Variable Resistors: Numeric Decade Resistors, 1-200 KOhms, 1 Ohm Resolution

[Jdo300]

Hello all,

I am working on a control circuit for a project that uses a 555 timer as a frequency controler. I need a variable resistor that has a range of 1-220 KOhms and I found some good 10-turn potentiometers that would seem to do the trick. But later, I was looking around on eBay and found these dial in numeric decade resistors (not sure if that is the right name). I attached a picture of what one looks like but could someone tell me what the official name of these things are? Also, does anyone know where I can buy some that have a range of at least 200 KOhms with a 1 Ohm resolution?

Thanks,
Jason O

 

[Jdo300]

I found a link to the manufacturer's website for the particular POT shown in the above pictures.

http://www.bourns.com/pdfs/3680.pdf

But most of the suppliers are charging at least $45 and up for just one. Does anyone know of any cheaper sources for these?

Thanks,
Jason O

 

[NoTime]

You're not going to be able to buy anything switchable that can do 1 ohm resolution at 200k ohms.
Or measure it for that mater.

You could see if you can find a decade resistance box.
They sell these as test equipment.

I think some are capable of doing a 1 ohm step, but that's not the same as resolution.

 

[waht]

Seems like you need a lot of precise control in your circuit. It is impractical to achieve presicion with analog circuitry. 10 turn pot is as high as you are going to get. It is rare to find, but there were 15 turn pots. Even if you had the switchable resistors box with 1 ohm resolution. Your circuit wouldn't change when you change 1 ohm dial becasue such low resistance would be absorbed in stray resistance of the wiring.

You are better off building a synthesizer for presicion. Otherwise a 10 turn pot will do the trick.

 

[Jdo300]

You're not going to be able to buy anything switchable that can do 1 ohm resolution at 200k ohms.
Or measure it for that mater.

You could see if you can find a decade resistance box.
They sell these as test equipment.

I think some are capable of doing a 1 ohm step, but that's not the same as resolution.

Hi, I thought resolution was the same as step. What does resolution mean?

I think some are capable of doing a 1 ohm step, but that's not the same as resolution.

What is a synthesizer and how do they work?

 

[Averagesupernova]

They've been discussed on this board. Do a search.

 

[es1]

It sounds like you just want to create a circuit that can output multiple frequency square waves. If this is true can I suggest an alternative: Have the 555 time circuit only output the highest frequency you need and then use counters type circuits to divide that output down to create your lower frequencies.

 

[es1]

If you're using the 555 timer to vary duty cycle then counters can still be used. Just set the 555 to output a clock at the resolution necessary for your duty cycle control then have the counters measure high and low time.

Personally, I find dealing with counters easier than with working with oscilator circuits. And the counter circuits tend to need less attention on your lab bench once they work.

 

[Jdo300]

Hmm that sounds nifty. That is exactly what I am doing. And I'm trying to find the simplest way to accurately set the frequency I want to use since I am working with a very wide range. In an ideal world, it would be nice if I could just dial in the exact frequency I want the circuit to run at. Is there a way to simply do that? or would it require immensly complex circuits to accomplish it?

 

[es1]

well the simplest way is to use a waveform generator from the lab. :) but I'll assume you don't have one.

there are a lot of ways to go really and the right way depends on what you have available and your what specific needs are (i.e. frequencies, do only rising edges matter, duty cycle, etc. etc.)

i'll assume you are working with reasonably slow clocks (low 100s of KHz) and want 50% duty cycle.

a pretty robust and simple way to go would to be to have your 555 timer circuit (or, even better, just an asic that outputs a fixed clock) clock a counter with a settable threshold. have the carry clock a toggle circuit. this way the 555 clock could be divided by 2*threshold giving you an extra bit for the divide.

 

[es1]

actually, now that I think about it just check the part catalogs (like digikey or jameco). I bet this problem is common enough that some one has an asic that solves it in one swoop for you.

 

[Jdo300]

Yeah it would be great if they have one. You are right about my criteria, I only need a simple square wave pulse at 50% duty cycle, nothing fancy there. But I just need a simple and effective way to vary the large frequency range accurately enough. It seems like the 555 timer would work just fine if I could find a simple enough method to vary the resistance easily. But if a widget does exist that would allow me to do that, does anyone know what it is? I wouldn't even know where to start looking.

 

[es1]

Integrated solutions abound. do a google for "clock divide by n"

However this part should do the job, along with your 555, if you only need to divide by powers of 2 that is. I suspect this is the case since for most control apps the exact clock frequency is not important as long as it lands in the correct decade. (and if precise frequency is important to your app how can you varying it over such a large range)

anyway,
http://www.datasheetcatalog.com/datasheets_pdf/7/4/H/C/74HCT393.shtml

Just cascade the two counters and feed the first your fastest clock and change the divide via one jumper to the corresponding output.

 

[Jdo300]

Hi thanks for the help. Though I'm still confused understanding how dividing the clock speed by 2 is going to give me the precise frequency I want. Could you explain this to me?

 

[es1]

Say you input a 100KHz constant frequency into this part, call this frequency f.

So the frequency at
Q1=f/2=50KHz
Q2=f/4=25KHz
Qn=f/(2^n)
and of course you can just use f directly.

Usually a control system uses a clock to measure time where the minimum time that can be measured is equal to the period of the clock. So as long as the period is sufficently smaller (i dunno, say ten times) than the time constant of whatever it is you want to control then everything will be ok. This is why the exact frequncy is not so important, just it's relative order of magnitude to whatever it is you are trying to control. (Of course the specific ratio is application dependant, for some applications the clock may need to be 2 or 3 orders of magnitude faster. The point is, there is a minimum period necessary but it's usually ok to use a slightly smaller one.)

 

[Jdo300]

So then the only frequencies I can produce are frequencies that are fractions of the original (F = 1/2^N)? Is there an alternative method if I want to be able to sweep the frequency through the whole range in discrete incraments rather than only fractions of the clock speed?

 

[es1]

They really are a lot of ways to go.

Just do a google for "clock divider 74hc" you should find something to fit your needs.

If you really want to do a continuous sweep over such a large frequency range than the 555 with an active circuit to generate the trigger is probably your best bet. I don't have time now to design one but I am sure there is a simple opamp circuit that will do it.

If you don't mind trimming in lab (each time) then maybe a bunch of pots in series and a smaller capacitance will do the trick.

And of course, see if you can just check out a function generator from one of the labs of your college. I bet you can...

 

[Jdo300]

They really are a lot of ways to go.

Just do a google for "clock divider 74hc" you should find something to fit your needs.

If you really want to do a continuous sweep over such a large frequency range than the 555 with an active circuit to generate the trigger is probably your best bet. I don't have time now to design one but I am sure there is a simple opamp circuit that will do it.

If you don't mind trimming in lab (each time) then maybe a bunch of pots in series and a smaller capacitance will do the trick.

And of course, see if you can just check out a function generator from one of the labs of your college. I bet you can...

Hi,

I do have access to a function generator that I have been using for my experiments. But I am trying to build my own independent control circuit to do some testing with. As for the idea of using POTS, so far that is the best thing I have come up with, but what does putting the capacitors in series with the POTs do?

Thanks,
Jason O