Hi Kartick,
I've had this same difference between left and right channels on every DDDac I've built (around 20 now). I can't explain it but can't here it either. It's nothing to do with offset. It's simply channel imbalance
I have some "old" boards in my DDDAC, and I added some "new" boards later on. You can mix them without any problems.
Hi David,
I also think it's not an offset issue as much as it is a channel imbalance. Actually, what triggered me to go measure the voltages across the outputs was that I was sensing the right channel sounding slightly more prominent than the left and as a result the overall image shifts to the right ever so slightly (but it is pretty noticeable). When I play the same track (same recording from the same PC), with another dac, the image comes back to the center.
I believe the Vout readings would be the cause of what I am hearing. Also, I would like to center the image with the DDDAC. Can I remove this imbalance in any way? Thanks.
This should not be the case. The DC bias has nothing to do with the AC signal after conversion. Seems you have another issue. I measure the channels to be within tenths of dB equal. Not noticeable.
Best way is to download somewhere a test tone of 400 Herz or something, not critical, anything between 200 and 2khz will do the job anyway, play the test tone and measure at your speaker with a simple Multimeter the AC voltage.
Do this with the dddac and your other DAC
Now you know if you have channel imbalances. If there are come back and I will give some more tips how to solve
Do you use multiple boards by the way ?
Yes Doede, I use 4 boards with the old motherboard.
I have test tones but can you please explain how can i measure the ac voltage at the speakers?
Also, the voltage difference that i am getting between left and right, is it high enough to cause audible issues ? David mentions that he got this kind of variance in 20 of his builds.
I have test tones but can you please explain how can i measure the ac voltage at the speakers?
Also, the voltage difference that i am getting between left and right, is it high enough to cause audible issues ? David mentions that he got this kind of variance in 20 of his builds.
Ok so I guessed and played a 1000Hz test tone at normal listening volume and got these figures:
Left channel: 0.418v
Right channel: 0.387v
does this seem correct or did i goof up?
What seems to be the reason for this variance?
Interestingly these readings change with volume and with the frequency of the test tone played.
Left channel: 0.418v
Right channel: 0.387v
does this seem correct or did i goof up?
What seems to be the reason for this variance?
Interestingly these readings change with volume and with the frequency of the test tone played.
I read David's comment only on DC bias differences…. which is a completely different thing
Last edited:
so the channels are unbalanced so much is clear 418/387 = 1,08 This equates to (only) 0,7dB - this is not much but probably audible. The DAC should not have this difference by the way.
I asked for the number of boards as if one board would not be playing with the rest (this happened !) you would see like 100% versus 75% obviously with 4 boards and that would account for 2,5dB - so we can rule this out.
never the less, I would check solder joints and re-solder/flow the connections between the boards for starters
And when playing the "other DAC" you get values which are closer ?
volume change is of course having effect on the readingas output voltage goes up/down
with frequency as most multimeters are not suitable for measuring over a wider AC range, normally up to 400Hz or so they are ok and reading should be stable
No, in post#6922, he is confirming that he gets the same voltage difference that i get between the left and right channels. I don't think he mentions dc bias differences.
Anyway, one more thing could there be a problem if my 12v psu was outputting say 11.7v and that would lead up to this imbalance? I dont know if there should have been a change but i adjusted the voltage to make it reach to 12v and then when i measured the dc voltage across the left and right, it measures equal.
Stranger things are happening !!!
Thank you. I knew that the boards can be mixed, I am just looking for a deal on the original boards. I bought my board in 2014 the price is considerably more now. Also, I don't need the Tent shunts as I am using Ian's LiFePO4 battery board to power the analog section of the DACs. A very noticeable improvement btw. I can't say whether or not it's true but Ian claims this power source bests any regulator of any design.
The DC voltage is the result of the bias current flowing out of the DAC output through the load resistors to GND. The DC bias is not related to the AC signal voltage. The slight imbalance you have is normal, and there is no point in balancing the DC level of the left/right channels.
You need to block the DC voltage using capacitors between the DDDAC output and the input of our (pre)amplifier. Alternatively, you could use a transformer to couple the balanced outputs of the DDDAC to the amplifier. It's all in the DDDAC documentation.
But there's definitely something else at play here.the v out from the left channel has been consistently lower by around the amount Kartick measured whereas with identical circuits the v out should measure the same. As I said, I've never heard this as an imbalance between the channels, but I'm very interested to know the cause
With 2.66V and 2.70V, the "bias imbalance" is about 1.5%. This is as close to nothing as I can think of, and I don't see how this could make a difference in any way.
I am using cinemags and the dc on the outputs is 0. So what you say in the second para is not the case. I don’t need to take care of the same.
Hey David, as Doede suggested, i am yet to measure the readings by playing the test tones via my other dac (Monarchy Audio), which, with normal music content (not test tones), centres the image better than dddac at the moment. And the imbalance appears to even out or go away.
That is my only concern that if the vout difference is not so huge, why is it surfacing out as a channel imbalance?
Off the topic, but please can someone confirm the current requirements of :
1. Wave IO board
2. Allo Kali board
I have been experimenting with different power supplies feeding Kali (when it is used in isolation, i.e., without an sbc, between the wave io and dddac) and the results are pretty interesting. Only feeding the Kali with different supplies changes the sound of my whole system in a significant manner. Did anyone else have such experiences. Basis my observations, I need to assess if I can make some changes to how I am feeding the WaveIO board.
1. Wave IO board
2. Allo Kali board
I have been experimenting with different power supplies feeding Kali (when it is used in isolation, i.e., without an sbc, between the wave io and dddac) and the results are pretty interesting. Only feeding the Kali with different supplies changes the sound of my whole system in a significant manner. Did anyone else have such experiences. Basis my observations, I need to assess if I can make some changes to how I am feeding the WaveIO board.
Cause and effect: as I've said I've never heard any channel imbalance but I've measured this voltage difference on every pair with the difference ranging from a couple of percent to around six percent. Being a curious type I'd like to know why.....
I had an imbalance of 3-4mV with 33ohms TX2575 (with much better than 0.1% tolerance).
This reduced to 1mV after I improved on balanced current-out tracks' impedance. I also improved ground uniformity/integrity. My DAC has 4 original DAC boards.
With time (a year later), the imbalance increased up to a whopping 2mV, but only on one channel; the other stayed at 1mV.
This reduced to 1mV after I improved on balanced current-out tracks' impedance. I also improved ground uniformity/integrity. My DAC has 4 original DAC boards.
With time (a year later), the imbalance increased up to a whopping 2mV, but only on one channel; the other stayed at 1mV.
I don't see why you would be able to hear a channel imbalance of the DC bias. Humans don't hear DC, and the DC is blocked by a capacitor or transformer. The reason for the imbalance would be slight tolerances between the load resistors or in the bias currents of the DAC chip(s).
If a an output transformer is used, taking care of an imbalance of the POS and NEG outputs of the same channel may be useful to minimize the DC current flowing through the primary winding. But we are discussing a 0.04 V "imbalance" between left and right channels here, which is irrelevant.
Last edited: