Resistors in Parallel


Hi,

I'm working on a vintage tube power amp in which I've been changing resistor types in the signal path, revoicing it with excellent results.  I've tried just about all of the boutique resistors currently available. I've settled in on a combination of carbon composite Rikens,  Audionote non magnetic tantalums, and Shinkoh tants, all in select positions.  I haven't tried the Shinkohs in this amp but use them in other amps I've modded and like the sound!.  Recently, I purchased some Shinkoh tantalum resistors in values that will have to be paralleled to get to the target value needed.

My question is, will paralleling resistors have a degrading effect on the signal? 

I'm thinking about the way using a .01 uf bypass capacitor does when added to a coupling capacitor.  I'm in the camp that hears signal smearing when a bypass cap is used and prefer not to use them.  That said, in the case of Shinkoh resistor availability, my only choice is to parallel if i really want to use them in this amp and I do. Does anyone have experience in trying this and what were your considerations and conclusions?  

Thanks guys!

 

ballen1

It is very common to use resistors in parallel to increase the wattage.  For example, to get a 20W 5ohm resistor put 2 10 watt 10 ohm resistors in parallel (space them apart for heat dissipation).  

Generally, I don't see people considering this a compromise.  However, it is obvious that you create a loop that could generate a voltage from rogue emf.  

Maybe someone who actually designs high end systems will chime in but I would say that I would prefer to avoid multiple resistors in parallel but I do it when I need to.  I did it in the upgraded crossovers for my speakers replacing sand cast resistors with path.

Jerry

Wouldn’t one resistor letting more current through with less resistance create less heat (and noise)?

 

Parallel caps work (when they work) because caps in real life have series inductance and resistance. 

Non-inductive, low temperature coefficient resistors are pretty close to ideal resistors. 

Now if you had some high temp coefficient resistors.... then I could see maybe a compensating parallel resistor... but heavens why?  Just get a better resistor. 

Another thing is resistors are cheap and small vs. caps.  The expense of top of the line copper caps + their sheer physical size at large uF values makes a parallel cap save on money and space.  No such reasons for resistors.

PS - All other things being equal, higher wattage resistors have lower noise than low wattage resistors of the same type, but, again, top end metal film or non-inductive wire wound are just so low noise.

 

At least this is what I think I remember from when I wanted to buy them.

It's not a problem. I've paralleled resistors to dial in a specific ohm value that is not standard. If you really want to dive deep into the weeds there is the wire inductance in the paralleled resistor leads which can be a lot higher than the inductance of a single resistor. Like everything else in audio -- there is always a price for lunch.

OK, thanks guys.  I'll proceed with confidence...  I did notice that Audio Research parallels resistors in a number of its components, i.e. Reference 5SE Preamplifier.

Tarey

 

There is a reason why large competent engineering driven speaker companies do not add really small capacitors across larger ones. I mean Wilson, Magico, Revel, and companies like that.

It is normal to parallel resistors especially if manufacturing volume is low. You can buy a small number of values in larger quantities to create new values as needed. If you can, parallel resistors close to the same value. That reduces the worst case heat in any resistor. If you are using non inductive resistors, even the cheapest ones, inductance will never be an issue at the frequencies we are working at. Follow Erik's advice and get a low temperature coefficient resistor. You can always use a bigger one to prevent self heating. I would not stress too much, the voice coil resistance is probably going to change more than the resistor. Better not to have 2 things changing though.