Is it worth replacing output resistors or just the caps???

[budx3385]

Lots of gurus say you can improve the output of any source by replacing old capacitors with Blackgate NXs or better (like Dynamicaps or Mundorfs).

What about the series resistors? Aren't they worth upgrading? The Rotel CD player I'm working on has a 100 ohm resistor on each output channel. Should I replace them with, say, a Caddock film resistor? or maybe just remove them?

What do you think ?

 

[slowpogo]

The effect a resistor has on the signal is proportional to its size: a 500k input impedance resistor, for example, can really affect the sound depending on what kind you use. I went from metal films to tantalums on my Pimeta input, and it was really worth it: the sound was still crisp but the metallic edge was removed, a pleasing warmth took its place, and the bass tightened nicely.

But a 100 ohm resistor, I believe, will have very little effect on the sound. Replacing it isn't worth the trouble.

I can attest to the benefits of changing output caps, however...I replaced the caps on my Echo Mia sound card with Black Gate NX, and it made a huge difference in the sound (in a good way). Soundstage went totally 3D, details jumped out of the woodwork everywhere, bass went from slightly anemic to deep and full (if a bit bloomy). And those caps do get better with burn-in.

 

[Clutz]

Quote:

Originally Posted by slowpogo /img/forum/go_quote.gif The effect a resistor has on the signal is proportional to its size: a 500k input impedance resistor, for example, can really affect the sound depending on what kind you use. I went from metal films to tantalums on my Pimeta input, and it was really worth it: the sound was still crisp but the metallic edge was removed, a pleasing warmth took its place, and the bass tightened nicely.

The effect a resistor has on a current in a circuit is proportional to its resistance. However, your logic that this means that it only makes sense to replace high resistance resistors with different types doesn't make any sense.

 

[majkel]

Quote:

Originally Posted by budx3385 /img/forum/go_quote.gif What do you think ?

You should try.

If they are carbon resistors, there might be an improvement when you replace them for metal film resistors. The latter have tolerances ususally not greater than 1% and higher power dissipation within the same size. Also "gurus" say you can lower the resistance at the CD output but I wouldn't do it without referring to the shematic. Sometimes it's not just a resistor connected to the op-amp buffer output but muting circuits added, and it's all matched together. I'm actually planing to upgrade my CD player - replace op-amps, some resistors and capacitors. Clock improvement planned later. Rectifying diodes are already Schottky, so no upgrade needed IMHO.

 

[budx3385]

I wandered through the Michael Percy catalog (percyaudio.com) and he sells several "boutique" resistors in various sizes, including 100 ohms. The "low noise" part seems enticing - when I changed the output coupling caps (installed BG NX 100uF in one and Dynamicap 5uF in another) the improvements in clarity were absolutely astounding. I wonder if that could happen changing the carbon resistors out. I mean, aren't they just a random bottle of dust, so to speak? Musical clarity - that seems like a worthwhile goal !

I want to hear from someone who has done this, please !!!

 

[slowpogo]

Quote:

Originally Posted by Clutz /img/forum/go_quote.gif The effect a resistor has on a current in a circuit is proportional to its resistance. However, your logic that this means that it only makes sense to replace high resistance resistors with different types doesn't make any sense.

Sorry, by "size" I meant "amount of resistance."

I'm not sure what you mean by the last part, though. From personal experience I've learned that, at least in a Pimeta, replacing a 1k resistor in the signal path yielded no audible difference that I could hear, while replacing a 500k or 1M resistor yielded very audible differences, even just between metal film types.

I know some of this has to do with the resistor's position in the topology. But I'm talking about comparing resistors within the feedback loop.

 

[Logistics]

Which order did you replace those resistors in? Cause it seems reasonable to consider that if you do the extremely noticable one first, the 500kOhm for instance, you may THEN notice a difference in the smaller 100Ohm when it comes time to do it.

 

[kipman725]

you shouldn't notice a differance at all. As long as you don't plan to short the outputs at any point you mayaswell remove the resistors.

 

[budx3385]

remove it? yeah, that's one way to swap it out (ha).

But I thought if there is a series cap and a shunt resistor plus inductor, that it's part of a high pass filter and load match for the output.

 

[12Bass]

Quote:

Originally Posted by slowpogo /img/forum/go_quote.gif I replaced the caps on my Echo Mia sound card with Black Gate NX, and it made a huge difference in the sound (in a good way). Soundstage went totally 3D, details jumped out of the woodwork everywhere, bass went from slightly anemic to deep and full (if a bit bloomy). And those caps do get better with burn-in.

You might also want to try different op amps after the D/A converter. I replaced the NJM2068s on my Gina24 with LM4562s and noticed a marked improvement in overall clarity and ability to hear into the mix. The LM4562s also lowered the measured THD in RMAA by about half. FWIW, I also removed the output electrolytics as my amplifier already has coupling capacitors at the input (the DC offset is rather small anyway...) - this also helped extend the bass down to 10 Hz or so.

 

[slowpogo]

Quote:

Originally Posted by 12Bass /img/forum/go_quote.gif You might also want to try different op amps after the D/A converter. I replaced the NJM2068s on my Gina24 with LM4562s and noticed a marked improvement in overall clarity and ability to hear into the mix.

I thought about doing this with my Echo Mia, which also uses 2068s. Except, the 2068s are a larger size than the typical SOIC opamp we would use in a headphone amp. I was worried that I'd have to go through a bunch of crap with the 4562s, extending their legs with resistor leads or something. Were you able to do it without a lot of trouble?

 

[majkel]

slowpogo, your example doesn't confirm only high resitance resistors matter to the sound. I had an issue when was buffering headphone out with 75 ohms resistors for some reason and couldn't find the source of bad sound. After all I realized I used another type of 75 ohm resistors, and after placing the same ones as previously, I got what I expected to hear. Actually, the role of the resistor matters, not the actual resistance.
Regarding op-amps, if there is an JFET input type of op-amp, I wouldn't start with bipolar op-amps on the I/V stage without analyzing the application. Bipolar op-amps have higher input current which might contribute to the converted current and lead to some conversion errors. It's better to check, however if there is an audible improvement it means the aim is reached.

 

[12Bass]

Quote:

Originally Posted by slowpogo /img/forum/go_quote.gif I thought about doing this with my Echo Mia, which also uses 2068s. Except, the 2068s are a larger size than the typical SOIC opamp we would use in a headphone amp. I was worried that I'd have to go through a bunch of crap with the 4562s, extending their legs with resistor leads or something. Were you able to do it without a lot of trouble?

While the LM4562s are slightly smaller, they should fit without problems. The legs should be plenty long enough to connect to the solder pads. Just be careful not to destroy the pads when you remove the old op amps....

FWIW, I've got a Mia as well and have been thinking about modifying it too (probably inputs only, for recording purposes). So far, I've done the three input op amps and one of the output op amps on my Gina24 (output channels 1&2). I left the NJM2068s in output channels 3-8 because I don't use them for critical listening. Quite a noticeable improvement in fidelity!

 

[slowpogo]

That's good to hear about the 4562s..the Mia has two inputs and two outputs, and four opamps. Would this be because they are balanced ins/outs? Two of the opamps are marked differently--still 2068s but with other, different numbers as well. I was wondering if this also has to do with being balanced, perhaps those opamps are out of phase or something?

Basically, I'm just wondering if I can replace both the output opamps with 4562s, even though the stock ones are not marked identically...

 

[Pars]

Quote:

Originally Posted by budx3385 /img/forum/go_quote.gif remove it? yeah, that's one way to swap it out (ha). But I thought if there is a series cap and a shunt resistor plus inductor, that it's part of a high pass filter and load match for the output.

Bud,

If its anything like typical Rotel output stages, the 100 ohm series resistors are there in combination with the muting transistors... most solid state outputs don't like being shorted to ground (mute). So they short the output to ground via 100 ohm resistor to keep the opamps happy.

So, if you aren't using muting transistors or relays, then you can jumper these. Otherwise, if you feel like spending $6+ per resistor, Caddocks such as the TF020 or the Vishay bulk-foil resistors could be worth a shot. You can also get what they call naked Vishays from Texas components. http://www.texascomponents.com/pdf/tx2352.pdf