No 1% resistors available

Working currently with active filters I am facing again the problem of getting 1% resistors (rarely available locally or with a 5 weeks delay - yes, what you read - to have them on the counter). Simulation is nice but I want to build things.

The gymnastics of assembling 3 or even 4 resistors (series/parallel combinations) to reach each value used to be fun but it is time consuming.

Many times I've read pejorative comments about this and, honestly, I tend to feel a little like that.

What to do? How much could I round values?

Suggestions welcome.

You have simulation. Use standard 5% values and see how much your filters changes.

remember to randomly assign values within the tolerance range when you simulate. that goes for the caps as well as the resistors.

... or use preset resistors

If you can't buy ordinary accurate resistors then do gardening instead of electronics.
I will buy some of your flowers if you buy some of my modern inexpensive electronic products.

Desagradable respuesta, Audioguru.

Quite in line with your style, indeed.

LTSpice is pretty sophisticated if you take the time to learn how to use it's more advanced features. You can program full equations for variables including randomness into almost any parameter anywhere in your schematic if you do it properly. Bellow is a simple example.


If you set that to say a resistors value, when LTSpice runs it will generate a random number between 0 and 1 using the number in the brackets (in this case 1) as a seed. Because the seed is static this will of course generate the same results every time you simulate. However if you use the .step parameter sweep function as well you can run a simulation multiple times with any scaled randomized variables in any range you want, you just have to tweak the math to use the 0-1 number that rand returns to get the variability you want..



Please take careful note of the use of the squiggly brackets, they determine how LTSpice will interpret things. In the above example the step parameter will run the simulation with a SEED value between 1 and 100. I had to add the +1 to avoid the possibility of a divide by zero error because zero is a possible return value of rand. I ran the above simulation on a simple transistor circuit with the variable component being a resistor and I got a different current value every stepped run.

The above example is pretty crude and you'll have to work out the math yourself to set 5-10% variables, but it should give you a better idea of what LTSpice is capable of. ANY variable for any component can be set through the math parser which has a full set of advanced math functions. Skies the limit!

Don't forget that LTSpice allows support for the input if .wav files as piece wise linear voltage sources and can also output files as well, so even without a single piece of hardware you can do advanced audio filter design and actually HEAR what the circuit would do. Depending on how much you want to get into the simulation. I've seen a few threads on the LTSpice Yahoo user group that actually use fully developed math simulations for speaker cabinet design which allow you to punch in the physic makeup of the speaker box and actually let you hear what it will sound like (assuming of course you're using studio grade monitor headphones)

An old trick with carbon comp resistors is to file a larger wattage resistor until the value is reached, then passivate with nail polish.

KeepItSimpleStupid said:

An old trick with carbon comp resistors is to file a larger wattage resistor until the value is reached, then passivate with nail polish.

Click to expand...

of course you had to start with a lower value resistor..... and carbon comps are noisy. one job i worked at was initial test of precision DACs. to trim the DAC steps as well as the overall gain, we used 5% resistors. if a value required was 1.01k, a 910 ohm resistor would be measured, and a second series resistor be selected to make up the difference. this is the process you've already described as being time consuming, but it does work, and it is used in industry quite regularly. National Semiconductor used a process where several resistors were populated in series with solder blobs in parallel with each one. the trimming process would remove solder blobs until the value was correct.

atferrari said:

Working currently with active filters I am facing again the problem of getting 1% resistors ... Suggestions welcome.

Click to expand...

Maybe the PCB houses like us should provide the purchasing service. If purchased resistor separately , the cost of shipping would be indeed too high. However, if shipping them with PCB, with many cases, there would be almost no increase in shipping cost.

From the garden - wrong phototropism - Series / parallel

unclejed613 said:

this is the process you've already described as being time consuming, but it does work, and it is used in industry quite regularly.

Click to expand...

Good to hear that!! Now that I think of it, I remember the case of a friend of mine, trained at a well known company in Switzerland where that process was common practice.

Last night I completed a 6-pole active filter with some values implemented playing the series/parallel game. It seems to work OK.

Sceadwian:

Thanks for the reply. I learnt a lot about LTSpice but that part of the simulations is still unexplored.

I am writing this from the garden where I enjoy my other passion, gardening, casually suggested as an alternative to the lack of resistors by the ETO resident Swami somewhere above in this thread.

This morning, after retrieving my old lousy garden rake from the tools shed, I spent some time trying to convince my potus, who has its phototropism wrong, to not invade my living room. Would turning the pot 180º do the trick?

Thanks for replying.