Resistor Ohm Meter

[TheMadMan]

Hi All

3:21am local - couldnt sleep so I was thinking about the problem.

How About This:

Using a 555 timer which has a nice formula for working out the pulse timing.
(t (time) = R (resistance in ohms) * C (Capacitance in Farads)
We can easily calculate what the resistance is if the capacitance is known and we can measure the frequency (or time) being emitted by the 555.

The circuit:
Some sort of PIC.
An array of capacitors that can be individually connected to the 555 by the PIC using transistors.
Of course, a bunch of 7 segment display units.

The program:

The PIC then starts by connecting the highest value cap and counts the pulses coming from the 555.

If it is below a certain frequency, the next cap is connected and the pulses re-counted.

If it is above a pre-determined frequency, reset and try again.

This goes on until the pulse count is within a pre-determined set of values.

At this time, we know what the cap value is, we know what the pulse count value is and we can therefore work out what the resistor value is.

Take the calculated resistor value and send it to the display.

Pause for a second (so that the display does not constantly flicker if it is a low value resistor and the resistance across the fingers is stuffing things around) and then start again.

Only thing we need now is the upper and lower limits of the ohm range we will want to measure and using this, the number of caps needed and the values.

If we use serial to parallel convertors, we should be able to cram in up to 16 caps using 2 pins and the display can be 9 digits long using another 2 pins. Add another pin for the 555 output makes 5 so a PIC12 should be able to do it quite comfortably and we still have another pin available if we need more caps.

I do realise that this approach may not give the ultimate accurate reading but at least you will be able to quickly tell the difference between a 470 and a 4K7. Apart from the fact that I have difficulty in making out what color the little stripes are, the table confuses the heck out of me.<img class="inlineimg" src="images/icons/icon5.gif" alt="Question" border="0" />

Using similar methology, it should not be too difficult to build a unit that can measure capacitor values as well. Mmmmm.....

Ok, concept done... maybe my brain will let me get some sleep now.

What do you people think?

Best

[mackrackit]

Have you read the PBP manual yet?

Look at RCTIME, POT, and the ADCIN commands.

[T.Jackson]

<img src="http://www.picbasic.co.uk/forum/attachment.php?attachmentid=2073&stc=1&d=119164380 4" align="left" border="1">
I have used this relatively simple, "poor man's" schema in a few past projects. Never tried to do all the maths with just the PIC though, instead I just spat out 10 BIT samples to a host computer and performed all the calculations there. Worked O.K.
<br/>

[T.Jackson]

I think an interesting OHM meter project would be something that goes slight beyond the traditional OHM meter as we know it. Something that's purpose specific, for say; quickly determining resistor values. Instead of showing a readout of 2178Ω for a 2K2 resistor, it would display 2K2 (which is ultimately what the user is looking for) A 2K2 resistor with 1% tolerance, on a traditional meter could measure in anywhere between; 2178Ω - 2222Ω. It takes our brain some slight extra time to compute that, that's a 2K2 resistor as apposed to the display directly showing 2K2. This would significantly speed things up if the user had intentions of measuring a considerable quantity of resistors.

Another feature might include averaging! The user samples 100 or so resistors and system spits out an average. So if you're going to do an OHM meter project - think different and it might all be worth while. Otherwise, that $5 meter on eBay looks like the go.

[Pic_User]

Hi Trent and Aubrey,

I like the idea of a PIC resistor sorting display.

I have unskilled (young) student help (paid), sort 5% resistors back into single value bins.

There are 169 standard values (1 Ohm to 10 Meg-Ohm 5%).

Either they have to sort by color code or use a digital meter.

Digital meter:
When they get a value, they have to calculate which standard value is within 5% (prone to errors!).

And, in the teaching lab.
I could even place one at each student (learning) lab station to avoid the same human error problem Aubrey (TheMadMan) mentioned for the auto-placing robot. It would help the students building projects.


A nice “go / no go” method would be great, too.
I have visions of 169 LEDs. One over each bin.
The student labor could “swipe” the resistor (from the pile to be sorted), an LED over a bin would light. Bingo, less error, more speed. The LED would stay lit till the next value is “swiped”.

Now all I have to do is teach them not to touch both leads on the higher resistance value.

I do not have the PIC skills or the programming skills to know how to begin either of these interesting projects.

Any Ideas?

I have occasion to sort several thousand resistors. Every few months.

-Adam-

[Melanie]

2K2 Example...

2K2 +/- 5% = 2090-2310R

oh... 2310R... is that a 2K4 (+/-5%) Resistor?

and... 2090R... is that a 2K0 (+/-5%) Resistor?

Your students have now sorted all your Resistors into their respective bins. Next lesson, a student dips his hand into the 2K0 Resistor bin and pulls one out marked Red-Red-Red. Now, is it a Resistor that's been put into the wrong bin, or a really poor 2K2 that's closer to 2K0? Does he then learn from his experience that Red-Red-Red means 2K0?

I don't think you could encompass the entire Resistor selection, but certainly the basic 82/84 (12 step) range (1.0, 1.2, 1.5, 1.8, 2.2, 2.7, 3.3, 3.9, 4.7, 5.6, 6.8, 8.2).

[T.Jackson]

I guess one possible solution would be to have roundup & round down buttons. Pressing these buttons would effectively remove the tolerance from the reading by either adding or subtracting 0.25 to 10% (user selectable) - of the resistor's value. From memory there's about 180 different values in the 5 band 1% range. This is the E? something range (totally forgotten) The more obvious solution would be to database the entire range of resistors. When a reading is taken, the system would basically look for a closest match and display the results. Lot of resistors to catalog though.

[TheMadMan]

2K2 Example...

Thats why placing the results on a color screen makes good sense.
If a value falls into the overlap of 2 (or more) values, show them all with the correct color code for each.
A look at the resistor and then the screen to pick up the "correct" value, should work well.

BTW. Have you had any experience interfacing a PIC to a color CRT monitor?

Best

[TheMadMan]

Hi Adam
Take it you are a teacher/lecturer or something similar.
The majority of the reply to your comments is covered in my previous post.

There are 169 standard values (1 Ohm to 10 Meg-Ohm 5%).

Didnt know that - you live and you learn!

The screen display idea may be just the ticket for the lab top units you need. Maybe you could get the students to build them as a project for point credits. Get the local cash register company to "donate" some screens from old cash register units. As a plus, the students would unknowingly "learn" the color code for each value without even knowing it.

But first there must be a circuit diagram and in my experience, that is normally where things fade away. Fact of life I guess.

I have visions of 169 LEDs. One over each bin.

About 21 serial to parallel converters would give you a pin for each of the 165 bins so we are probably talking about a 40 pin unit with a fair amount of memory for the LED Selection array/s.

A nice “go / no go” method would be great, too.

Would probably need a keypad to set the "base" value to be tested against.
An easyer way would be a unit where you put a (example) 1K 1% resistor between 2 crocodile clips as a master and when you touch the 5% resistor onto the 2 test contacts, it works out whether the test piece is within 5% of the master. If it is, flash green led, if not, flash red led.
This will probably be far easyer to design and build.

I do not have the PIC skills or the programming skills to know how to begin either of these interesting projects.

I probably hav'nt either - YET! But a small detail like that should not stop us trying.

I have occasion to sort several thousand resistors. Every few months.

LUCKY!!!
Every time I run out of something, it means waiting for Saturday and then making a 150km round trip to re-fill the drawers. Very frustrating sometimes.
Sometimes I "substitute" but that cost me 4 opto-isolators and 8 leds this week. Bad News.

Best,

[TheMadMan]

Instead of showing a readout of 2178Ω for a 2K2 resistor, it would display 2K2 (which is ultimately what the user is looking for) A 2K2 resistor with 1% tolerance, on a traditional meter could measure in anywhere between; 2178Ω - 2222Ω. It takes our brain some slight extra time to compute that, that's a 2K2 resistor as apposed to the display directly showing 2K2. This would significantly speed things up if the user had intentions of measuring a considerable quantity of resistors.

Another feature might include averaging! The user samples 100 or so resistors and system spits out an average. So if you're going to do an OHM meter project - think different and it might all be worth while. Otherwise, that $5 meter on eBay looks like the go.

Nice out the box idea!!! Like the "2K2" readout.

Here's an even more "out the box" idea:
Somewhere I've got an article saved where an interface to a normal PC screen is shown.
So how about using one of the small screens (like those on cash registers) as the output device.
If it is a color screen, you could also display what the band colors are.
The human brain will probably start to associate the patern (as opposed to the individual bands) with the value and very soon will associate the patern with the value.
Unfortunately, if you then ask them "what are the band colors", they probably wont be able to tell you but if you put a resistor in front of them, they will recognise the color patern and tell you "2K2" or whatever.
This is why people can read but some are terrible at spelling. They recognise the patern of letters in the word and know the meaning but dont ask them what individual letters make up the word.
Best,

[T.Jackson]

Nice out the box idea!!! Like the "2K2" readout.

Here's an even more "out the box" idea:
Somewhere I've got an article saved where an interface to a normal PC screen is shown.
So how about using one of the small screens (like those on cash registers) as the output device.
If it is a color screen, you could also display what the band colors are.
The human brain will probably start to associate the patern (as opposed to the individual bands) with the value and very soon will associate the patern with the value.
Unfortunately, if you then ask them "what are the band colors", they probably wont be able to tell you but if you put a resistor in front of them, they will recognise the color patern and tell you "2K2" or whatever.
This is why people can read but some are terrible at spelling. They recognise the patern of letters in the word and know the meaning but dont ask them what individual letters make up the word.
Best,

Well that's an interesting, yet rather scary thought.

[TheMadMan]

I have used this relatively simple, "poor man's" schema in a few past projects. Never tried to do all the maths with just the PIC though, instead I just spat out 10 BIT samples to a host computer and performed all the calculations there. Worked O.K.
<br/>

This method should be able to work quite well but possibly will not be too accurate if the test piece DUT is too big or too small.
As far as I can remember, the ADC converts the voltage received (between 0v and 5v) to a number between 0 and 255. You will have to adjust the value of R1 to get a reading that is accurate for high value resistors and another R1 value for low value resistors and probably a number of times in between.
This method will probably work very well in cases where you are trying to get a number of resistors as close as possible to identical, closer than 1%.
Must add this to my library.
All the best,

[T.Jackson]

This method should be able to work quite well but possibly will not be too accurate if the test piece DUT is too big or too small.
As far as I can remember, the ADC converts the voltage received (between 0v and 5v) to a number between 0 and 255. You will have to adjust the value of R1 to get a reading that is accurate for high value resistors and another R1 value for low value resistors and probably a number of times in between.
This method will probably work very well in cases where you are trying to get a number of resistors as close as possible to identical, closer than 1%.
Must add this to my library.
All the best,

That schema is relatively orthodox stuff that I was formally taught at TAFE.

[TheMadMan]

Hi All

Have you read the PBP manual yet?
Look at RCTIME, POT, and the ADCIN commands.

Although I have written major apps using GWBasic back in the DOS days, I'm still a total newcomer to PICBasic. The PBP manual is required in-bed reading at this time (Wife bitches like a drain!!!!) and I'm still busy absorbing the info.
My gut feeling after reading up on the 3 commands mentioned is that the range that will be able to be handled is not wide enough to cover from 100ohm to 10K (in my opinion the range things fall into for "normal" projects without having "extra components" to the PIC) measurement range limits.
After all, the PIC range is a general solution that has tremendous versatility but is not able to be all things to all men or women in Melanie's case.
Would someone who has the experience and knowledge at thier fingertips care to comment on whether a PIC's RCTIME, POT or ADCIN commands would be able to do the job in the 100ohm to 10K range?
Best