I've found the same. Usually they match much better than 1% but you have a few outlying values that break the trend. I guess that's what you're paying for.
One thing to be aware of with SMD resistors in their voltage rating. I believe that as a general rule, large package sizes will have a higher rating. This can have an impact. Douglas Self mentions this in his book. He states "Using SMD resistors in the feedback network of a power amplifier has been shown to be a bad idea." Unfortunately, he doesn't explain his reasoning for this, so you have to go to the reference to research it. The reference points to post 121 in this thread:
Essentially, the 50V rating of the SMD device the poster was using caused excessive distortion. I think Douglas Self's advice to avoid SMD resistors in a overreaction. He should state to be aware of this specification.
Essentially, the 50V rating of the SMD device the poster was using caused excessive distortion. I think Douglas Self's advice to avoid SMD resistors in a overreaction. He should state to be aware of this specification.
I also specified RR series thin films for some custom stepped potentiometers. They’re great, 0.5%, 25ppm, virtually immeasurable distortion and noise low enough I can’t visually distinguish it from the analyzer’s residual. Normal pots are in the 500-1000ppm range, for starters, cermets are around 400-500ppm, and conductive plastic 150ppm.
Susumu also has the RG series.
If I was chasing the bottom in terms of noise while building a SOTA spectrum analyzer or passthrough current was a projected factor, other options would be appropriate in key areas. In audio equipment, I don’t worry about chasing the absolute bottom.
Probably best to use good thin metal film or else MELF. A problem with audio resistors which I have seen, IMHO is a function of their Excess Noise properties. Excess Noise mostly consists of current noise. It is noise above and beyond thermal noise, and it depends on how much current is flowing through the resistor. If resistor current is proportional to the audio signal, then signal-correlated noise is produced. Also IMHO, resistors designed especially for audio are compromised in some ways in order to optimize them for lower Excess Noise. IOW, there are always engineering tradeoffs to consider. In addition, audio resistors may not be optimal for audio test equipment design depending on whether Excess Noise is more important that other parameters for instrumentation purposes.
Last edited:
Best to get a soldering microscope. Prices for suitable long-working-distance, binocular/stereo, optical microscopes start around $200 - $300. It will change everything about soldering, handling small parts, inspection, etc.
https://www.amazon.com/AmScope-SE400-Z-Professional-Microscope-Magnification/dp/B005C75IVM/ref=sr_1_1_sspa?crid=256GZ4GUD9HZN&dib=eyJ2IjoiMSJ9.eX13iA3CMh1RolTks91eyLXrWrfTidcA7uHXA5ZoRO2J26vd0LCTo2uq3g3HMMPz9QeGZLvixUg0wxuane7KT6_1lDGcKTU7cjS5Dvqr2tJv2TZEpxLAzIcnPsoaSLFvrR-c5frkvIXGXfdINEOZahtQ0SaOXaqrmW-bGEzOLZHBuirbrPtr0w3olyvQdEBXa7oC0inr7puLUAXZJVWR4L6ugcB7ljeBkpGbjoBQi3k.P8iGfkC-X5WSRBVOpBAFqmDynTrICkpe3XIcLKYY_Lc&dib_tag=se&keywords=amscope+soldering+microscope&qid=1711219753&sprefix=amscope+soldering+,aps,155&sr=8-1-spons&sp_csd=d2lkZ2V0TmFtZT1zcF9hdGY&psc=1
People building switched-resistor FIRDACs have been able to hear the differences in resistors used in the output array. It turns out the sound of resistor excess noise may not be what someone expects. To me it blurs the sound and obscures some low level details. In some cases bass can be strongly affected if the excess noise is more strongly of the 1/f type.
Some resistors which I found to be among the worst I tried included Susumu RG. I expected them to be very good based on their reputation. They were among the most blurry. Also, the blurry effect is IME similar to some of difference between well implemented good off-the-shelf audio clocks as versus SOA clocks (the least blurry). Interestingly, signal correlated noise including clock phase noise, resistor excess noise, etc., can be viewed in a very high resolution FFT. A good post on the subject at: https://www.diyaudio.com/community/threads/phase-noise-in-ds-dacs.387862/post-7063038 The rest of the thread that follows may also be of some interest. Regarding clocks, IMHO blur is not the only difference between good and better clocks, but it is one effect similar to that of some other noise sources.
Some resistors which I found to be among the worst I tried included Susumu RG. I expected them to be very good based on their reputation. They were among the most blurry. Also, the blurry effect is IME similar to some of difference between well implemented good off-the-shelf audio clocks as versus SOA clocks (the least blurry). Interestingly, signal correlated noise including clock phase noise, resistor excess noise, etc., can be viewed in a very high resolution FFT. A good post on the subject at: https://www.diyaudio.com/community/threads/phase-noise-in-ds-dacs.387862/post-7063038 The rest of the thread that follows may also be of some interest. Regarding clocks, IMHO blur is not the only difference between good and better clocks, but it is one effect similar to that of some other noise sources.
Last edited:
I'm gonna try Susumu RR and RS. Yageo RT are good. IMHO, Texas Components SMD metal foil that we tried here might be slightly better than the best metal thin film, but not enough in our consensus to justify the price difference in almost all cases. Also, could be I tried Susumu RR before if that's what was in the FIRDAC output array of a MarcelvdG RTZ FIRDAC sent to me by Acko for some evaluation of the type that I like to do. Marcel's dac sounded fine from the output resistor perspective, at least we though so here (could be some other issues though).