I've just started using RightMark with a turtle Beach Santa Cruz sound card as a budget ( $0) distortion analyzer. I ran "baseline" loopback tests to so I would know what the soundcard's contributuin is. Along the way, I decided I really didn't want my new PC connected to a power amp's output regardless of using a 20:1 voltage divider or even an opamp for isolation - just to paranoid. So I hauled an old Dell (433MHz) out of the garage and put the Santa Cruz in that. I re-ran the baseline tests. In my new PC the Santa Cruz came in at .006% but in the old Dell it read .0038%. These are repeatable. Odd. Is it possible that the older slower Dell is just less noisy? Would that be due to the slower processor? That seems very unlikely in that even the slower machine is operating way beyond the audiuo range. Also in the older one the power suppy is actually closer to the soundcard.
Anyway I would have thought the loop back test wouldn't be affected much by the motherboard or the rest of the PC. This makes me wonder about how valid any data I generate might be. I'm puzzeled.
Any comments?
Anyway I would have thought the loop back test wouldn't be affected much by the motherboard or the rest of the PC. This makes me wonder about how valid any data I generate might be. I'm puzzeled.
Any comments?
too many possible variables
case ground shielding effectiveness from line up to GHz, ground loop - case ps ground impedance, internal conducted+radiated coupling, pci slot location/relative to ps or disk drive wire harness, random additional bus/disk/video activity during meas...
also you need to look at the spectrum not just a thd #, the calc is usually thd + noise - averaging can reduce noise floor
even with thd limitations measurement resolution can be effectively extended with IMD difference freq measurements - the difference products won't be washed out by source thd and can be shifted to quieter frequency ranges
adding a few outboard analog filters could easily add 40+ dB to properly designed measurements
- not as simple as pushing a button and getting a #, but naive use of even a AP System can give meaningless results
case ground shielding effectiveness from line up to GHz, ground loop - case ps ground impedance, internal conducted+radiated coupling, pci slot location/relative to ps or disk drive wire harness, random additional bus/disk/video activity during meas...
also you need to look at the spectrum not just a thd #, the calc is usually thd + noise - averaging can reduce noise floor
even with thd limitations measurement resolution can be effectively extended with IMD difference freq measurements - the difference products won't be washed out by source thd and can be shifted to quieter frequency ranges
adding a few outboard analog filters could easily add 40+ dB to properly designed measurements
- not as simple as pushing a button and getting a #, but naive use of even a AP System can give meaningless results
The older, slower Dell has much more thorough case shielding than the new, fast PC. I just realized this today when the cover was off. Luckily this is the one that I was intending to use for testing.
One thing I picked up on while comparing them is that the 3.5mm phono plugs and cables less than wonderful and that's the good ones. I used the same good cable when making the comparrison by the way. Some resultd in intermittent connections and some had really high cross-talk despite being "stereo".
I'm going to try to get as much out of the current setup as I can before consiering a more upscal card, but when I do, one criteria is to be one of the musician's cards with independent out/input for each channel.
Anyway, I will be trying to check out somer of the suggestion/comments that have been posted.
One thing I picked up on while comparing them is that the 3.5mm phono plugs and cables less than wonderful and that's the good ones. I used the same good cable when making the comparrison by the way. Some resultd in intermittent connections and some had really high cross-talk despite being "stereo".
I'm going to try to get as much out of the current setup as I can before consiering a more upscal card, but when I do, one criteria is to be one of the musician's cards with independent out/input for each channel.
Anyway, I will be trying to check out somer of the suggestion/comments that have been posted.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.