You've got a lot of time on your hands! Good insight.
I like the questions you raise, but your means seems passive-aggressive and statements verge on exaggerated. Although I'm sure not intended in that way.
I'd suggest this is a bit harsh to post this, without asking Ian privately as to why there is a difference, as there could be many reasons. As I'm sure you will understand building products is never a linear business model. It could be simply to do with the volumes he has ordered for specific parts as he expects the 3.3v to sell less or part shortage of the pandemic etc.
Keep asking the right questions, but it would be nice if it could be in a constructive way and give Ian the opportunity to answer privately for anything that could be controversial
Since we are talking about commercial devices rather than diy (no schematics, nothing to build and so on), I believe that Ian is free to sell his devices at the price he feels most appropriate, like everyone is free to buy or not his devices.
BTW, this thread is very singular, it looks like a battle between believers and unbelievers rather than a place to discuss technical aspects in a constructive way.
When I did ask technical question, typically I got no reply.
Maybe I had to ask privately?
Maybe, worth a shot
I suppose as a commercial product, as you say, he has no requirement to share his IP.
Considering your building a competitive product, private communication might be more effective.
Looking forward to try your battery supply too.
Last edited:
I think there is a big difference, I share my projects, schematics, BOM and so on, the spirit of diy audio.
So I'm not a competitor, since my designs disregard commercial aspects, sometimes I lost money, and mostly there is no budget limit to reach the best performance as possible.
Ian's competitor is Allo, not me.
And I reply to all questions as soon as I find the time (since this is a hobby and not a business for me).
It's very curious that nobody is interested in technical discussion, objective measurements, design choices and so on.
So I'm not a competitor, since my designs disregard commercial aspects, sometimes I lost money, and mostly there is no budget limit to reach the best performance as possible.
Ian's competitor is Allo, not me.
And I reply to all questions as soon as I find the time (since this is a hobby and not a business for me).
It's very curious that nobody is interested in technical discussion, objective measurements, design choices and so on.
Thank you for your kind reminder, but if you want me to ask why there is a difference between the 3.3V and 5V versions, I don’t think it’s necessary to ask because it only involves Ohm’s law. Under the formula of V=IR, the difference between the 3.3V and 5V versions is of course the resistance. Earlier, I mentioned that I bought the wrong 5V version, and I should buy two 3.3V. I once asked Ian how to change 5V to 3.3V myself, but Ian replied that it is unlikely that I could modify it myself. At the same time, Ian further stated that he did not have time to help me make changes. Even if it was only exchanged, he was not available. So, what else do you think I can ask?
Well, since this site is called DIY Audio, it is to be able to DIY, so I first announced that I initially found the difference between the 3.3V and 5V versions, and then began to study whether to replace the components on the 5V version with 3.3V , I can complete the modification of both modules are 3.3V output. Furthermore, based on these research results, I will redraw the circuit diagram according to my own ideas and change it into a more ideal module. For example, only one Jumpers is needed to switch between the 3.3V and 5V versions. There is no need to distinguish between the 3.3V and 5V versions, and it is more flexible in use.
For example, EA6 and EA8, which can be seen everywhere on the board, are reference voltage ICs. ICs of the same brand are linked as follows. The datasheets of EA6 and EA8 have detailed instructions on how to use them, and then compare the relevant components on the board to easily draw the circuit schematic.
Diodes Incorporated|Diodes Incorporated AZ432ANTR-E1|Voltage References|LCSC
Since this may infringe on Ian's intellectual property, I will publish the designed circuit diagram file for free, without involving commercial activities, and only provide it to those who are interested in making changes.
Last edited:
After the test, I feel very disappointed. ConditionerPi did not bring any improvement. Maybe I already use an isolator, and any interference from Pi has been isolated.
In this way, UcConditioner is probably the same. I bought Ian's new toy, but I didn't feel more happy for the first time. I have bought more than 800 dollars in total, and I shouldn't buy it again.
In this way, UcConditioner is probably the same. I bought Ian's new toy, but I didn't feel more happy for the first time. I have bought more than 800 dollars in total, and I shouldn't buy it again.
Exactly as expected, see my post #1185.
There are other ways to improve the system but they involve to redesign the whole architecture of the chain, as I suggested several times.
While it was clear at design level, it could be measured, but when I would help with objective measurements I got no reply, nobody is interested to know the real jitter of the system.
BTW, if I was a competitor, like someone claims, I would not disclose the right architecture, I would keep it a secret to be implemented in my design.
Finally, these are audio toys (nothing to do with diy) offered at reasonable prices, it's just an audio shop to get devices at honest quality/price ratio.
But if you are looking for the best then it's another thing, developing time are longer and costs are much higher.
Keep in mind we have spent the last 4 years to develop a pair of SOTA oscillators.
There are other ways to improve the system but they involve to redesign the whole architecture of the chain, as I suggested several times.
While it was clear at design level, it could be measured, but when I would help with objective measurements I got no reply, nobody is interested to know the real jitter of the system.
BTW, if I was a competitor, like someone claims, I would not disclose the right architecture, I would keep it a secret to be implemented in my design.
Finally, these are audio toys (nothing to do with diy) offered at reasonable prices, it's just an audio shop to get devices at honest quality/price ratio.
But if you are looking for the best then it's another thing, developing time are longer and costs are much higher.
Keep in mind we have spent the last 4 years to develop a pair of SOTA oscillators.
Not true - I'm interested
Just like I'm interested in knowing what (measured) improvement to expect from implementing one of your clocks.
I can send you a RPi/FifoPi/DAC HAT setup with supplies if you are interested.
I live in Milan so postage costs should be limited.
If interested please do drop me a message.
Makes sense, hopefully you didn't loose too much money - that may be taking the passion too far! I suppose it is a dying industry.
Hopefully your high standards don't put of entrepreneurial people like Ian, as there isn't enough competition as it is. There just isn't any real money to be made in this space.
Thats nice, looking forward to seeing the design. I will give it a try
I suppose, in the nature of DIY that YunYun and Andrea are advocating. That I should take the time to add my experience with Ian's ultracapacitor and LifeP04 supplies.
I have been tinkering with Ian's DAC since September and introduced ultracaps to every 3.3v rail and the sonic gains were significant.
Upon the release of his Mk3 board and UCconditioner range, I decided to implement them slowly. The MK3 board, may have made an improvement but nothing substantial - however the firmware upgrade for 5v was welcome. I was sceptical that 3.3v UChybrid boards would make a negative impact vs them being hardwired and balanced by a single resistor - I always go for simplicity where possible. I was pleasantly surprised that there was no degradation to the power supply and it sounded the same before.
(Tangent) I then upgraded my ESS dac to his new 9038plus.. That was a surprisingly decent step-forward. Clearly a better thought out design. Increased copper, meant I had to do less to combat the RFI/EMI from the clocks underneath (I still do) and there are no more components underneath that could potentially create a short.
I then introduced the 5v Ucconditioner to my digital source, clean side of my USB board and the battery board. Underlying supply for the external 5v was the Shanti, which I would consider an average power supply at best (despite its low noise rating). This combination surpassed my expectations, so much so I have sent them out on loan for confirmation.
At that point I wanted to base-line the DAC, so I started looking at comparable ESS designs. I ended up choosing to get a Auralic Vega G2. ESS SDM conversion and lots of marketing material around their clocking and class A output stages, seemed like a good benchmark for it. I wasn't expecting the Ian dac powered by the Battery PS and Ultracaps to better it. Well I was in for a big surprise, it far surpassed it and I sent the Vega G2 back three days later. I tested without the ultracaps and the results were a lot closer, but still lent towards the Ian Stack.
It has taken a significant amount of optimisation to get to here though.
With further testing with friends, I would say the ultracap 5v board would improve most supplies. Providing that the cables are thick and location to target is as close as possible. What we did find though, is that if the powersupply was good enough the impacts were less. For example, the MPaudio 10x LT3045 (single rail) had a very very marginal improvement if at all. Importantly though it didn't negatively impact it.
I am on the group buy for Andrea's clocks and looking forward to see what that brings to the Ian dac. I also plan to buy his entire R2R stack, to compare. Great to have options and a bit of healthy competition.
I have been tinkering with Ian's DAC since September and introduced ultracaps to every 3.3v rail and the sonic gains were significant.
Upon the release of his Mk3 board and UCconditioner range, I decided to implement them slowly. The MK3 board, may have made an improvement but nothing substantial - however the firmware upgrade for 5v was welcome. I was sceptical that 3.3v UChybrid boards would make a negative impact vs them being hardwired and balanced by a single resistor - I always go for simplicity where possible. I was pleasantly surprised that there was no degradation to the power supply and it sounded the same before.
(Tangent) I then upgraded my ESS dac to his new 9038plus.. That was a surprisingly decent step-forward. Clearly a better thought out design. Increased copper, meant I had to do less to combat the RFI/EMI from the clocks underneath (I still do) and there are no more components underneath that could potentially create a short.
I then introduced the 5v Ucconditioner to my digital source, clean side of my USB board and the battery board. Underlying supply for the external 5v was the Shanti, which I would consider an average power supply at best (despite its low noise rating). This combination surpassed my expectations, so much so I have sent them out on loan for confirmation.
At that point I wanted to base-line the DAC, so I started looking at comparable ESS designs. I ended up choosing to get a Auralic Vega G2. ESS SDM conversion and lots of marketing material around their clocking and class A output stages, seemed like a good benchmark for it. I wasn't expecting the Ian dac powered by the Battery PS and Ultracaps to better it. Well I was in for a big surprise, it far surpassed it and I sent the Vega G2 back three days later. I tested without the ultracaps and the results were a lot closer, but still lent towards the Ian Stack.
It has taken a significant amount of optimisation to get to here though.
With further testing with friends, I would say the ultracap 5v board would improve most supplies. Providing that the cables are thick and location to target is as close as possible. What we did find though, is that if the powersupply was good enough the impacts were less. For example, the MPaudio 10x LT3045 (single rail) had a very very marginal improvement if at all. Importantly though it didn't negatively impact it.
I am on the group buy for Andrea's clocks and looking forward to see what that brings to the Ian dac. I also plan to buy his entire R2R stack, to compare. Great to have options and a bit of healthy competition.
Hey Yunyun,
I see you are using stock clocks here. They are very bad and I would look to replace them asap. That is by far the limiting factor, I cant actually enjoy listening to music with them regardless of optimisations I have made.
I know Ian has built the UcConditioner to help the raspberry PI, but that is genuinely like trying to polish a turd when it comes to replaying audio in my limited experience. You would be much better of using an allo usbsig (I2S) if budget restratined, which will be okay. If you really want to go all out, get a good USB to I2S board - slave it to the FIFO (obviously with better clocks) and an excellent USB source. J-Cat USB XE is a better USB source - compared to say a USBsig, but at a price.
Drop me a pm with your entire stack and configuration, I will try my best to help you get more out of your investment.
Edit: Forgot to mention, stacking anything as noisy as a PI or even a Usb sig under the FIFO - is far from ideal. If you do, at least get some copper tape, wrap it around some card and then protect it in kapton tape to reduce upwards radiation.
Last edited:
sebbyp,
The clock I really use is the pulsar clock with metal cover Ian said is the best on McDualXO. Even if it is USBRIDGE, it still uses USB, and USB itself is the source of huge amount of jitter, why should I use it? The real best signal source is still CD transport, so if you look carefully, you will find I have used ReceiverPi.
The price is not expensive but very good CD transport is cec TL3 3.0. It’s much better than spending a bunch of meaningless money on USB.
The clock I really use is the pulsar clock with metal cover Ian said is the best on McDualXO. Even if it is USBRIDGE, it still uses USB, and USB itself is the source of huge amount of jitter, why should I use it? The real best signal source is still CD transport, so if you look carefully, you will find I have used ReceiverPi.
The price is not expensive but very good CD transport is cec TL3 3.0. It’s much better than spending a bunch of meaningless money on USB.
UcConditioner should be really useful, because if conditions permit, UcConditioner can directly provide 3.3V to the digital circuit, and avoid using LDO in the digital circuit.
However, we can't avoid using LDO on the raspberry pi, so ConditionerPi is not really helpful, which is easy to understand. Even if I can learn from theory, I still try to buy it directly, but I advise everyone to stop wasting money to buy ConditionerPi. As for the LDO module used to charge the super capacitor, there is no need to buy it specially, unless you don't have an LDO module on hand. The reason is very simple, let me say it again, as Ian himself said when UcConditioner is supplying power, it will no longer use the current of the charging LDO module, so no matter what is used for charging, the overall effect is the same. . In other words, the short life and expensive LiFePO4 is not worth buying.
When you use Auralic Vega G2, because you can't directly input 3.3V, you can't fully obtain the advantages of UCconditioner. You should buy Buffalo-IIIse-Pro without LDO.
Last edited:
I am glad you have a better clock! The pulsar is brilliant.
Hmm, well I've seen a few implementation improvements you can make and yes I saw you use the ReceiverPI as I do too.
I will leave it up to you then
Sometimes ignorance is bliss in this hobby and ultimately saves some money.Over and out for me and best of luck!
This doesn't make sense? Its an AC input with an internal PSU.
edit: Leaving note, I agree with the benefit of bypassing LDO's and I haven't had a chance to experiment with different power supplies behind UC's yet to understand any impacts. Although the Ian battery board is a better performer with a regulated 19v linear power supply (even though it shouldn't be as it's floating). However, my USB signature (loaded with LDO's) responded really well to the 5v UC conditioners - although I haven't tried the conditionerPI. So I wouldn't suggest people are wasting their time, that may be a bit strong
Last edited:
You can't say that, it's not correct,and i refrain from saying more,the development of Supercapacitor power supplies took months of work and these are products that we find Nowhere else.
If you read every post of Ian carefully, you will find that he mentioned that the power quality of LiFePO4 or supercapacitors far exceeds that of any LDO.
The commercial brands DAC does not support you to easily remove the LDO. Therefore, in the commercial brands DAC, the power quality is completely limited by the LDO. Even if you use LiFePO4 or supercapacitors to provide power, when the current enters the internal circuit of DAC, it will still use the LDO, which is a waste of work.
In contrast, if you look carefully at the photos Ian once posted, you will find that the LDO board on the BUFFALO-IIISE PRO only plugs in 1.3V LDO, and the other 3.3~3.6V LDOs are not plugged in, but power input and input terminals are directly connected by wires to avoid LDO. All the digital circuit boards is designed by Ian can automatically sense, no need to pull out the LDO board.
edit: Leaving note, I agree with the benefit of bypassing LDO's and I haven't had a chance to experiment with different power supplies behind UC's yet to understand any impacts. Although the Ian battery board is a better performer with a regulated 19v linear power supply (even though it shouldn't be as it's floating). However, my USB signature (loaded with LDO's) responded really well to the 5v UC conditioners - although I haven't tried the conditionerPI. So I wouldn't suggest people are wasting their time, that may be a bit strong
I know my post will be more controversial, but to emphasize that I have always supported Ian, not with my lips, but with my money to support. In order to prove this, I specially posted transaction records from 2017 to 2020. I support Ian, but I am not blindly supporting him. If something is really bad or impractical, I will say it.
I also believe that 5v UC conditioners is a good 5V power supply, but this does not mean that conditionerPI is meaningful, especially since Ian's IsolatorPi has been used in raspberry pi. In the same way, if there are already UC conditioners, there is no need to buy the more expensive LinearPi, and the generally cheap LT3045 LDO can already be used to charge the UC conditioners.
I have a friend who specializes in researching and selling all kinds of linear power supplies and sells them all over the world. I asked him about LiFePO4 battery power supply a long time ago. My friend told me based on actual test experience that the output terminal of LiFePO4 still must use LDO or capacitor, because the output of LiFePO4 will have a similar ripple phenomenon. A long time ago, Ian was still developing LiFePO4 battery power supply. I already proposed it, but Ian refused to listen. Many people also think I am heresy. Okay, it's a waste of your own money anyway. As a result, Ian finally admitted that LiFePO4 needed capacitors to solve the problem.
However, Ian’s solution to the problem is to connect the output terminal of LiFePO4 to a supercapacitor which is actually a battery... Why not just use a supercapacitor? Superfluous is just talking about this situation.
Last edited: