Hi guys,
I'm looking to boost the signal out of my miniDSP so that I can run a lower gain on my TPA3116 amp. I have a 2.1 setup with the miniDSP as an active crossover driving 2 separate TPA3116 amp boards, one running in bridged mode for the sub. I only need to boost the input signal for the sub amplifier.
I have one of these chinese (likely not genuine TI) NE5532 preamp boards coming https://www.amazon.com/gp/product/B07HG37N1X/
I only need one of the two channels, but I know that OP-Amps are not always happy with one unterminated input/output.
The TI spec sheet says that it is OK to tie it to ground or either supply rail. I was thinking that I would do this directly at the chip on the board, bypassing the rest of the circuit for the second channel. (as opposed to just connecting the input or output of the board to ground). Considering that this is likely a fake and not a real TI chip, do you think it is still OK to do this?
If not, what do you recommend...
I'm sure I'm not the first person with this problem, but sometimes tribal knowledge is the key
I'm looking to boost the signal out of my miniDSP so that I can run a lower gain on my TPA3116 amp. I have a 2.1 setup with the miniDSP as an active crossover driving 2 separate TPA3116 amp boards, one running in bridged mode for the sub. I only need to boost the input signal for the sub amplifier.
I have one of these chinese (likely not genuine TI) NE5532 preamp boards coming https://www.amazon.com/gp/product/B07HG37N1X/
I only need one of the two channels, but I know that OP-Amps are not always happy with one unterminated input/output.
The TI spec sheet says that it is OK to tie it to ground or either supply rail. I was thinking that I would do this directly at the chip on the board, bypassing the rest of the circuit for the second channel. (as opposed to just connecting the input or output of the board to ground). Considering that this is likely a fake and not a real TI chip, do you think it is still OK to do this?
If not, what do you recommend...
I'm sure I'm not the first person with this problem, but sometimes tribal knowledge is the key
I believe the issue of not using and not connecting one OP-AMP channel (out of more channels) relate to if there are no external components mounted for the unused channel (in your case there are). Without a feedback resistor for the unused channel, the unused OP-AMP may behave as a comparator sensing a noise signal and that can hamper good performance of the channels in use.
Conclusion the same - no issue.
Conclusion the same - no issue.
Yes, exactly. Sorry, my mind is usually thinking component-level, not board level. I agree with both of you guys.
I intend to mod the TPA3116 amp to use differential inputs to reduce noise (made a significant difference with the other amp I have in the setup).
Where does it make the most sense to start the differential chain?
I see that I have 2 options:
1) miniDSP--(differential with - tied to output ground)--> NE5532 (1 channel for +, 1 channel for - (amplifying both sides of the pair in parallel)) --> Differential inputs to TPA3116
2) MiniDSP--(normal, ground referenced)--> NE5532 (use only 1 channel) --> Differential inputs to TPA3116
My first thought is to go with option 1, but I'm not sure how well matched the components are on the independent channels of the NE5532 board (or how well they need to be matched, practically speaking).
What are your thoughts?
I intend to mod the TPA3116 amp to use differential inputs to reduce noise (made a significant difference with the other amp I have in the setup).
Where does it make the most sense to start the differential chain?
I see that I have 2 options:
1) miniDSP--(differential with - tied to output ground)--> NE5532 (1 channel for +, 1 channel for - (amplifying both sides of the pair in parallel)) --> Differential inputs to TPA3116
2) MiniDSP--(normal, ground referenced)--> NE5532 (use only 1 channel) --> Differential inputs to TPA3116
My first thought is to go with option 1, but I'm not sure how well matched the components are on the independent channels of the NE5532 board (or how well they need to be matched, practically speaking).
What are your thoughts?
I have the MiniDSP balanced 4x4 and the 10x10. Both have significant noise on output so by lowering the input gain on one for the other you might be robbing Peter to pay Paul (or whoever). The 10x10 is head and shoulders better than the 4x4 but an L-pad on your tweeter might be a good shout down the line if it is particularly sensitive
I believe that graham(x2) is onto something very important. You do not gain anything by introducing a more perfect signal transmission between units that do not perform well for a start. If the power amplifier on its own or the source itself has important noise, that noise is not going to disappear by introducing balanced line signal transfer. It is like installing better carpets in your car when actually the exhaust system is falling apart.
Good points all. I misspoke that there was signal noise, i meant that there was ground loop noise and some hum. (Both of which were reduced in the other amp after going differential with the input and reducing the gain).
The issue is that the amp is already at 32db gain and i’d rather go down than up. Its not so much as the amp or dsp are noisy, i just want to prevent noise due to the longer wires and high gain in the 3116. I also plan to use a pot on the sub input to control the level.
I actually got the pre-amp board this afternoon and connected it up (using option 2 without doing the differential mod). Levels are good so i can turn down the gain and do the differential mod tomorrow.
Here’s a silly question: i only have a 1 channel pot. I’m going to have to get a 2 channel pot in order to use it with a differential signal, right? I can’t imagine it would work if i only used it on one of the differential pair.
The issue is that the amp is already at 32db gain and i’d rather go down than up. Its not so much as the amp or dsp are noisy, i just want to prevent noise due to the longer wires and high gain in the 3116. I also plan to use a pot on the sub input to control the level.
I actually got the pre-amp board this afternoon and connected it up (using option 2 without doing the differential mod). Levels are good so i can turn down the gain and do the differential mod tomorrow.
Here’s a silly question: i only have a 1 channel pot. I’m going to have to get a 2 channel pot in order to use it with a differential signal, right? I can’t imagine it would work if i only used it on one of the differential pair.
If you use a mono-potentiometer on only one of two balanced input, it will have a limited effect. 6dB. You need to use a stereo-potentiometer on both (balanced) channels simultaneously. Two stereo potentiometers for a stereo amp.
Balanced line signal transfer is relevant for hum-currents running in the ground wire. But you may also chose to fight the hum by looking at the power supply(ies).
Last edited:
Exactly! I had some hum in only the un-balanced input amp so I suspected that first. Waiting for some time to complete the mod, but hopefully that will address it.
I also got a new power supply (I found the one that you mentioned in a post previously). The power supply didn't fix it. In addition to the hum, there is a digital noise that appears to modulate with the movement of an antenna within the setup.
I'll post some pictures. too difficult to explain!
I also got a new power supply (I found the one that you mentioned in a post previously). The power supply didn't fix it. In addition to the hum, there is a digital noise that appears to modulate with the movement of an antenna within the setup.
I'll post some pictures. too difficult to explain!
Ok. Here is my setup at the moment.
Forgive the horrible wiring, i’m going to mount everything to a single sided copper clad board to function as a huge star ground/ground plane for all of the boards.
The sub amp (noisy) is the mono one on the upper right. I’ve just completed the balanced input mod right before i took this, but won’t have time to test it. MiniDSP (4x4, not balanced. Taken from my parts bin, unneeded for another project, not necessarily what i would buy for this. but the price was right! is next to it with the green preamp board hanging in free space at the moment.
FauxFrench, I’ve been reading about issues with the SMPS causing interference. This is indeed one of my issues. I tested this setup with my huge linear benchtop supply (24v) and the ‘digital’ sounding interference went away (although the hum remained, as this was before the balanced mod). What type of filtering could get rid of noise on the PS? It sounds almost like a tick tick or a ppppt ppppt with a constant repetition around 1 hz.
I also followed your guidelines (favorite post in the sticky) and narrowed down the source of the noise to the input board (arylic up2stream, upper left). Moving the wifi antenna around changes the intensity of the interference. Swapping out the antenna for the external stub antenna i eventually have for this had no real effect (naturally).
I tried powering the arylic board with a different power supply, no change.
What to try next? I won’t have time to play with it for a few hours, but i have no reason to expect the SMPS interference (if that’s what it is) will go away just by the balanced input mod.
Forgive the horrible wiring, i’m going to mount everything to a single sided copper clad board to function as a huge star ground/ground plane for all of the boards.
The sub amp (noisy) is the mono one on the upper right. I’ve just completed the balanced input mod right before i took this, but won’t have time to test it. MiniDSP (4x4, not balanced. Taken from my parts bin, unneeded for another project, not necessarily what i would buy for this. but the price was right! is next to it with the green preamp board hanging in free space at the moment.
FauxFrench, I’ve been reading about issues with the SMPS causing interference. This is indeed one of my issues. I tested this setup with my huge linear benchtop supply (24v) and the ‘digital’ sounding interference went away (although the hum remained, as this was before the balanced mod). What type of filtering could get rid of noise on the PS? It sounds almost like a tick tick or a ppppt ppppt with a constant repetition around 1 hz.
I also followed your guidelines (favorite post in the sticky) and narrowed down the source of the noise to the input board (arylic up2stream, upper left). Moving the wifi antenna around changes the intensity of the interference. Swapping out the antenna for the external stub antenna i eventually have for this had no real effect (naturally).
I tried powering the arylic board with a different power supply, no change.
What to try next? I won’t have time to play with it for a few hours, but i have no reason to expect the SMPS interference (if that’s what it is) will go away just by the balanced input mod.
Attachments
Last edited:
I did something similar with two TPA3118 and one TPA3116 amps. I used isolated DC-DC converters and balanced connections from the DSP.
Have ytou snooped around for ground loops? If the two amps share the same power ground then they don't need both signal grounds connecting. That will cause a loop.
The advice given to me was to take it apart (disconnect) everything and put it back together ensuring each device works as it should one by one. Then maybe in different orders.
I too used the Arylic board but found it annoying and noisey so my friend opted for an Echo Dot in the end.
Are you using the line in on the Arylic? I found that very noisy and didn't like sharing the audio ground with anything else.
Have ytou snooped around for ground loops? If the two amps share the same power ground then they don't need both signal grounds connecting. That will cause a loop.
The advice given to me was to take it apart (disconnect) everything and put it back together ensuring each device works as it should one by one. Then maybe in different orders.
I too used the Arylic board but found it annoying and noisey so my friend opted for an Echo Dot in the end.
Are you using the line in on the Arylic? I found that very noisy and didn't like sharing the audio ground with anything else.
Just to clarify - your MiniDSP is balanced out?
It may be that knocking the gain on the TPA3116 to 26db is all you need and you can keep the signal chain as simple as possible.
Some pics would help my little brain
Nope, not balanced. I indeed want to turn down the gain to 26! but then the signal is too low to match up with the more sensitive mids/highs at any decent volume. This is the reason for the preamp. Send the 3116 a hotter signal so that i can keep the gain lower.
My thought with the balanced wiring to the amp was just to eliminate the ground loop. Worked great on the other amp i have. Although not as nice as actual balanced outputs, having the twisted pair carrying the signal makes a big difference for interference.
I think the problem you may be having is that the amps are running from the same PSU without sync. You can get round this by using balanced connections throughout. Here is how mine turned out and it was a ball of cables just like yours.
6 Channel TPA3116 + DSP
If you aren't too financially or emotionally attached to the MiniDSP I would recommend the 3E audio DSP which has 4 balanced outs. Problem gone.
6 Channel TPA3116 + DSP
If you aren't too financially or emotionally attached to the MiniDSP I would recommend the 3E audio DSP which has 4 balanced outs. Problem gone.
Very good suggestion about the sync! I had this idea myself. I did the mods to add a sync line (blue 30ga wire if you can see it there) and make the mono amp the slave by changing the gain resistors. Oddly, I didn’t hear a difference.
I think one of you guys will be able to recognize the source of the noise just by hearing it. It is very distinct. Oddly enough, it reminds me of the sound that would come through a (poorly installed) car audio system when a text message would come through on a GSM phone.
Admittedly I haven’t played with class D amps with analog input very much. I have a pair of them (Helix, i forget the models but they have integrated DSP/DAQs in them) in my car, but they are driven by digital signals over fiber. Completely digital signal path. Amazing noise floor.
I digress. I’ll take a few minutes after work this afternoon to record the interference.
I might have to replace the minidsp, but as this ministereo is really just for my computer speakers/ home office stereo my expectations aren’t super high.
Thanks for the help!
I think one of you guys will be able to recognize the source of the noise just by hearing it. It is very distinct. Oddly enough, it reminds me of the sound that would come through a (poorly installed) car audio system when a text message would come through on a GSM phone.
Admittedly I haven’t played with class D amps with analog input very much. I have a pair of them (Helix, i forget the models but they have integrated DSP/DAQs in them) in my car, but they are driven by digital signals over fiber. Completely digital signal path. Amazing noise floor.
I digress. I’ll take a few minutes after work this afternoon to record the interference.
I might have to replace the minidsp, but as this ministereo is really just for my computer speakers/ home office stereo my expectations aren’t super high.
Thanks for the help!
Beautiful craftsmanship on your system!
I’m driving one of those 3” tang band W3-1876s minisubs in a 3d printed (carbon fiber reinforced PLA) transmission line enclosure (of Hexibase fame) and a pair of monitor audio Radius 45 satellites that i have mounted to my monitor VESA mount.
I’m driving one of those 3” tang band W3-1876s minisubs in a 3d printed (carbon fiber reinforced PLA) transmission line enclosure (of Hexibase fame) and a pair of monitor audio Radius 45 satellites that i have mounted to my monitor VESA mount.
Attachments
Cheers! I am very jealous of your 3D printing.
I used a type like this. One for each DSP board and the arylic. Possible (definitely) overkill but it is a very quiet system considering its assembly.
https://cpc.farnell.com/c/electrica...verters/isolated-board-mount-dc-dc-converters
I used a type like this. One for each DSP board and the arylic. Possible (definitely) overkill but it is a very quiet system considering its assembly.
https://cpc.farnell.com/c/electrica...verters/isolated-board-mount-dc-dc-converters
Well, no chance for a recording.
I just toasted my mono 3116 board somehow. Failed the ol’ smoke test. I don’t see anything shorted but it is the chip itself that burned. (You can see right next to the heatsink pad). Do you see anything suspect? (No, that red wire isn’t actually touching the neighboring pin).
It was working fine until i did the differential input mod and decreased the gain. You can see before and after.
Good thing i have a spare, but it is some work to do the updates.
I just toasted my mono 3116 board somehow. Failed the ol’ smoke test. I don’t see anything shorted but it is the chip itself that burned. (You can see right next to the heatsink pad). Do you see anything suspect? (No, that red wire isn’t actually touching the neighboring pin).
It was working fine until i did the differential input mod and decreased the gain. You can see before and after.
Good thing i have a spare, but it is some work to do the updates.
Attachments
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.