit for building circuits that don't have to be programmed to work.

[Mechatrommer]

Wouldn't they have to be cheaper than the cheapest MCU to be worth using?

Of course not.  There's a lot of value/merit for building circuits that don't have to be programmed to work.

i think the chip need to be programmed through serial interface, otherwise it will use whatever default value in EEPROM i guess. what i dont really get (quick reading did not give me anything) is what is the real advantage compared to the classic 555?

[smackaay]

I don't really know what the advantage is especially since it's so expensive (almost $2.00) vs 0.46 for a normal 555. It's a neat chip but I can only see it being useful for extra long delays. In those cases, why not just use a pic10f200 or something.

[AndyC_772]

A new 555? What next, a really low power kind of cassette player? A 6502 processor in a QFN package? Maybe a slightly better kind of fax machine?

[Mechatrommer]

if they want to create a new breed of 555, i want the "high speed version", up to few tens MHz osc and pwm freq. i wont mind if its $10.

[ftransform]

This is a game changer and I am going to buy a dozen.

Out with the old and in with the new.

[BravoV]

I see nothing overly special about this? It can do delays to 1 day, lower quiescent current then the CMOS version but after skimming through a 20 page application example I don’t see the point. I really can’t see anybody dropping over 2 bucks on it when you can buy a 25 cent micro that can do the same and then some.

Maybe for those "unexplored" audiofool market ?  Just can't imagine the buzzwords they'd use in using this "new" 555 !

The only distributor that seems to stock it is Jameco and they actually carry an evaluation kit for $75.00 WTF!

For those who are really desperate getting one, here what I would do ...

1. Get a Cypress Semiconductor PSOC mcu in 8 dip or so package for about $ 3.5 , the spec -> CY8C27143-24PXI Product page
2. Use this firmware made by Cypress in one of it's application note example -> Simulating a 555 Timer with PSoC

You will have a cool fully functional 555 IC that internally has :

- Vcc down to 3 volt
- 16 KB Flash
- 0.25 KB SRAM
- 8 bit cpu core
- 6 GPIO
- 8 configurable digital blocks

Features wise, how is this sound for a 555 chip ? 

[codeboy2k]

the benefit is the lower power. and it just works out of the box. They claim < 5uA active operational current. 

It would be interesting to compare it to a 25 cent micro as AcHmed99 says... although which one are you talking about?  the lowest price micro I know of is the PIC10F200 and it's 0.34 on a reel of 3000.  Maybe alibaba has it cheaper, but is it real?

They also claim .005%/C timing stability, so that would be also useful to compare to a low cost micro over temperature changes.
The 10F200 is about 0.1% stable over temperature.

Furthermore, the 10F200 uses about 300uA at 3.3V (175uA @2V) and 4Mhz. So you would have to do some slick programming to get it to sleep, and wakeup to change an output pin state, then sleep again, etc.. During the sleep, it's only drawing about 0.1uA,  if you disable all peripherals, but I think you would still need the timer or the WDT, which draws about 2-8uA over temperature. The timer needs the clock, so its not really sleeping, then it's back to 300uA (or 175uA). But if you can use the WDT then it's only using an average of 3uA. If I read the datasheet right, the smallest WDT timeout will be 15ms. (you can use the post-scaler to make it larger, but not smaller).So now you're limited to 15ms on and 15ms off, 50% duty cycle and that's 30ms or 33Hz ! not very good.. so I don't think you can use the WDT, in fact as a 555 timer here. cross that off the list.

You could try to get fancy with the comparator during sleep, and use an RC ramp outside the processor chip, just like the 555 does.
So you can sleep with comparator interrupts enabled, and wake up, run a few instructions at full power, flip an I/O port and go back to sleep.
Perhaps that might use less than 5uA on average, I don't know. That's what I mean by it's worth a study.  You might be successful, however,  the programming effort is much larger, and the complexity might introduce a subtle timing bug, or worse total failure, that breaks your system.

 as w2aew says: "There's a lot of value/merit for building circuits that don't have to be programmed to work"

So I think there is a market for this more expensive 555, if you need a timer running continuously on a battery device, that perhaps uses up less power than most micros alternating between sleep/wakeup mode, and you just want it to work out of the box. $2.00 is not so expensive anymore if that fits your needs.

If you just need a timer, then perhaps the F200 is ok, but if you need continuous running while perhaps the rest of your system is sleeping, and you need low power, or stability over temperature,  then perhaps the CS555 is going to work in that niche.

I would love to evaluate what I just wrote maybe they would like to send me samples

[bombledmonk]

Not in larger packages, but I remember seeing a presentation from Linear Tech back when they were launching their TimerBlox series.  They pitched them as resistor [only] settable replacements to 555 timers, but at the time, they were going for $7 a piece.  I thought they were nuts, but I later heard they sold fairly well.  Prices have gone down somewhat, but still rather crazy in my mind.

[Analogtech]

I don't see any real point to this other than the ability for long delays as mentioned before if they really wanted to make a new 555 timer they should have made a more stable one to start and higher frequencies would be real nice.the old 555 timer has been in use for decades and is still a damn good timer chip for basic purposes. They should make a new 555 which can actually output more current than 200ma a 1.5-2 amp output would be nice but is probably asking for too much and the beginners might be put off the whole hobby if they needed to program a chip to blink an LED

[codeboy2k]

You bring up some good points. About the closes PIC would be an XLP series which has a worst case Idd of 30uA over temp running at 500kHz 1.8V its one buck at Newark cut tape.
[... ]
I’ll revise my original statement. For a device that cost $2.34 cut tape it seems a little pricey. For less then half that you can get the PIC  XLP so it would offer additional functionally maybe at reduced performance in some aspects compared to the CS55 as to whether this is worth it to someone? Well I see it’s only available at Jameco.

Yes, the $1 XLP seems like a better deal if you factor in that you can also use some of it's other functions, and even get multiple timer waveforms from its I/O ports.

Say for example you use an XLP to create 10 timers in software, different waveforms on multiple output ports.  Now you've saved $2.34 x 10 plus 10 x 5uA over the CS555, that's $23.50 and 50uA, replaced with the XLP at $1 and 30uA, at the cost of some programming time.  Really, it comes back to "depends on the design need".

This thread makes me likely to consider the XLP if I ever needed more than a few timers on a board, and was not as concerned about battery life. If I was REALLY concerned then the CS555 might win. If I needed lots of timers, the XLP definitely wins.