Well hello there – I have no knowledge of amplifiers or electronics in in the diy/repair sense but I was hoping there may be some chance I can work this one out with some help from people who are much more knowledgeable than me.
My almost new DTA3116S amplifier from Dayton Audio (yes the smallest one in the known universe) is 100% what I wanted but there is a slight problem – the left channel plays louder and I’m afraid there is not going to be any returns/warranty thing because the shop I bought it from (PartsExpresss) is located on another continent… and I don’t really have everything that came in the box anymore.
Details:
Amplifier is connected to 220 volts/Europe (with another cable, but the same power source).
I have done tests with signal coming from my computer which as it turns out has an extra 0.2-0.3 db output / left channel (connected out to in and measured in Soundforge).
I measured voltage at the output of the amplifier with speakers connected:
4000 hz tone at 250 millivolts – right channel was at minus 5.6%
4000 hz tone at 1000 millivolts – right channel was at minus 3.2%
4000 hz tone at 2000 millivolts – right channel was at minus 2.3%
4000 hz tone at 3000 millivolts – right channel was at minus 2.35%
I also measured the voltage without any speakers connected (just one measurement) – 4000hz tone, Left ch = 1704 millivolts / R chanel = 1671 millivolts – that is a 1.93% difference.
I also connected to amp to the line in of the computer and I discovered a 1.1 db difference between the channels – and I even tried to increase the volume but something in the amp cut the power or something – output signal was gone and replaced with single strong impulses (rare ones)… not going to try this one again it was weird and a warning for not doing things that I don’t really know if they are ok or not; the amp was not damaged.
I also discover while testing that one of the speakers (these are full range) sounded weaker than the other and I had them swapped so that they balance the 2 discrepancies.
Can I hear the difference ? yes … however after the swapping took place it’s more of a “I’m not sure” on most recordings but on white noise it’s still clear that the left channel is louder.
I honestly don’t know how to solve this one maybe it’s just a question of adding some cable length for left channel… I really don’t know. I’m planning to fix the speaker loudness difference (probably replace speakers/enclosure) and after that it’s going to be more of a problem (replace or not) but I would really really like to have this one 100% functional as it should.
My almost new DTA3116S amplifier from Dayton Audio (yes the smallest one in the known universe) is 100% what I wanted but there is a slight problem – the left channel plays louder and I’m afraid there is not going to be any returns/warranty thing because the shop I bought it from (PartsExpresss) is located on another continent… and I don’t really have everything that came in the box anymore.
Details:
Amplifier is connected to 220 volts/Europe (with another cable, but the same power source).
I have done tests with signal coming from my computer which as it turns out has an extra 0.2-0.3 db output / left channel (connected out to in and measured in Soundforge).
I measured voltage at the output of the amplifier with speakers connected:
4000 hz tone at 250 millivolts – right channel was at minus 5.6%
4000 hz tone at 1000 millivolts – right channel was at minus 3.2%
4000 hz tone at 2000 millivolts – right channel was at minus 2.3%
4000 hz tone at 3000 millivolts – right channel was at minus 2.35%
I also measured the voltage without any speakers connected (just one measurement) – 4000hz tone, Left ch = 1704 millivolts / R chanel = 1671 millivolts – that is a 1.93% difference.
I also connected to amp to the line in of the computer and I discovered a 1.1 db difference between the channels – and I even tried to increase the volume but something in the amp cut the power or something – output signal was gone and replaced with single strong impulses (rare ones)… not going to try this one again it was weird and a warning for not doing things that I don’t really know if they are ok or not; the amp was not damaged.
I also discover while testing that one of the speakers (these are full range) sounded weaker than the other and I had them swapped so that they balance the 2 discrepancies.
Can I hear the difference ? yes … however after the swapping took place it’s more of a “I’m not sure” on most recordings but on white noise it’s still clear that the left channel is louder.
I honestly don’t know how to solve this one maybe it’s just a question of adding some cable length for left channel… I really don’t know. I’m planning to fix the speaker loudness difference (probably replace speakers/enclosure) and after that it’s going to be more of a problem (replace or not) but I would really really like to have this one 100% functional as it should.
PE data doesn't match picture of ampboard, picture shows gain is 36dB. It seems you mention the computer has a small difference in left right output, that difference will be amplified ~60 times if picture is correct. Next there is a volume pot that will have unbalance especially in lower part and if gain is 36dB you might never get out of lower part of volume pot, I don't know.
I don't know what "picture" of ampboard we are talking about.
Yes there is a 0.2 db louder left channel from the computer I'm using but the difference in sound is there indifferent of the source connected. I use headphones about as much as speakers and I have never noticed that 0.2 db difference.
Volume knob is at 35-50% the rest is done from the coomputer.
Yes there is a 0.2 db louder left channel from the computer I'm using but the difference in sound is there indifferent of the source connected. I use headphones about as much as speakers and I have never noticed that 0.2 db difference.
Volume knob is at 35-50% the rest is done from the coomputer.
PE shows picture of ampboard. The gainresistors shown in that picture are for ~60 times amplification. The output difference divided by ~60 should be the input difference.
Differences are not very high btw, but if annoying you could lower gain or see if you can adjust balance computersound.
Differences are not very high btw, but if annoying you could lower gain or see if you can adjust balance computersound.
To be honest I don't understand everything you said but if I'm correct you are saying the difference of 2-3% is in the neglectable area? and that these might be due to the 0.2 db difference at the input???
Adjusting the balance is not a solution because I regularly use headphones.
I should do another test to see if there is no db difference at the amp input if I still get the imbalance.
I attached an image with the inside of the amp (mine) just in case it helps.
Adjusting the balance is not a solution because I regularly use headphones.
I should do another test to see if there is no db difference at the amp input if I still get the imbalance.
I attached an image with the inside of the amp (mine) just in case it helps.
Attachments
OOPS! I should have checked this before posting, but since I don't know these things I thought that a 0.2-0.3 db difference at the source it's not big enough to matter; I replaced the source with a laptop and i got a 0% difference at 404 milivolts, as voltage increased to 1500 milivolts the error got to 1.5% but the left channel was the one that was quieter this time.
I think this is not a problem in the amp at all but one from the source greatly amplified initially by the fact that one of the speakers had a weaker nominal SPL.
So that makes average sound cards not exactly left-right balanced? - correct me if I'm wrong.
I think this is not a problem in the amp at all but one from the source greatly amplified initially by the fact that one of the speakers had a weaker nominal SPL.
So that makes average sound cards not exactly left-right balanced? - correct me if I'm wrong.
Most unbalanced part would be the volumepot, normally. I can't read second gainsetting resistor. Below/next to c12 you see a 104 (100k) resistor, that makes 40-20-10 times, next to that 104 is the resistor that determines that.
And what would your conclusion be? I don't see anything wrong the amp, it's the source that I see as the problem because once I replaced it the imbalance turned the other way (right channel became louder by a different margin).
Yes.
And if you want to decrease the difference at the speakers, you could lower amplifier gain. It isn't 60, but still 40 or 20 times, to get it to 10 times you remove the 104 resistor. That might have other benefits, but only you know how much room there is in both volumecontrols you use. If you end up with an amplifier that can not play loud enough to your taste, you should not do it. Amplifier volume pot position you mentioned suggests you use less than 1/3rd of current gain, but if computer volume now is all the time or often at max, lowering gain might not be convenient.
And if you want to decrease the difference at the speakers, you could lower amplifier gain. It isn't 60, but still 40 or 20 times, to get it to 10 times you remove the 104 resistor. That might have other benefits, but only you know how much room there is in both volumecontrols you use. If you end up with an amplifier that can not play loud enough to your taste, you should not do it. Amplifier volume pot position you mentioned suggests you use less than 1/3rd of current gain, but if computer volume now is all the time or often at max, lowering gain might not be convenient.
Datasheet of TPA3116 chip specifies +/-1dB gain setting error, although it does not specify if this is absolute or mismatch. Anyway, since common quality resistors are 1% and gain setting depends on 2 resistors, amplifier gain mismatch up to 2% is normal, unless trimming or 0.1% resistors, which are not common in consumer equipment. This is usually better than common dual gang potentiometers with 5% or 10% mismatch (plus more with aging). +/- 1dB is approx. +/-10%. This is hearable. +/-1% is about +/-0.1dB and is on the threshold of human audibility.
The ampboard gainresistors themselves have no influence on gainerror other than when chip selects 20/26/32/36dB at startup, after startup they play no role in gain, they could be changed and chip continues to play in selected gain. The steps are not 100% the same for different voltagelevels, switchingfrequencies, production tollerances, that is the margin in datasheet. Other resistors in computeroutput might have little influence, but normally volumepot is indeed biggest influence.
In this case the gain setting resistors are etched inside the chip and selected with CMOS switches, which are driven from the digital result of an ADC of external divider voltage, latched at startup. 1% is by far not as easy to achieve in integrated circuits, 2% can be considered quite good for a gain setting of an IC. As a reference to get the sense of gross numbers in IC production I can cite ST Microelectronics datasheets.
I could probably get away with a reduction in maximum power available but I'm not going to make any compromises like that.
I'm going to presume that a more expensive amplifier (something T class maybe/ same sort of power) has better components so the risk of such problems is minimal; I still wait for someone to confirm there is nothing wrong with my amplifier (at this point this is my conclusion); as far as the computer I was going to upgrade it anyway...
I'm going to presume that a more expensive amplifier (something T class maybe/ same sort of power) has better components so the risk of such problems is minimal; I still wait for someone to confirm there is nothing wrong with my amplifier (at this point this is my conclusion); as far as the computer I was going to upgrade it anyway...
the only way I was able to get good gain matching between channels on a preamp or a power amp is to use a step volume pot with 0.1% resistors. The gain setting resistors in the pre amp and power amp were also 0.1% resistors. You can replace the volume pot in the power amp with step volume control. If you did a frequency sweep of your headphones and your speakers there would be spl differences from 20 to 20K. Its just the nature of electro mechanical devices. You would have to pay huge dollars to get speakers and headphones that are matched to less than 2% across the frequency band.
Weakest link in your Dayton amp is the volumepot, that seems to "correct" a little the unbalanced source you use. So improves balance of your unbalanced computer.
You do not limit power perse when you reduce gain/amplification factor, output power remains the same, watt remains the same, you just need more mV signal going into the ampboard to reach same outputlevel. If you can not deliver enough signal to the ampboard, then yes you limit output. Again, all you told is position amplifier volumepot, that suggests you use less than 1 third of gain at the most. Again I don't know if that position correlates to max position on your computer, if it does, then reducing ampboard gain might limit output. Gain is not same as power.
You do not limit power perse when you reduce gain/amplification factor, output power remains the same, watt remains the same, you just need more mV signal going into the ampboard to reach same outputlevel. If you can not deliver enough signal to the ampboard, then yes you limit output. Again, all you told is position amplifier volumepot, that suggests you use less than 1 third of gain at the most. Again I don't know if that position correlates to max position on your computer, if it does, then reducing ampboard gain might limit output. Gain is not same as power.
I can tell you where the volume knob was - at 30-70% depending on what was the voltage, but that does not take into account the source... which was in the 50% region I think, original sinus wave was at minus 7 db I think...etc. matters less, the fact that the volumepot is the weakest link is where I might make an intervention (replace) but the task is beyond my current levels of repair skills.
I just hope things will be acceptable or even 100% ok after I replace computer and even the speakers.
I just hope things will be acceptable or even 100% ok after I replace computer and even the speakers.
At the rear of volumepot you can read value and type, A50K for example, an A is a audio/log pot, a B is a linear pot. For a B pot 70%=70%, for a A audio pot, 70%=25%, input is still attenuated 75% in that position.
Value of the pot often is a bit of a mismatch, but maybe PE ordered a good value, curious 🙂
Value of the pot often is a bit of a mismatch, but maybe PE ordered a good value, curious 🙂
Well I made more than just one picture but it's not clear in the picture what it says but I read it as 850k, 5 maybe a 6 or a B... don't know... B60K or B50K I guess...
B50k, linear 50kOhm. That goes to an amplifier chip that with gainsetting resistors shown in your pic has an inputimpedance of either 30k or 15k, PE had no say or didn't look at it. If you turn pot without computer playing music you have two zones where noise is louder than at max position, equal positions around 12 o'clock because linear pot. Because you use computer to set volume, avoid those areas for Dayton volumecontrol.
Lowest gainsetting would fit the B50k pot better, but you might need the gain, if 70% equals 4V you do not have much left unless .....computer being unknown. If you ever buy a new pot you can go for 10k or even 5k (if source can drive headphone the 5k won't be a problem, but not every source can drive a headphone and some might not be able to drive the much easier than headphone 5k pot)
Lowest gainsetting would fit the B50k pot better, but you might need the gain, if 70% equals 4V you do not have much left unless .....computer being unknown. If you ever buy a new pot you can go for 10k or even 5k (if source can drive headphone the 5k won't be a problem, but not every source can drive a headphone and some might not be able to drive the much easier than headphone 5k pot)
- Status
- Not open for further replies.