Astable 555 circuit with a period of up to 10mins

danadak

Joined Mar 10, 2018
4,057
JulesP, that schematic is a small % of the stuff that is onchip. Each item,
'like PWM is a component, a component being a resource all on one chip.
The board has a little more to allow for programming the device.

Note clock used is internal to get he accuracy I stated earlier. If you want
even more accuracy, you can use a external crystal attached to chip for this,
but I do not think you need it.

eetech00, You are correct I was not paying attention. There is no precision RC based
spec. But you can get an idea looking at prop deal vs T and V, looks pretty sloppy.
And add in Cap behavior with T and V I would go out on a limb and say it less
than precision, being a generous statement.

Regards, Dana.
 
Last edited:

eetech00

Joined Jun 8, 2013
3,859
eetech00, You are correct I was not paying attention. There is no precision RC based
spec. But you can get an idea looking at prop deal vs T and V, looks pretty sloppy.
And add in Cap b ahaviour with T and V I would go out on a limb and say it less
than precision, being a generous statement.

Regards, Dana.
Yes...I was gonna recommend a crystal based timing circuit for the CD4060B.

eT
 

danadak

Joined Mar 10, 2018
4,057
JulesP -

Is Clock_1 from the 555 setup to deliver 100Hz? And on the PWM what is 'tc' and where does the Reset come from? Which component is the CD4060?
No CD40460 needed, the PWM accomplishes that.

The clock, PWM,A/D all internal, on same chip, one chip.

tc is terminal count, when the PWM period rolls over and reloads
its count (automatically) to generate the next period cycle. This
output pulses for one input clock. Used in other applications for
timing control. This application does not need that connected to anything.

Reset has a logic 0 tied to it, so effectively disabling reset, as you do not
need that to be used. Note part has a lot of drag and drop basic logic elements
on it, like gates, flops, muxes, comparators, but in this design you did not need
any of that stuff so it goes unused on chip. Attached is a component list, a compo-
nent an onchip resource. This chip has most, but not all, of these onchip. This being
a low end part does not have some of the more sophisticated stuff like DSP engine,
20 bit DelSig, DMA......

If you look at right hand window in screen shot it is resources used/not used
on chip. As you can see many were not used. Its a summary, not everything
in part. For example you could add LCD display (external) and connect to part
and display settings.....messages......

Regards, Dana.
 

Attachments

Last edited:

Thread Starter

JulesP

Joined Dec 7, 2018
383
Thank you all for the contributions. What frequency does the 555 produce with 470n, 1k and 120k? 100Hz?

I will get some 4060s and have a play and let you know how it goes in due course.

Jules
 

AnalogKid

Joined Aug 1, 2013
10,987
There are no "standard" crystal frequencies that are exact binary multiples of 1/600 Hz. Thanks to the wonder of television, you are in luck.

If you are in the US, an NTSC color subcarrier frequency crystal is easy to find and cheap. 3.579545 MHz. Follow this with 31 stages of binary division and you get an almost perfect 600 second period with an error of less than 0.012%. A 31 stage divider can be done with three 4000-series CMOS chips, such as a 4060 as the oscillator and first 14 stages, followed by two more 4060's or other counters.

Or, a CD4521, 24-stage counter plus oscillator, followed by anything with 7 or more stages.

ak
 

MrChips

Joined Oct 2, 2009
30,712
There are no "standard" crystal frequencies that are exact binary multiples of 1/600 Hz. Thanks to the wonder of television, you are in luck.

If you are in the US, an NTSC color subcarrier frequency crystal is easy to find and cheap. 3.579545 MHz. Follow this with 31 stages of binary division and you get an almost perfect 600 second period with an error of less than 0.012%. A 31 stage divider can be done with three 4000-series CMOS chips, such as a 4060 as the oscillator and first 14 stages, followed by two more 4060's or other counters.

Or, a CD4521, 24-stage counter plus oscillator, followed by anything with 7 or more stages.

ak
Just for the fun of it, look up MM5369 which uses the 3.579545 MHz color burst crystal.
However, you can do just as well or even better with a 32768Hz watch crystal.
 

crutschow

Joined Mar 14, 2008
34,285
If you don't want to program a microprocessor, I would go with a CD4060 timing circuit as others have suggested.
Here's an example circuit.
For your purposes you not not need any of the circuit in red (the relay, transistor, and diodes D1, D2).
For 5 minute switching (10m cycle) you would take the output from Q7 (pin 6).
You would then adjust the oscillator frequency with R4 to give an LED blink on-time (or off-time) at Q4 (pin 7) of exactly 75 seconds (oscillator period of 4.6875s).
That will give you a 300s (5 minute) switching on-time at Q7 with an exact 50% duty cycle.
 

sghioto

Joined Dec 31, 2017
5,380
I think you could make it work with the SE555 using a modified # 3 circuit. Values shown are estimated and may need some tweeking. Timing cycle is appx. = 0.7CR.
SG
EEE 555 5 min delay.PNG
 

AnalogKid

Joined Aug 1, 2013
10,987
Just for the fun of it, look up MM5369 which uses the 3.579545 MHz color burst crystal.
However, you can do just as well or even better with a 32768Hz watch crystal.
I'm very familiar with the 5369, and still have a couple. However, its output is 60 Hz, not 1/600 Hz, so the output would have to be divided by 36,000, a non-binary increment (480 x 75) requiring decoding.

Same for the watch crystal plus a CD4060. You need more dividers plus decoding (16 x 75).

ak
 

crutschow

Joined Mar 14, 2008
34,285
But it would be a real pita to adjust to a new frequency.
It certainly would.
It likely will take a lot of tweaking and waiting for the 5 minute timeouts until the time is within the desired 3%.

That's why the CD4060 circuit is preferred. You just have to wait 75s to calibrate the time (or 4.6875s if you have a digital oscilloscope).
 

danadak

Joined Mar 10, 2018
4,057
Here is a solution where only one instruction code line has to be typed, an
instruction to start the PWM -

TCPWM1_Start();

You have to type that into main.c file, using the tool editor.

The solution gives you a fixed 600 sec output.

FYI for future work.

So again this is one chip does it all, outputs exact duty cycle of 50%,
and frequency tolerance of +/- 2% over temp. No external crystal, and
just a cap to bypass power supply. No calibration. If you use $4 board
no external components. Bypass cap already on board.

upload_2018-12-9_7-36-49.png

Note advanced users can actually design a 4060 internal to this part,
and use it in solution. But in this case why reinvent the wheel, just
use standard resource (TCPWM) inside part.


Regards, Dana.
 
Last edited:

Thread Starter

JulesP

Joined Dec 7, 2018
383
Given the board space I have already set out for the battery swapper, I will go for the 4060 but there are several options being suggested. Using the 4060 (crutschow) on its own and being fed from a 555 timer chip (Alec_t). Again because I already have a 555 in place I will use that but all your generous contributions show that, in electronics at least, there is more than one way to solve a problem. :)

When I have built this I will post again to check if I have it right to trigger an automotive type relay that can switch 30A. I don't know yet if the 12V high will be ok on its own or if I will need some attenuation or amplification.
 
Last edited:

crutschow

Joined Mar 14, 2008
34,285
If you want to control a relay from the CD4060 output then you will need a transistor buffer, as shown in the circuit I referenced.
To keep it oscillating, you just need to leave out diode D1.
 

AlbertHall

Joined Jun 4, 2014
12,345
There are a few problems with that cicrcuit.
First, automotive relays tend to need a lot of coil current and that may be more than the 2N2222 can supply.
Can you measure the resistance of your relay coil?
 

Thread Starter

JulesP

Joined Dec 7, 2018
383
There are a few problems with that cicrcuit.
First, automotive relays tend to need a lot of coil current and that may be more than the 2N2222 can supply.
Can you measure the resistance of your relay coil?
Yes I will but it’s coming from overseas and hasn’t arrived yet. Will post detail when it has arrived.

Thanks
 
Top