[BobM]

I just made a voltage divider to reduce the output of a device from 7V (professional equipment) to the more standard 2V (or thereabouts) for consumer equipment. Here's the basic schematic and a calculator, for those who are not familiar with a voltage divider network:

http://www.raltron.com/cust/tools/voltage_divider.asp

As you can see from trying different values for the calculator, many different combinations will work. I happened to have a set of very good Caddock TF020 10K resistors on hand that I used for R1, and some Holco metal films of about the right value for R2 to get close to 2V output.

But I could have used a 1K resistor for R1, and a correspondingly smaller R2 resistor and it still would have worked. Or I could have gone smaller yet. So here's my question, for you engineers out there...

What is the difference using a relatively smaller R1/R2 value resistor vs relatively larger value resistors? Is there a capacitance or inductance difference large to small, or does it just not matter at all? My guess is there is probably an optimal way of doing this, because as we all know, everything effects everything else.

Thanks,
Bob

[JoshK]

The difference quite simply has to due with how much current you are willing to burn across the resistors.  The two resistors making the voltage division are from voltage potential to ground.  V=I*R or I = V/R.  Since your V is 7V (to ground) and the two resistors added together are the total resisance, the I is then calculated from this.  As R gets larger, I gets smaller and visa versa.

[Steve Eddy]

What is the difference using a relatively smaller R1/R2 value resistor vs relatively larger value resistors? Is there a capacitance or inductance difference large to small, or does it just not matter at all? My guess is there is probably an optimal way of doing this, because as we all know, everything effects everything else.

What you're doing is trying to strike a balance between the increased output impedance of the device whose output you're trying to attenuate, and the load impedance that device will see.

Using a higher R1/R2 presents the source component with an easier load to drive, but at the expense of a significantly higher output impedance which can have implications for increased noise as well as high frequency rolloff due to cable capacitance.

Using a smaller R1/R2 doesn't increase the output impedance as much, but the source component will have a more difficult load to drive which can increase distortion.

You might want to try dropping R1 down to about 5k ohms or a bit less.

You could also use a step-down transformer which would give you the attenuation you need, but without increasing output impedance or reducing the load impedance seen by the source component.

se

[NewBuyer]

...You could also use a step-down transformer which would give you the attenuation you need, but without increasing output impedance or reducing the load impedance seen by the source component.

se

Would this perhaps be preferable, as the step-down transformer would present a lower output impedance (square of turns ratio, etc)?

[BobM]

Using a higher R1/R2 presents the source component with an easier load to drive, but at the expense of a significantly higher output impedance which can have implications for increased noise as well as high frequency rolloff due to cable capacitance.

Using a smaller R1/R2 doesn't increase the output impedance as much, but the source component will have a more difficult load to drive which can increase distortion.

Steve, thanks. This is exactly the info I was loooking for.

Since this device with the voltage divider is the output of a DAC, the signal will be going through interconnects from the DAC to the preamp. Increased impediance would therefore not be too big an issue, I believe. If the capacitance increased then I would think it would be more of an issue.

In any case, the new Caddock resistors are making it sound better than before, when I was only using Holco's. Wider and deeper soundstage - more holographic. I guess good parts are worth it (like I didn't know that already)!

Enjoy,
Bob

[art]

Yes, he could use a 4:1 step-down transformer.....pick up some CMRR, RF rejection.

And a big hole in his wallet. Not sure that is what he had in mind.

But, since someone else has already diverted this thread in that direction......

Typical 4:1 step-down is terminated in around 2.5 kilohms. You could go higher, but that is where they usually look best. In which case, you would present a 10 kilohm load to the source.

(Assumes step-down is at the load. Best bet.)

But back to pads................

It all depends where you put it. On the source end, too high of an output Z mucks up HF response. Too low.......it sucks current.

Again, if you put the pad (as you would with the transformer) inside, or at the load, you have a much higher range of resistance loading to play with. Without having to worry about HF loss.......noise pickup........current loading........

I kinda think he had this sort of answer in mind.........but I could be wrong.

[Steve Eddy]

Would this perhaps be preferable, as the step-down transformer would present a lower output impedance (square of turns ratio, etc)?

Correct.

se

[BobM]

Again, if you put the pad (as you would with the transformer) inside, or at the load, you have a much higher range of resistance loading to play with. Without having to worry about HF loss.......noise pickup........current loading........

I kinda think he had this sort of answer in mind.........but I could be wrong.

Yup, that's where it's located.

As for step down's - I tried that a while ago and got mixed results. True, the sound was a bit more fluid and "vinyl like", to use and OVER-used phrase. But the slam factor was gone and a wee bit of the transparency. I used Edcor coils for that, mostly because they were inexpensive. I do believe that using a better quality coil would result in better sound, but Jensens are expensive, for sure, and that's not what I was looking for.

The higher quality resistors I used for this experiment worked wonderfully.

Thanks for all your help guys,
Bob

[Steve Eddy]

I used Edcor coils for that, mostly because they were inexpensive. I do believe that using a better quality coil would result in better sound, but Jensens are expensive, for sure, and that's not what I was looking for.

If you ever are, and you're not already aware of them, CineMag offers the same level of quality as Jensen, but their prices are usually around 20% less.

The higher quality resistors I used for this experiment worked wonderfully.

Then there ya go.

Thanks for all your help guys,

No problem. Have a great weekend!

se