28-04-2015, 09:09 PM
The joy of electronics is that if you ask a dozen engineers to suggest a circuit, you'll get a dozen different answers 
Regarding the 555, I also use it in an educational setting. We have them on a special breadboard system made by Feedback, and it's a pretty good setup. The 555 timer is mounted inside a plastic box with pins that fit into the breadboard, and alongside the IC is a handful of decoupling and protection components that are intended to make it foolproof in the application (the same deal applies to the 741 as well). Despite that, we have had a few fail on us.
Until a couple of years back, we also had students assemble basic LED flashers as a soldering exercise. Both conventional and SMT. A fair few of the SMT ones got damaged by solder bridges and overheating, as you might reasonably expect, but they generally all worked by the end of the session. The main problem was the PP3 only lasted a short while because the bipolar 555 takes a lot of current. I proposed a 2-transistor series flasher, but it was felt that the educational benefits of soldering an SMT IC outweighed the battery life issue. Fair enough, I suppose...
Of course, they are used in the reset circuit of the BBC Micro, so that's good!
So why did I suggest a transistor alternative? Mostly just personal preference, I guess. Even though the IC is tiny and costs next to nothing, I do get some engineering satisfaction from doing with 2 transistors what the 555 does with 30 (that's a guess, BTW). Additionally, I have been caught out in the past - I have developed circuits using the bipolar versions with the expectation of fitting CMOS versions when they arrived from Farnell or wherever, only to find that the CMOS versions behave differently to the bipolar ones. I never did get to the bottom of that on the main project I remember, but I did stumble across the Veroboard prototype in the attic earlier, and am tempted to dig out the notes and see if I can work out what was going on. At the time - 20 years back - it was easier to ditch the 555 circuit and come up with another solution... I used a 4093 (quad Schmitt input NAND gate).
Also, the bipolar version can cause quite large rail glitches when it changes state, so you have to be careful about decoupling. OK, arguably that's just good practice, but it's variable from sample to sample - and manufacturer to manufacturer - so for volume production, you have to be very careful indeed. Far better engineers than me have been caught out by it
Obviously, I'm not suggesting that any of this is a real reason to not use the 555 here, but as I said earlier that I didn't fully trust them, I thought it was worth expanding on that.
Thinking about it, we have loads of the PCBs from the old exercise. I'll dig one out tomorrow to see if it lends itself to monostable operation. Failing that, those boards made by RS sometimes crop up. I have some, but unearthing them might be challenging!
Regarding MOS-FETs, they might work well here, and would be especially useful if battery life was a challenge - I had that thought earlier when musing the quiescent current issue. But personally, I'm not a fan of them. They are great in DC-DC converters and switched-mode power supplies, where the fast switching and low conduction losses are an obvious benefit. But for the bulk of what I do (audio engineering), they are markedly inferior to bipolar transistors, and as a result, I don't really have large stocks of them. But I know that I have thousands of general-purpose NPN transistors of the BC108 ilk, and I would imagine that many of us might have a similar balance in their stocks.
It's not clear to me without a diagram, but by way of a caution, if you swap the time cap and resistor around, there is a risk of ripple on the rail triggering the MOS-FET. Certainly, that would be a risk with a bipolar transistor - that's why I went for the 2-transistor solution. I guess it depends on the threshold voltage of the chose device.
Anyway, as I said earlier, it was just a suggestion - there are many ways to do it:
http://en.wikipedia.org/wiki/Monostable
http://www.daenotes.com/electronics/digi...tion-types
http://www.zen22142.zen.co.uk/ronj/drs.html
And I'm sure you can use a RaspPi if you wanted to give it an IP address with a web interface
Enjoy,
Mark
Regarding the 555, I also use it in an educational setting. We have them on a special breadboard system made by Feedback, and it's a pretty good setup. The 555 timer is mounted inside a plastic box with pins that fit into the breadboard, and alongside the IC is a handful of decoupling and protection components that are intended to make it foolproof in the application (the same deal applies to the 741 as well). Despite that, we have had a few fail on us.
Until a couple of years back, we also had students assemble basic LED flashers as a soldering exercise. Both conventional and SMT. A fair few of the SMT ones got damaged by solder bridges and overheating, as you might reasonably expect, but they generally all worked by the end of the session. The main problem was the PP3 only lasted a short while because the bipolar 555 takes a lot of current. I proposed a 2-transistor series flasher, but it was felt that the educational benefits of soldering an SMT IC outweighed the battery life issue. Fair enough, I suppose...
Of course, they are used in the reset circuit of the BBC Micro, so that's good!
So why did I suggest a transistor alternative? Mostly just personal preference, I guess. Even though the IC is tiny and costs next to nothing, I do get some engineering satisfaction from doing with 2 transistors what the 555 does with 30 (that's a guess, BTW). Additionally, I have been caught out in the past - I have developed circuits using the bipolar versions with the expectation of fitting CMOS versions when they arrived from Farnell or wherever, only to find that the CMOS versions behave differently to the bipolar ones. I never did get to the bottom of that on the main project I remember, but I did stumble across the Veroboard prototype in the attic earlier, and am tempted to dig out the notes and see if I can work out what was going on. At the time - 20 years back - it was easier to ditch the 555 circuit and come up with another solution... I used a 4093 (quad Schmitt input NAND gate).
Also, the bipolar version can cause quite large rail glitches when it changes state, so you have to be careful about decoupling. OK, arguably that's just good practice, but it's variable from sample to sample - and manufacturer to manufacturer - so for volume production, you have to be very careful indeed. Far better engineers than me have been caught out by it
Obviously, I'm not suggesting that any of this is a real reason to not use the 555 here, but as I said earlier that I didn't fully trust them, I thought it was worth expanding on that.
Thinking about it, we have loads of the PCBs from the old exercise. I'll dig one out tomorrow to see if it lends itself to monostable operation. Failing that, those boards made by RS sometimes crop up. I have some, but unearthing them might be challenging!
Regarding MOS-FETs, they might work well here, and would be especially useful if battery life was a challenge - I had that thought earlier when musing the quiescent current issue. But personally, I'm not a fan of them. They are great in DC-DC converters and switched-mode power supplies, where the fast switching and low conduction losses are an obvious benefit. But for the bulk of what I do (audio engineering), they are markedly inferior to bipolar transistors, and as a result, I don't really have large stocks of them. But I know that I have thousands of general-purpose NPN transistors of the BC108 ilk, and I would imagine that many of us might have a similar balance in their stocks.
It's not clear to me without a diagram, but by way of a caution, if you swap the time cap and resistor around, there is a risk of ripple on the rail triggering the MOS-FET. Certainly, that would be a risk with a bipolar transistor - that's why I went for the 2-transistor solution. I guess it depends on the threshold voltage of the chose device.
Anyway, as I said earlier, it was just a suggestion - there are many ways to do it:
http://en.wikipedia.org/wiki/Monostable
http://www.daenotes.com/electronics/digi...tion-types
http://www.zen22142.zen.co.uk/ronj/drs.html
And I'm sure you can use a RaspPi if you wanted to give it an IP address with a web interface
Enjoy,
Mark
