We have to agree to disagree for the moment. I wasn't replying to Davey.
But, since it's posted, his noise performance (shown in that image above) appears better than anyone else's measurements of the miniDSP by order of magnitudes. I am curious of the particulars as to why his approaches -130dB (I'm skeptical), while others are closer to -73 to -83dB in reported similar tests?
But, since it's posted, his noise performance (shown in that image above) appears better than anyone else's measurements of the miniDSP by order of magnitudes. I am curious of the particulars as to why his approaches -130dB (I'm skeptical), while others are closer to -73 to -83dB in reported similar tests?
You need to compute the total rms, the little red text at the lower left.
Think about it a second take Davey's setup and without changing anything substitute 103dB speakers for 85dB ones, it gets too loud. Attenuating at the amp preserves the original SNR. Attenuating the digital words keeps the noise and attenuates the music. Listening to CD's you are already at -93db or so at best using the whole full scale of the DAC.
I had a noise complaint from a friend using miniDSP HD 10 x 10. He was using Carver PM175 amplifier with low efficiency drivers. Carver amplified has input attenuators, and no setting with this resolved the issue; tweeter noise was unacceptably high. He shipped me the miniDSP to see what I could find. I had no problems with my various amplifiers.
Davey's wide-band noise measurement in post #14 is quite telling. miniDSP uses single pole low pass filter, and passes significant wide band noise. Carver PM175 has poor low pass filtering as well.
I shipped the miniDSP back to my friend, and had him purchase a Behringer A500 amplifier, and the noise problem went away.
Apparently miniDSP wide band noise modulates with power amplifier, placing audible noise into output.
Davey's suggestion of placing small capacitor acrosss outputs of miniDSP is step in right direction, but in many cases a two pole low pass filter may be required dependent on low pass filtering/bandwidth limiting of power amplifiers used with miniDSP products.
Yes, it would be nice for miniDSP to own up and revise output filters.
Davey's wide-band noise measurement in post #14 is quite telling. miniDSP uses single pole low pass filter, and passes significant wide band noise. Carver PM175 has poor low pass filtering as well.
I shipped the miniDSP back to my friend, and had him purchase a Behringer A500 amplifier, and the noise problem went away.
Apparently miniDSP wide band noise modulates with power amplifier, placing audible noise into output.
Davey's suggestion of placing small capacitor acrosss outputs of miniDSP is step in right direction, but in many cases a two pole low pass filter may be required dependent on low pass filtering/bandwidth limiting of power amplifiers used with miniDSP products.
Yes, it would be nice for miniDSP to own up and revise output filters.
Kouiky,
The "others" are either not measuring correctly and/or misinterpreting the measurements/numbers. I suggest to perform some measurements yourself.
I've switched to my E-MU Tracker-Pre (lower noise than my other soundcard.)
Here's another noise floor test with 96khz SR. Same graphing scale and setup as the one in post #41.
I believe Barleywater is correct and that, in some cases, out of band noise modulates with certain power amplifiers to create noise within the audio band. This probably accounts for the differences of opinion regarding audible noise of the miniDSP 2x4 unit.
If switching power amplifiers is not an option, another might be to couple the outputs via Jensen Iso-Max transformers.
Dave.
The "others" are either not measuring correctly and/or misinterpreting the measurements/numbers. I suggest to perform some measurements yourself.
I've switched to my E-MU Tracker-Pre (lower noise than my other soundcard.)
Here's another noise floor test with 96khz SR. Same graphing scale and setup as the one in post #41.
I believe Barleywater is correct and that, in some cases, out of band noise modulates with certain power amplifiers to create noise within the audio band. This probably accounts for the differences of opinion regarding audible noise of the miniDSP 2x4 unit.
If switching power amplifiers is not an option, another might be to couple the outputs via Jensen Iso-Max transformers.
Dave.
Attachments
Last edited:
This is a different issue, I would think some RFI/EMI protection would be present in the amp. I would think it would be hard to get a sense of broadband white noise from this but who knows. Anyway this was not the described problem, -86dB device directly into 30dB PA = -56dB SNR. The gain structure issues are real and there are numerous resources.
It is very easy to find any number of folks with 100dB or so speakers for whom the simple attenuation works.
Last edited:
Thank you Scott. Now that I know where to look for that, Davey's value is within range of everyone else's measurements.
I did, actually.scott wurcer said:
I had written the same thing regarding gain sequencing and speaker efficiency prior to condensing post #40, part of which remains; Attenuation should be after the DSP's outputs and at the power amp's inputs. This assumes a scenario where there's no way to reduce the amp's inherent gain. I'm not sure why you mention shortening word length though, I made no reference to such digital domain attenuation.
There is quite a leap from miniDSP's -78/-83dB to a hypothetical Compact Disc player's -93dB noise floor. A CDP is always attenuated prior to the amplifier through a preamplifier or passive attenuator. It seems that the DSP should be treated similarly as part of the gain structuring sequence, thereby attenuating noise along with the audio signals, ie: post #40.scott wurcer said:
I know that was just my way of saying that's what you do when you attenuate before the DSP, it runs with only some of its bits not the best idea.
So, in my case I should connect my O2+ODAC combo between my computer and MiniDSP and do the DAC there instead of in my HK amp, run full volume there and then connect from MiniDSP to normal input of HK amp. Will give that a try, I think I have enough cables for such a setup.
Edit: I did indeed have enough cables for it and it was a very quick setup. Just removed my headphones from the O2+ODAC, plugged in a cheap 3.5mm->2xRCA straight to the MiniDSP. Connected the MiniDSP to a normal RCA input on the HK. Set the volume to the second last step on the HK before it mutes totally. Slowly turned the volume up on the O2. I'm not sure though, but I think I hear some distorsion if I trn the volume on the O2 all the way up but it's hard to tell, can only play on low volume at the moment because I don't want to disturb my flatmates who are studying. Is there any danger to the O2 by running it on full volume?
Last edited:
That's probably not going to work. By using a normal input on your HK amp and reducing its volume control way down, you're likely forcing the signal level into clipping internal to the miniDSP.
This is a significant alteration to the gain structure of your setup.
Optimizing that scheme would require some experimentation and testing to determine the voltage gain at various settings of the HK volume control.
Dave.
This is a significant alteration to the gain structure of your setup.
Optimizing that scheme would require some experimentation and testing to determine the voltage gain at various settings of the HK volume control.
Dave.
Ah, you are right. I have RevA (which is max 0.9vrms input IIRC) and the ODAC puts out 2.0vrms. So if I then run the O2 on full output it gives 2.0vrms to the MiniDSP causing it to clip? Shall I change the jumpers on my MiniDSP or should I leave the volume on the O2 at like 50%? I guess changing the jumpers is probably a better solution?
Also, I don't need to be able to put out crazy high volumes from my HK, at least not in my current setup.
With the miniDSP connected to a normal input on your HK amp is the noise situation improved? Try the volume control at various positions.....even maximum.
You're using just two output channels from your miniDSP?
Yes, you could reconfigure the input jumpers and try connecting your source directly to the miniDSP input and then using your HK volume control as your system volume control.
If you don't have any huge +db equalizations programmed within the miniDSP that might solve your issue.
Dave.
You're using just two output channels from your miniDSP?
Yes, you could reconfigure the input jumpers and try connecting your source directly to the miniDSP input and then using your HK volume control as your system volume control.
If you don't have any huge +db equalizations programmed within the miniDSP that might solve your issue.
Dave.
Yes, my noise situation is greatly improved. At the moment I am only using two of the output channels. Connecting my computer straight to the MiniDSP is not an option, I want a high quality DAC and not the sound card on my motherboard. I don't think I have any big +db equalizations in my current setup but I will have an extra look at it this weekend.
Well, for "source" I meant your ODAC.....not your computer analog outputs.
The ODAC uses an ES9023 that has integrated 2.0 volt outputs, so it would be a good match with the input of your miniDSP unit with the jumpers in the 2.0 volt position.
So, clipping the output of the miniDSP ADC is not possible....but you still might generate clipping in the DSP portion with programming above 0db.
The fact that you're using just two channels of the miniDSP unit opens up a quite different usage option. If you see what I mean.
Dave.
The ODAC uses an ES9023 that has integrated 2.0 volt outputs, so it would be a good match with the input of your miniDSP unit with the jumpers in the 2.0 volt position.
So, clipping the output of the miniDSP ADC is not possible....but you still might generate clipping in the DSP portion with programming above 0db.
The fact that you're using just two channels of the miniDSP unit opens up a quite different usage option. If you see what I mean.
Dave.
Last edited:
Alright, thank you. Yes, I could get another ODAC to use in my setup and that is something that I will consider, so I don't have to switch the cables on my O2+ODAC combo whenever I want to switch between headphones and speakers.
But my final question is, is there any danger in turning up the volume on the O2 to maximum? Can it damage the amplifier in any way by forcing it to work harder than it is designed for? Or shall I not worry since I just have 1x gain on the O2 amplifier?
As long as you have your O2 configured for 0db gain then,yes, you can turn the control to maximum and just leave it there. Use the volume control on your HK amp as your main volume control.
Dave.
Dave.
For those using the MiniDSP as a 2 x 4 (stereo 2-way crossover), the best way to obtain good S/N and full volume control is to put a 4-gang pot between the MiniDSP and your power amp; setting your amplifier's gain to clip at the maximum MiniDSP output (0.9v rms). This 4-gang 10K audio taper pot is under $3: Bourns PTD904-1020K-A103; available from Mouser.
That will work, but you have to be careful input signal levels don't overload the ADC and also programming internal signal levels (EQ boosts) within the miniDSP don't generate clipping.
You can also then have tracking issues with a 4-section potentiometer.
Dave.
You can also then have tracking issues with a 4-section potentiometer.
Dave.
The tracking issue can be mitigated by using a multi-gang linear pot instead (Bourns PTD904-1015K-B103) and slightly modifying it (credit Rod Elliott). Unless you pay serious money, the standard "log" pot you buy from electronics shops is not log at all, but is comprised of two linear sections, each with a different resistance gradient, resulting in a pronounced 'discontinuity' as the control is rotated. Instead, take a linear pot, and connect a 1% tolerance resistor (10 to 15% of the pot’s value) across its wiper and end terminals. This will achieve a far more logarithmic curve than a standard log pot—a good approximation to true log pot operation is obtained over at least a 25 dB range. Treat all sections of a multi-gang pot the same way.
Last edited:
What are you guys using to power the miniDSP? It has a large input range. USB power was very noisy for me so I'm using a 12V adapter. I used a battery once and the noise was very low but that was not a convenient option.
- Status
- Not open for further replies.