Cayin RU6: R-2R USB Dongle DAC with Head-Amp
Jun 7, 2022 at 1:07 PM Post #2,582 of 3,665
Multiple things, LG v35 with UAPP, Ibasso DX160, PC, Tablet.. It does not matter, the result is always the same, because as I said, it is not a static I believe it is just imperfection of the DAC.
Unfortunately I still am unable to understand what you are describing and it would seem as if you've already made up your mind rather than figure out if it's an isolated problem with your gear or your particular RU6.
 
Jun 7, 2022 at 1:22 PM Post #2,583 of 3,665
Unfortunately I still am unable to understand what you are describing and it would seem as if you've already made up your mind rather than figure out if it's an isolated problem with your gear or your particular RU6.
Well, did you try what I suggested? lower the source volume almost to 0 and increase it on the DAC, and pay attention to what happens to noise during volume steps. I feel like you are the one disregarding my claims without checking it. If I hear any counter arguments or counter evidence to what I say I will change my mind easily.
 
Jun 7, 2022 at 1:25 PM Post #2,584 of 3,665
It is not static, it is a weird noise during the very quiet sounds (but not during the silence). It can only be noticed during very quiet parts of high dynamic range music. But for the demonstration you can reduce the source volume a lot(almost to 0), then increase cayins volume and check the noise difference between 64 and 65 for example, you will know what I am talking about.

Please don't. R2R is demanding on the data integrity of incoming bitstream, when you lower the source volume to a certain level, the R2R decoding algorithm fails to perform correctly.

It is VERY IMPORTANT to keep the source at full output level when you use R2R-designed DAC. If you need to control the volume at the source level, delta-sigma DAC probably suits your needs better.
 
Last edited:
Cayin Stay updated on Cayin at their sponsor profile on Head-Fi.
 
http://en.cayin.cn/
Jun 7, 2022 at 1:48 PM Post #2,585 of 3,665
I have a theory about the volume control, please correct me if there is any information which disproves this.

The resistor volume control has only 8 steps (probably by having 3 resistors with different combinations) and substeps between them change volume only digitally (at the dac level or before). This means that when the volume is for example 30, it is digitally maxed out and at 31 it switches to a lower resistance resistor combination. Now this would increase volume and noise floor a lot, but to compensate for this it decreases volume digitally. Therefore what we get is increased noise by a lot, with barely louder music, hence the damaged SNR. Then moving from 31 to 40 the noise does not change because the volume is increased digitally, and at 40 the dac is using its full potential again.

Anyway, one thing is clear that if we want to maximize quality, we should listen to volumes right below the 'critical' steps. But it is really frustrating that the noise level in this dac is already not great and it does not even use its less than stellar full potential because of imperfect volume control. I wonder why cayin could not implement completely analog volume control, maybe like 6 resistors and 64 steps or smth.

The volume control of RU6 is known as Resistor Array volume control, it is not a new technology, you can find this type of volume control in many high-end integrated amplifiers and pre-amplifier. We have explained our design consideration and implementation detail of the Resistor Array volume control at the opening post of this thread The Resistor Array volume control involves switching relays when you hop from one segment to another, so it is different from your theory. The RU6 volume control is 100% analog, so when you said "moving from 31 to 40 the noise does not change because the volume is increased digitally, and at 40 the dac is using its full potential again" is a completely wrong assumption.

we need to mute the output for a ~40s when a relay kicks in. This will introduce a small delay in volume adjustment at 10, 20, 30,... etc. A lot of users have experienced the delays, but I didn't recall any user complaints regarding high noise level at the step when you jump from one volume segment to another. Theoretically, this will happen when your RU6 fails to mute the switching relays properly, but the mute mechanism is very reliable theoretically, I'll be very surprised if this happens in your RU6.

I have copied and pasted the Resistor Array volume control paragraph from Post #1 to here for your reference, but please go through the complete opening post to make sure you understand and operate the RU6 properly.

Cayin tested numerous off-the-shelf volume options, unfortunately they cannot meet the high precision, low noise, and low power consumption requirements in dongle DAC application. For instance, we have used PGA2311A in several DAPs and R01 Audio Motherboards, we tried to repeat the trick but the RU6 developed unacceptable background noise. There are better volume chipset in the market, but they are either too big physically, or drain a lot of power, making it inappropriate for Dongle DAC application. Eventually Cayin bite the bullet and developed a high precision resistor array volume control circuit that provides 99 steps volume control through 9 segments of resistors and switching relays.
1654623121399.png


Resistor Array volume control is not a new technologies, it have been around for a very long time. They are very high quality volume control design, extremely transparent when implemented correctly and you can find them in a lot of high-end preamp and integrated amplifies. Each segment of resistor array can only provide ~10 steps of volume control, this is obviously not enough in real life application, so you need to "hop" to another set of resistor array for another 10 steps, and so on, and so on.

We don't recall anyone used resistor array volume control in a Dongle DAC, probably not even in portable DAP. There is no way we expected this final outcome when we started the RU6 project back in early February 2021 (when we completed R01 R&D). We basically completed the RU6 circuit design in very fast track but we stuck at the volume control issue for 3 months trying different solutions and going nowhere. Using Resistor Array Volume Control is our final trump card. It is basically an out of proportion implementation for Dongle DAC and that us another month to finalise the furrent 9 segment resistor array design.

Our main hesitation with resistor array volume control is not about technical difficulties or budget consideration. The Resistor Array volume control involves switching relays when you hop from one segment to another. The relay will produce very mild pop sound with speakers and is inaudible when you are 2 or 3 meters away from the speaker. Unfortunately the pop noise will become very annoying to sensitive IEM users. For this reason, we need to mute the output for a short moment when a relay kicks in. This will introduce a small delay (around 40ms) in volume adjustment, definitely a set back from user experience point of view. For dedicated audiophiles who put audio performance, this is a small price to pay in order to implement R-2R resistor ladder network in a dongle DAC. but we can understand some users might consider this as a deal breaker, that's why we explained this in detail up front.
 
Cayin Stay updated on Cayin at their sponsor profile on Head-Fi.
 
http://en.cayin.cn/
Jun 7, 2022 at 2:33 PM Post #2,586 of 3,665
First of all, thank you for responding.
Please don't. R2R is demanding on the data integrity of incoming bitstream, when you lower the source volume to a certain level, the R2R decoding algorithm fails to perform correctly.
I am not doing this, of course. It was just to demonstrate issue that is difficult to notice without reducing volume (but is still there at max volume during very very quiet parts of music).
A lot of users have experienced the delays, but I didn't recall any user complaints regarding high noise level at the step when you jump from one volume segment to another.
I am not experience noise during switching, what I said is that noise that is already there increases at these volume steps (and is constant during other volume steps).
Here is the quick image to demonstrate what I meant all this time, and I could not communicate well enough, since everyone misunderstood.
Notes_220607_141633.jpg

The RU6 volume control is 100% analog, so when you said "moving from 31 to 40 the noise does not change because the volume is increased digitally, and at 40 the dac is using its full potential again" is a completely wrong assumption.
I stand corrected, I trust that it is not digitally controlled, but could you please explain why the noise behaves the way I showed on diagram? Or would you say I am wrong about this and noise always increases proportionally at every step?
 
Jun 7, 2022 at 4:58 PM Post #2,587 of 3,665
First of all, thank you for responding.

I am not doing this, of course. It was just to demonstrate issue that is difficult to notice without reducing volume (but is still there at max volume during very very quiet parts of music).

I am not experience noise during switching, what I said is that noise that is already there increases at these volume steps (and is constant during other volume steps).
Here is the quick image to demonstrate what I meant all this time, and I could not communicate well enough, since everyone misunderstood.


I stand corrected, I trust that it is not digitally controlled, but could you please explain why the noise behaves the way I showed on diagram? Or would you say I am wrong about this and noise always increases proportionally at every step?

Now we are getting somewhere.

These are the noise characteristics of an R-2R DAC circuit, it has nothing to do with the Resistor Array volume control, in fact, the resistor array volume helps to control the R-2R noise significantly. If we use an off-the-shelf volume chip such as PGA2311A, the R-2R noise will be even worse, significantly worse. (we have also explained this in the opening post of the RU6 thread).

We didn't explain the noise pattern specifically in the RU6 thread, but we have explained it in detail in N6ii thread. We only mentioned the noise problem when we discuss Line out option from RU6 (or to put it correctly, the noise pattern has discouraged the implementation of line out from RU6.

The R-2R noise pattern of RU6 is similar to R01 Audio Motherboard (for obvious reasons). You can read the R01 Audio Motherboard announcement HERE. For convenience sake, I have copied and pasted the more related paragraph below:

The R01 is a headphone-output-only Audio Motherboard. Like E01, It does not have a line-out feature. We have been fighting very hard to fit the R-2R Resistive Ladder Network into the Audio Motherboards PCB. Line out from the sigma-delta DAC chipset is relatively straightforward because we can take the LPF output as the foundation of the line out signal. The R-2R design requires more work to deliver a dedicated line out. The background noise pattern of the R-2R network also makes it inappropriate for pseudo lineout implementation. In other Audio Motherboards, the background noise basically remains the same level at all volume levels, so it will affect the sensitive IEM users as they tend to listen to low volume settings (i.e, same amount of noise but lower output level). On the other hand, the background noise of the R-2R ladder DAC circuit will increase as you turn up the volume. While sensitive IEM users might welcome this pattern, it makes pseudo line out inappropriate because when you max out the volume of the amplification circuit, the background noise becomes audible.

We have also mentioned the R-2R noise pattern HERE, HERE, and HERE.

If you want to look into the fine detail of R-2R noise, they are mainly thermal noise: when you pack a larger number of resistors into a very tight space, the resistors will warm up and generate thermal noise. The noise is output-related, when you drive your circuit to a higher output level, thermal noise will increase. When you implement R-2R in a more spacious environment, the thermal noise will decrease. Unfortunately, R01 Audio Motherboard and RU6 Dongle DAC are among the smallest and most cramped implementations ever, that's why RU6 suffers more thermal noise than other R-2R implementation.
 
Cayin Stay updated on Cayin at their sponsor profile on Head-Fi.
 
http://en.cayin.cn/
Jun 8, 2022 at 12:18 PM Post #2,588 of 3,665
Now we are getting somewhere.

These are the noise characteristics of an R-2R DAC circuit, it has nothing to do with the Resistor Array volume control, in fact, the resistor array volume helps to control the R-2R noise significantly. If we use an off-the-shelf volume chip such as PGA2311A, the R-2R noise will be even worse, significantly worse. (we have also explained this in the opening post of the RU6 thread).

We didn't explain the noise pattern specifically in the RU6 thread, but we have explained it in detail in N6ii thread. We only mentioned the noise problem when we discuss Line out option from RU6 (or to put it correctly, the noise pattern has discouraged the implementation of line out from RU6.

The R-2R noise pattern of RU6 is similar to R01 Audio Motherboard (for obvious reasons). You can read the R01 Audio Motherboard announcement HERE. For convenience sake, I have copied and pasted the more related paragraph below:



We have also mentioned the R-2R noise pattern HERE, HERE, and HERE.

If you want to look into the fine detail of R-2R noise, they are mainly thermal noise: when you pack a larger number of resistors into a very tight space, the resistors will warm up and generate thermal noise. The noise is output-related, when you drive your circuit to a higher output level, thermal noise will increase. When you implement R-2R in a more spacious environment, the thermal noise will decrease. Unfortunately, R01 Audio Motherboard and RU6 Dongle DAC are among the smallest and most cramped implementations ever, that's why RU6 suffers more thermal noise than other R-2R implementation.
Thanks a lot for this very informative answer.
I do not want to sound very stubborn but one thing is still unclear. As you quoted "the background noise of the R-2R ladder DAC circuit will increase as you turn up the volume", which makes perfect sense, but the noise I am talking about is not increasing at every step.
The noise level is exactly the same at all volumes from 21 and 30 for example, and then jumps up to at volume 31(as I showed on the diagram).
 
Jun 8, 2022 at 12:32 PM Post #2,589 of 3,665
Thanks a lot for this very informative answer.
I do not want to sound very stubborn but one thing is still unclear. As you quoted "the background noise of the R-2R ladder DAC circuit will increase as you turn up the volume", which makes perfect sense, but the noise I am talking about is not increasing at every step.
The noise level is exactly the same at all volumes from 21 and 30 for example, and then jumps up to at volume 31(as I showed on the diagram).

I don't hear what you described, and I seriously doubt human hearing can compare noise levels accurately, especially when we are using different IEM or headphone.
 
Last edited:
Cayin Stay updated on Cayin at their sponsor profile on Head-Fi.
 
http://en.cayin.cn/
Jun 8, 2022 at 12:46 PM Post #2,590 of 3,665
I don't hear what you described, and I seriously doubt human hearing can compare noise levels accurately, especially when we are using different IEM or headphone.
If you tried the test I described, lowering volume of the source and increasing it on the RU6 and then checking the noise levels you would know right away what I am talking about, because it is very apparent.
So either my unit is defective (which I doubt) or no one here really tried to understand what I said. Anyway, thanks for your time.
 
Jun 9, 2022 at 2:55 PM Post #2,592 of 3,665
does anyone have the details with the warranty of this product? Mine loses its power and connection constantly when anything touches the usb c

Does it happen with usb-c to usb-c cable (when connected to your smartphone) as well as usb-c to usb-A cable (connected as usb dac to your laptop)? If both are failing and you tried multiple cables, then usb-c plug on RU6 could be at fault. Otherwise, could be an issue related to one of the cables.
 
Jun 9, 2022 at 7:06 PM Post #2,593 of 3,665
I talked to ddHiFi and I believe they will be working on usb-c adapter version, similar to the one above with a separate data and charging ports. Otherwise, just go with a usb-c splitter, I have one, talked about it before in this thread, it works, but adds bulk. Plus, many companies that make these splitters use cheaper usb-c connectors. ddHiFi quality is usually better.

For now, your alternative is something like this:

cayin-ru6-x15.jpg cayin-ru6-x16.jpg
That’s why I said the USBC version was coming soon
 
Jun 10, 2022 at 12:33 AM Post #2,594 of 3,665
Does it happen with usb-c to usb-c cable (when connected to your smartphone) as well as usb-c to usb-A cable (connected as usb dac to your laptop)? If both are failing and you tried multiple cables, then usb-c plug on RU6 could be at fault. Otherwise, could be an issue related to one of the cables.
tried multiple cables as well as connecting to my macbookj.
 
Jun 10, 2022 at 3:09 AM Post #2,595 of 3,665
tried multiple cables as well as connecting to my macbookj.
Sounds like ur RU6 connector plug is indeed faulty. Call your dealer~
 

Users who are viewing this thread

Back
Top