Pdxwayne
Major Contributor
- Joined
- Sep 15, 2020
- Messages
- 3,219
- Likes
- 1,172
Interesting, you are located in Malaysia?
-
WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!
Yet another A/B test of DACs
- Thread starter raif71
- Start date
- Joined
- Feb 23, 2016
- Messages
- 20,759
- Likes
- 37,603
The latency, and propagation delay differences between DACs is why I didn't mention this method. Usually not an issue, but every once in a while some DACs due to something they are doing will be tardy putting out the signal. So the two signals are out of phase enough to be an issue.
Ok I'm gonna be the one that admits my confusion here.
I get the first bit - plug the left analog-outputs from each DAC into the amp. One into the left input of the amp, one into the right.
I don't follow the rest - are you saying to take the positive speaker output from the amp (say the left side) and connect it to (say) the positive terminal of one speaker. Then take the positive speaker output from the other side of the amp and connect it to the negative terminal of the same speaker?
(and presumably both the negative speaker outputs from the amp are completely disconnected?)
That's not a comparator. This is a comparator!(apologies to Croc Dundee)
Hi John, I designed and played with the AB box few years ago. It was possible to match level of 2 analog paths, send it to 2 amplifiers and to switch amplifier outputs to 1 pair of speakers. Speaking about DACs, it is impossible to tell the difference if one of them is not really horrible or flawed. With power amplifiers, there is a bigger chance, however it is very difficult and the listener is to be trained.
Re null method, yes it was popular but there are caveats. The differences in phase response (as a result of small differences in amplitude response) or time delays like latency make the nulling process unreliable or even impossible. And dangerous if a beginner will play with power amp outputs. Current DACs have so same frequency response in audio band that a simple test at say 400Hz and a multimeter will give a good result. Or a pink noise, if available. The null method would be just confusing.
restorer-john
Grand Contributor
Exactly.
A perfectly balanced "bridge" is formed where each positive swings up and down and the net swing over the speaker is zero (no sound).
It's the quickest way to balance channel levels or L-R balance with a common gound system. It also shows relative balance on a volume control through its range.
restorer-john
Grand Contributor
It's a good thing for people to understand polarity, cancellation and absolute differences without any test gear. Sure, non-linearities (freq/phase etc) will manifest, but I like it a lot. It highlights channel differences faster than just abut anything else too. And as the difference signal is so small, there's little chance of hurting a thing.
restorer-john
Grand Contributor
solderdude
Grand Contributor
Method won't work with an early (< 2004) E30 though as it was phase inverted ... unless he had upgraded the firmware and only used USB input.
... unless he had upgraded the firmware and only used USB input.
restorer-john
Grand Contributor
Absolutely true Pavel. I use several identical amplifiers with identical preamplifiers up front of the big Technics comparator (in the picture above). I can then match the speaker levels with preamp gain. Total of 12 stereo sources, 5 amplifiers and 10 pairs of speakers. The Yamaha comparator has individual speaker level controls for each speaker and that is implemented by switching the matching attenuator to the speaker pair into either the amplifier tape loop or between pre/power. Very clever and simple- one set of RCAs in and one set out for 10 pairs of speakers and a control signal from the comparator.
I'm pretty sure most of the comparators through the 80s and 90s used a similar setup, although it was done with micros and remote control. I like the big old skool units. They used to take a day or two to wire up back in my retail days.
Last edited:
Have you ever had an accident doing it? Neither have I.
Yes. When 2 power amp outputs connected by a speaker to find a null, HF oscillations created in one of them took the amp away together with the tweeter.
solderdude
Grand Contributor
yes
It's a non issue in reality but in the very first E30's the 'null trick won't work, the other methods will
The null test could be the most accurate IF the L-R channel matching of the amp (and volume control) are within 0.1dB.
You can check this by putting the exact same signal on L and R input. The speaker output result must be 'null'.
Of course you would have to start this test by only using the L or R channel and hear how loud the sound is. Then add the other channel and it should 'null'.
Last edited:
Again, in a dual-mono amplifier with 2 separate power supplies and no common power ground for L and R channel there may be issues with the speaker "null". And if I say "may", I mean I have experienced it. Please guys (@solderdude and @restorer-john ), please take into account you do not speak to people who are as experienced as you are. This is a big trouble of on-line discussions, a debate of experts and total beginners mixed in one stream that is seemingly equal. No. I have more than 20 years of experience with internet discussions and meetings with audio fans. The meetings were popular up to some 2010 and then started to cease. During the meetings, I was always surprised by lack of practical knowledge of most participants that were otherwise willing to join any technical online discussion. So please, stick with safe recommendations only. What is clear to you is not clear to vast majority of participants of the discussion.
solderdude
Grand Contributor
I do get your point. John just wanted to give OP (and readers) another relatively cheap method with stuff they probably already own.
Wouldn't recommend using good speakers for it or high powered amps.
It won't be a problem in most cases where folks would be using an ordinary stereo amp.
the biggest issue will be channel imbalance in practice so while one thinks they have it nulled to a T the reality might be the result is 'off'.
Chances one encounters the odd dual monoblock amps that actually oscillate under certain conditions (by lack of common ground) is much smaller than channel imbalance being below 1%.
By far the best method would be to measure DAC output using an ADC (line input) on a sound card and even then only use 1 channel of it.
Tests by Amir have shown there can be L-R differences in DACs and ADC's.
Using a multimeter is also a good option but you will need one with a 200mV or 2V AC measurement range. NOT the cheapies with 200V AC as the lowest AC volt setting.
When you really want to find out the differences between DACs level matching is important within 1% (0.1dB).
By ear or by SPL meter will certainly have greater errors than that.
Wouldn't recommend using good speakers for it or high powered amps.
It won't be a problem in most cases where folks would be using an ordinary stereo amp.
the biggest issue will be channel imbalance in practice so while one thinks they have it nulled to a T the reality might be the result is 'off'.
Chances one encounters the odd dual monoblock amps that actually oscillate under certain conditions (by lack of common ground) is much smaller than channel imbalance being below 1%.
By far the best method would be to measure DAC output using an ADC (line input) on a sound card and even then only use 1 channel of it.
Tests by Amir have shown there can be L-R differences in DACs and ADC's.
Using a multimeter is also a good option but you will need one with a 200mV or 2V AC measurement range. NOT the cheapies with 200V AC as the lowest AC volt setting.
When you really want to find out the differences between DACs level matching is important within 1% (0.1dB).
By ear or by SPL meter will certainly have greater errors than that.
restorer-john
Grand Contributor
Pavel makes a good point. We've been doing this for 30 or 40+ years and it's second nature for me to lift negatives, short together and tweak my level/balance/gain control for a null on a stereo amplifier with two speakers connected.
The OP was not looking to spend any money and I wanted to help. But in real terms, he should spend some money, buy a multimeter and learn how to make some basic measurements. He may still blow something up, but he's probably less likely to.
The OP was not looking to spend any money and I wanted to help. But in real terms, he should spend some money, buy a multimeter and learn how to make some basic measurements. He may still blow something up, but he's probably less likely to.
OP
raif71
Major Contributor
- Joined
- Sep 7, 2019
- Messages
- 2,345
- Likes
- 2,563
- Thread Starter
- #60
Eh...I have made plans to buy a multimeter, wait for it !!!Pavel makes a good point. We've been doing this for 30 or 40+ years and it's second nature for me to lift negatives, short together and tweak my level/balance/gain control for a null on a stereo amplifier with two speakers connected.
The OP was not looking to spend any money and I wanted to help. But in real terms, he should spend some money, buy a multimeter and learn how to make some basic measurements. He may still blow something up, but he's probably less likely to.
Edit: checkout posts 39 and 40
