resistors in series/parallel questions...


Background:
I'm trying to build some loading plugs for my phono preamp.

The pre unloaded is 100k ohms. My target value is 47k. After speaking ot the designer of the preamp, my understanding is that the correct resistor value to load the pre is calculated as follows:

R2 = 1/((1/Rt)-(1/R1))

Rt = desired load resistance
R1 = 100,000
R2 = resistor for loading plug

To get 47k, I need the new resistor to be a value of 88.7k, 89k, somewhere in there...

OK, now I need to determine what resistors to use in series or parallel to get the desired value. What issues do I need to consider in terms of power handling (designer said 1/4w is sufficient for this purpose)? It seems that it would be easier to get my desired value by going in series (with an 82k and an 6.8k, yielding 88.8k which is nigh perfect), but is there a better way of doing this via parallel resistors of common values or is series the easy way to go. If I go in series is it best to minimize the number of resistors in series by using the highest and lowest values that will give the desired value, or is it better to use similar values or does it matter? Assume I could use two 44.4k resistors. Would this be a better choice vs the 82k and 6.8k I mentioned? It seems like that since resistors are rated in terms of tollerance, that in theory, I should go with a bunch of very small values, since the overall error would be much smaller and possibly cancel each other out for some actual high and low values relative to stated resistance. Obviously the issues with soldering dozens of resistors in series negates this from a practical (and I'm assuming an electrical) standpoint. I think that I would gain power handling relative to each resistors rated power if I go in parallel, so is it correct that in series, I'm goin gto need to use higher wattage resistors?

If you're still reading, I'm sure it's obvious that I'm pretty clueless here, but I want to learn. I posted this only after spending some quality time with Google and learning the basics of parallel v. series calculations. Ohms law is now taped next to my monitor...

Thanks in advance for any primers, pointers, or help you can provide.

tw
twylie
The correct value should be 88.68k but any standard value resistor within 5% will work. You won't hear any difference. There is no need to solder multiple resistors together. Cartridge generates very little current, 1/4w is more than enough.
You can use whatever combination you want to acheive the value you want. The fewer the better, because the soldered joints can be problematic for sound. You don't have to get exactly 47k, because it is not that critical. 1/4 watt is fine for this. High quality resistors with 1% tolerance will sound better, and also cost more. If you want to get as precise as possible, you'll have to measure each resistor to get it's exact value, and then calculate. But, at 47k, a couple hundred ohms one way or the other is not going to amount to a hill of beans.
The above advice is good. However, it is very unusual to have a phono preamp with an input impedance of 100K. I suppose this is correct since you have spoken to the designer. Since you have access to the designer of the preamp, why don't you talk to him about just replacing the 100K resistors with 47K, which is a standard value.

If you have the skills to build some plugs, replacing a few resistors should be no problem.
All - thanks for your responses. Very helpful.

Twl - I'm having difficulty finding 1% resistors locally, but I'll probably test out and build some el cheapo rat shack ones as my first test since I'll likely have to order the higer quality ones. This will give me some practice and a reference for how much better the high quality ones sound.

Sidssp - thanks for confirming that I should be fine w/ 1/4w ones.

Herman - the phono stage is a Gold Aero DB-45 I picked up for a reasonable price from Hong Kong. It had the original manual and test sheet, but no loading plugs. The manual stated 100k internal loading (later confirmed). In an effort to understand how the loading plugs needed to be built I ended up trading email with Peter Russel, the designer of the piece. He is now VP of Operations for Penta Labs. I took a chance and emailed Penta as the maker of Gold Aero tubes and asked for a little help. Within a day Peter had emailed me the calculation and some more info on the pre. Nice guy and great service.

I thought about replacing the internal resistors and will eventually do this, but until I settle on a cartridge (still using the one shipped with my recent table purchase), I like the option of being able to pop phono plugs in and out of the back for testing purposes. Once I settle on a cartridge and required/desired value, I'll open it up and go after the internal resistor (and bypass the phono plugs while I'm in there...).

Zaikesman - yep, MI - Grado Sonata that shipped with my used table. I'm tweaking and saving my pennies to get into a Shelter 501 at Twl and many others recommendation.

Again, thanks to all of you for your help here. I'm really digging being back into my vinyl collection and appreciate the guidance of the board.
N0 Herman, but you're not the first one to ask me if that's an alias... You don't really thing Twl would have had to ask this quesiton do you?... Twylie = Trey Wylie
don't seat the close-tolerance resistance-value issue,
as much as being concerned about the quality of the resistors themselves...
In Other Words:
get some quality low noise noninductive parts from Michael Percy Audio. 1/8 watt or even 1/10 watt will be fine.
http://www.bainbridge.net/percyaudio/
Bob, I've been using the suggested 1/4 watt value for replacement loading resistors in my Camelot Tech Lancelot phonostage's sockets (they are Vishay VHS model, as recommended by the head tech guy over at Conrad-Johnson, who are located near me and permitted me to buy the parts over the counter). Will there potentially be an increase in transparency in using the lowest watt value possible for this application?
no that doesn't matter. As long as the resistor can dissipate the required input power (in this case it is negligible - signal only in the microwatt region) then you're just fine. CJ's recommendation of Vishay is excellent.
BTW I meant to say "don't SWEAT the value tolerance" even within 5% here is plenty close enough.