Real resistor values

Joined Mar 5, 2007
45
Hi,
I'm using a 555timer to drive my traffic light circuit.The task is to produce a circuit that sends a pulse every second, and then alter it so it sends a pulse every 10 seconds.
I know that I need to use the formula 1=1.45/(RA+2RB)*C. The values I used to find RB were: 1=1.45/(1k+2RB)*10micro. This resulted in rb= 72k. How ever this is all well and good when simiulating the circuit, but I understand I need to use prefered values as close to one pulse a second as I can.
How can I work out what prefered values I can stick too? Or is there a table showing the avaliable values to use?
Many thanks
Matt Seymour

beenthere

Joined Apr 20, 2004
15,808
Hi,

Actually, it's real handy having a parts catalog to refer to when trying to find actual resistor values. There are two tables, one for 5% resistors and the other for 1% resistors. You may be able to search these out on the web.

For really exact resistor values, you can use a potentiometer wired as a rariable resistor. Use a meter to set the resistance to the exact value needed, or live in the circuit to give the exact voltage required. Start with a 100K pot.

5% resistor values jump from 68K to 75K. The standard was set back when resistors were very poor - the actual value was only within 20% of the stated resistance. Then you could get 10% resistors, and finally the 5% tolerances we have now. With 20% values, you hand-selected for the resistor you needed.

The 1% decade may have 72.5K - I don't have a catalog to refer to.

Joined Mar 5, 2007
45
Ah I see. I dont have a cataloge to refer to either When you say 5% and 1% is this the tolerance your refering too? Which means for example if the stated value of the resistor was 1kohms then the actual value of the resistor could be anything between 5%less then 1k and 5% more then 1k right?
Many thanks so far
Matt

Papabravo

Joined Feb 24, 2006
12,301
Standard resistor values are actually calculated based on having the logarithms of the values evenly spread over each decade. Here is a handy online table and the explanation for how the table is caluculated. The original algorithm had nothing to do with a 20% tolerance. As you can see it applies universally regradless of the actual tolerance.

http://www.rfcafe.com/references/electrical/resistor_values.htm

Joined Mar 5, 2007
45
Hmm After a Bit of research, the closest I could get to produceing a pulse once every 10seconds , is to produce a pulse once every 10.58 seconds.
I managed this by using the following values: Ra= 1k, Rb=68k and a capacitor value of: 1micro.
What do I have to do to alter the circuit to produce nearer 10 seconds instead of 10.58 seconds?
Many thanks
Matt

Papabravo

Joined Feb 24, 2006
12,301
There are at least three approaches to your problem depending on the amount of expense and agony you are willing to endure.

Option 1: Use resistors with a tolerance tighter than 1%

Option 2: Use series or parallel combinations of resistors to get closer to the desired value

Option 3: Use a potentiometer to make fine adjustments on each board that you produce.

The thing about values and tolerance that you are obsessing on is that the value you get on an individual part can be anywhere within minus 1% to plus 1%. In series or parallel circuits these tolerances can add up quickly. In the absence of any information to the contrary from a manufacturer I would assume a uniform distribution of values withinn the tolerance range.