Need help with 555 timer circuit development

On Tuesday, 16 December 2014 15:56:00 UTC+11, Toyin  wrote:

 use a 555 timer circuit to generate a variable 1 to 5 minute timer to alte
rnate between 2 bank of batteries that are connected to a charger and a loa
d.  There are four switches where two are on when the other two are off. Pl
ease send me a private email to <olutoyin at yahoo dot com> if you are inte
rested and can help.  Thanks!

I'll provide advice for free. The 555 isn't really suited for producing 1 m
inute to 5 minute time intervals. The CMOS versions have high enough input  
impedance to make the idea practicable, but keeping a printed circuit board
 clean enough to exploit this isn't all that practical.

This really is a job for single chip microprocessor clocked by a 32,768 Hz  
watch crystal. It will give you more accurate time intervals, and will be l
ess tricky to get working.

Bill Sloman wrote:

   Beg to differ; one 555 running at (say) 1KC can drive a second one  
set like a 30:1 divider which in turn can drive a third set like a  
divider, etc; can get down to days with good accuracy..
   But the micro gives excellent precision and accuracy with no RC tweaking.

On 16/12/2014 08:46, Robert Baer wrote:


That sounds like making life unnecessarily difficult just to use 555's  
for the job. A CMOS 555 followed by a divide by 2^N chip should do the  
job admirably (although I think a PIC might be easier/cheaper).

A CMOS 4060 might well do it in a single chip oscillator & divider.

http://electronicsclub.info/cmos.htm#4060

(for the benefit of the OP)

On Tuesday, 16 December 2014 20:46:16 UTC+11, Robert Baer  wrote:

 to use a 555 timer circuit to generate a variable 1 to 5 minute timer to a
lternate between 2 bank of batteries that are connected to a charger and a  
load.  There are four switches where two are on when the other two are off.
 Please send me a private email to<olutoyin at yahoo dot com>  if you are i
nterested and can help.  Thanks!

 1 minute to 5 minute time intervals. The CMOS versions have high enough in
put impedance to make the idea practicable, but keeping a printed circuit b
oard clean enough to exploit this isn't all that practical.

 Hz watch crystal. It will give you more accurate time intervals, and will  
be less tricky to get working.

ng.

I think you've just gone mad. Iterating 555's just gives a linearly increas
ing delay, not a multiplication.

On 12/16/2014 6:01 AM, Bill Sloman wrote:


He's talking about injection locking the slower ones.  The problem is  
that the slowest one still has all the problems of a free-running 555 at  
that speed.

If the OP can't design his own 555 circuit, a 4060 is probably good  
medicine.

Cheers

Phil Hobbs

Phil Hobbs wrote:

   Absolutely!
   I just stuck with the 555 to make a point that they are useable, and  
with the right division ratio (no more than 30:1) and stable parts, one  
can get many hours for timing.

Bill Sloman wrote:

   NOBODY said anything about multiplying. Dividing is an inverse function.

Or a CD4060, this timer has a 14 bit divider attached.  
it's used in toasters etc 5 minutes is no problem.

snipped-for-privacy@yahoo.com says...


 I guess a timer driving a Flip flop would do..

 The Q and /Q can drive some trannies to switch the coils in.
  
 Not sure if the timer can spand that far however, you can  
switch in/out between flips flops to correct that problem.

Jamie

On Monday, December 15, 2014 11:56:00 PM UTC-5, Toyin wrote:

 use a 555 timer circuit to generate a variable 1 to 5 minute timer to alte
rnate between 2 bank of batteries that are connected to a charger and a loa
d.  There are four switches where two are on when the other two are off. Pl
ease send me a private email to <olutoyin at yahoo dot com> if you are inte
rested and can help.  Thanks!

I just want to thank everyone for your invaluable input. Since this is way  
over my head, I still need help and would contact some of the folks who had
 replied and shared their contact information.  Thanks again, I do apprecia
te the suggestions.

-Toyin

On 12/16/2014 7:57 PM, Toyin wrote:

Since nobody else acknowledged the elephant in the room, I'll take a stab.
Your objective isn't clearly stated...
If you're charging one battery while loading the other,
What happens at the switchover?
Break before make glitches the power off.
Make before break risks breaking something with high
current between the batteries.

It's likely that you'll have issues well beyond timing.
The microprocessor solution gives you the opportunity
to address them.  And having A/D converters available lets
you monitor everything and possibly decide that time is
not the optimum switching criterion.  Also lets you
log performance and shove data out a LED that that
you can read with a phone/PDA.

You're likely to be disappointed with any rational
approach to a 5-minute analog delay.

I'd back up a level in the system design and approach
the whole range of issues.

I'm curious to know exactly what problem the switchover
solves???

On 17/12/2014 07:19, mike wrote:

Just to add some numbers to that 555 v uP decision, LMC555CMN from  
Farnell costs 0.29 each on a reel of 5k. You can buy an 8 pin Cortex M0  
for 0.33 (3k price break)(and there will be cheaper processors if you  
look around). The extras to make the 555 give reliable 5 minute delays  


Michael Kellett

On 12/17/2014 09:40 AM, MK wrote:




Sure, but since the OP isn't comfortable building a 555 circuit, setting  
up a development system, learning C or PIC asm, and getting it working  
is probably more than a 4d job. ;)

Cheers

Phil Hobbs

On 12/17/2014 10:07 AM, Phil Hobbs wrote:



Given the lack of info on the objectives and quantities,
it's difficult to speculate.

My point, based on 40 years of watching very smart engineers
bang their heads against the wall trying to use devices unsuited for the  
job,
is that reexamining the system architecture is likely to be more
fruitful than trying to design around all the subtle gotchas
required to get a reliable 5-minutes out of a 555.

I'm all for using tools and devices matched to the designer's
skills and experience.  I suggest that it's far easier to
learn a very valuable programming skill that, in this case,
requires little more than copying an example program,
than to learn all about current leakage on a circuit board
under reasonable environmental stresses.

If you don't have the skill for either task, might as well
learn the skill for the one likely to work.

On 12/17/2014 6:12 PM, mike wrote:




BITD when you had to use wet electros for the job, I agree.  I suspect  
that a polymer aluminum might do fine with a CMOS 555, inelegant as that is.

1meg * 330 uF makes a TC of 5.5 minutes.  Connecting the cap from the  
trigger input to VDD makes the leakage shorten the delay rather than  
lengthening it (perhaps indefinitely).

Cheers

Phil Hobbs

On 12/17/2014 3:44 PM, Phil Hobbs wrote:




Well, assuming that half of 5.5 minutes is tolerable, this time,
with this humidity at this temperature.  And that's only one source
of leakage.

What's the cost  of  a 330 uF 5V polymer aluminum cap these
days?

A 555 is ill-suited even if you do have the skills
and budget to prevent catastrophic failure.

                      <Snipage>




Well, the ultimate Usenet solution would be a 555 clocking a PIC.

 mike wrote:


  


** Back in 1977, I made myself a darkroom timer for B&W printing from negat
ives in an enlarger. It was based on a NE555 monostable.  
  
A rotary switch selected resistors wired in 1:1.4 sequence, equating to sto
ps on a camera. Another switch multiplied time settings by 10. The range wa
s from 1 second to 450 seconds. 4x15uF tantalums provided the delay and the
 x10 times switch operated by pulling the voltage on pin 5 up and down for  
long and short delays respectively.  

The OP could use switched resistors with a maximum of say 5Mohms for time s
etting and a 20uF worth of film cap for delay. Then pull pin 5 towards the  
supply with a trim pot and resistor in series of about 1kohms each to get e
xactly 300 seconds -  which also calibrates the unit.  

5% accuracy and repeatability should be easily had.  


....  Phil  











  
  

On 12/17/2014 7:39 PM, Phil Allison wrote:

I'd agree that the 555 is one of the most versatile/useful devices in  
history.
Emphasis on "history".
In 1977, that solution may have been far better than the alternatives.
Add up the parts cost today and compare that to a microcontroller solution.

I built a long-duration timer out of a pair of unijunction transistors.
But I wouldn't recommend it today.

Based on what little we know about the application, I judge that
there are a lot more issues involved in sequencing the transfer
that will require additional timing functions.  Bite the bullet and
put in a microcontroller so you can fix stuff as you learn about those
problems.

I'd still recommend going back to the system architecture and taking
another look.