my 74LS92 is not counting the way i want...

[TommyM]

Hi, as I'm new here I hope anyone can help me.

I have the following problem:
As you can see at

**broken link removed** (also attached)

I want to switch four relais (tried it first with LEDs). Forget the switch at the beginning, this doesn't work like this.
The rest works fine, but this 74LS92 just does not count the way I want.
I want to count from 0 to 3 and then reset to 0 and start again.

What it does:
*) starts at the first LED
*) after first trigger it jumps to second
*) after second trigger it jumps to four :?
*) after third trigger it jumps again to second
*) then it jumps again to four :evil:
*) then to two
*) to four - to two - to four - to two - to four - ...
*) sometimes it starts at the beginning, 1 - 2 - 4 - 2 - 4 - ...
*) sometimes it really jumps to three
*) most of all it jumps between two and four

WHY ???? :?:

It would be perfect to get some hints on this, as I don't know a way out of this.

Thanks
Thomas

also a strange thing: I want to trigger the 74LS92 by a NE555 (pin3 out) over pin 14... doesn't do anything...

 

[hyedenny]

Hi,
In order for the 74ls92 to count, one or both or the master resets (pins 2, 3) need to be held LOW - right now theyre "NC," or floating.
Leave pin 12 tied to pin 1, but use pin 14 JUST for your clock input from the 555 timer.
Also, decouple your power supply to the chip with a cap!!! (maybe thats what "c29" was meant for??)

When all else fails, refer to the DATASHEET!!!!!!

 

[jbeng]

On my datasheet, pins 2&3 have no connection, pins 6&7 are the reset pins. When either of the reset pins are low, the chip should count. It should reset to 0000 when both resets go high. C29 appears to be meant to help debounce the switch. If you're going to trigger it with the output from a 555, leave out the switch, resistor & capacitor. Don't forget to decouple your power supply like hyedenny says. Here's a link to a 74LS92 datasheet: https://www.futurlec.com/74LS/74LS92.shtml (Also, I never mix logic families if I don't have to, but I see why you did...)
JB

 

[hyedenny]

Whoops - pins 2 and 3 are resets on the ls90... Sorry, and thanks for the correction.

 

[neon]

you want a count of 0,1,2,3 and you got a 7492 to do that what is wrong with 7474 flip flop and an lm555 driving it?

 

[jbeng]

Hyedenny, I attributed your reply to someone else ... sorry!

Anyway, with regard to what neon says, TommyM could do that, but why not go one further and use a 4017? It has the counter and decimal decoder in one package, thus eliminating a chip altogether.
JB

 

[TommyM]

Hey Guys,

Thanks for the respond.
My comments to yours:
*) I did check the datasheets of course
*) I use a 74ls92 because I want to be open for extending this circuit for more than 4 states.
*) if pins are not connected internally I think I don't have to give them a defined state, am I right?
*) am I right with connecting the reset pins like I did?
*) as I said, forget the switch at the beginning :wink:
*) if I use pin3 of the 555 then of course there is no switch, R or C for pin14 of the 74ls92 (which were indeed meant for debouncing...)
*) I will put a cap there for decoupling the IC (100nF ?)
*) 4017? I will check this chip...

Thanks
Thomas

 

[hyedenny]

I wish you could attribute ALL my mistakes to someone else!!

Why not take it TWO steps further and do it all with a uC?! ...but thats cheating.

 

[TommyM]

because this seems to be a little over the top for this "little" issue...

This has to work, I would be the first where it didn't, I think.

 

[Roff]

Are you clocking from a switch, or a 555, or...?
If you are clocking from a switch, you'll need to debounce it.

EDIT:
I guess you sort of answered this already.
The Motorola datasheet that I looked at made no mention of maximum clock risetime, but I'll bet that it won't work by debouncing the way you have shown. If you do it this way, I think you'll need a Schmitt trigger between the RC and the clock input.

 

[ljcox]

You don't need the diodes (D3 ~ 6) across the relays since the ULN2003 has internal diodes to protect the transistors. Since you have connected the COM pin to +5V so the protection is active.

 

[ljcox]

Hey Guys,

Thanks for the respond.
My comments to yours:
*) I did check the datasheets of course
*) I use a 74ls92 because I want to be open for extending this circuit for more than 4 states.
*) if pins are not connected internally I think I don't have to give them a defined state, am I right? Correct, but for the sake of others:- Unused inputs (ie. pins that are internally connected to inputs) should be connected either to Vcc or Gnd. Sometimes it does not matter which (ie. Vcc or Gnd) but in other cases it does. Read the data sheet
*) am I right with connecting the reset pins like I did? Yes
*) as I said, forget the switch at the beginning :wink:
*) if I use pin3 of the 555 then of course there is no switch, R or C for pin14 of the 74ls92 (which were indeed meant for debouncing...)
*) I will put a cap there for decoupling the IC (100nF ?) Yes, decoupling is essential, otherwise strange results may occur
*) 4017? I will check this chip...This is a CMOS device with decoded outputs. You can make it count in modulo 4 by connecting output 4 to the reset. It will replace IC9 & 10. There may be a 74LS version also.

Thanks
Thomas

 

[TommyM]

Thanks again,

I thought already about this diodes, and also came to the result, that i will not need them. I will leave them out.

I looked at the 4017, but didn't get how this will work for me...
Maybe with ljcox hint I will make it. Or does someone have a more detailed description? :roll:
Or anybode knows a 74ls which does the same?
And are 100nF enough for decoupling (or too much)?

 

[ljcox]

Thanks again,

I thought already about this diodes, and also came to the result, that i will not need them. I will leave them out.

I looked at the 4017, but didn't get how this will work for me...
Maybe with ljcox hint I will make it. Or does someone have a more detailed description? :roll:
Or anybode knows a 74ls which does the same?
And are 100nF enough for decoupling (or too much)?

Have you looked at the data sheet for the 4017? It is a decade counter with decoded outputs. So it is equivalent to your IC9 & IC10 combined.

If you do a search in this forum for 4017, you will see examples of its use.

100 nF is the usual decoupling capacitor. You need at leat one for every two ICs.

 

[TommyM]

Great, I found something and will try it like that.
If I face problems I will come back with them :wink:

Thanks so far
Thomas