• These commercial threads are for private transactions. diyAudio.com provides these forums for the convenience of our members, but makes no warranty nor assumes any responsibility. We do not vet any members, use of this facility is at your own risk. Customers can post any issues in those threads as long as it is done in a civil manner. All diyAudio rules about conduct apply and will be enforced.

Shanti Dual LPS 5V/3A , 5V/1.5A

Earthing...for us was very important.


We have done a lot of research on earthing using USbridge , Revolution Dac.


Simply put , earthing is one more step in removing all noise from a PCB


On Katana (uController) you have a hole next to DC input . If you connect the earth of Shanti to Katana ground (the hole is connected to Gnd), some noise will go away




Please expect all our new products to have very sophisticated earth connections (a lot of taught went into how to connect earth to analog Gnd PCB , of course not direct)





I will check for live wire orientation , but from what I know , its more important in smps less on linear transformer .


Do you have a picture from the Katana (uController) where to connect the earth from the Shanti?
 
Hello, I've received my Shanti and second the earlier comments that it is an impressive unit. I have used it with the 3.0 amp section powering my RPi 3B+ and Boss DAC 1.0, which works very nicely. I now want to put my Isolator 1.0 into the stack and have a couple of questions.

1) Shold the Isolator be set to Slave or Master mode? I can construct stories that justify both configurations, but want to do it right without taking a chance on smoke.

2) With this stack, where is the best place to attach the ground wire to Shanti?

Thanks for your help,

Skip
 
1) Shold the Isolator be set to Slave or Master mode? I can construct stories that justify both configurations, but want to do it right without taking a chance on smoke.

It's about I2S master/slave DACs.

Boss and Katana will run in master mode.
E.g. Piano21 or Audiophonics iSabre would run in slave mode.


2) With this stack, where is the best place to attach the ground wire to Shanti?


With an "isolator" there shouldn't be a common ground for best performance.
That's why Allo suggests to use their battery pack.

However.

I'd say. If you consider the Shanti ground terminal your starground point try to run a ground wire to all devices attached to avoid ground loops.
 
Hi.

One more comment to the Shanti "delayed & sloping power down" behavior.

Let's assume you power your Pi with Shanti.
In this case you can't simply turn off Shanti and pull the SD-card right away.
It might corrupt your SD card. Be careful!

The safest way would be to pull the plug. Those of us who hardwire the wires obviously have an issue with that method.
As an alternative you simply wait until the PI power LED is off. But that'll take a while.

I'm not sure yet if that "sloping" turn-off behavior is causing other issues.

We need to keep an eye on that.
 
Last edited:
Just been talking about "other issues" !?!?

Did anybody notice?

When turning off Shanti on its power switch I can hear a very strange sound (high-pitched-sucker-sound ;) ) on the Katana side.
I guess it's related to the Katana input buffer caps / filter. Not sure what's happening there.

For sure it doesn't sound like healthy or normal behavior.

Perhaps Allo should quickly check if there's some weird interaction ongoing.


The more I think about it, the more I get the feeling that a relay on the Shanti output is needed.
 
Last edited:
I am back in the office after some corsican sun .:)


OK so lets start.


Yes we are aware of a sound (rather nasty) when turning off completely shanti using the AC switch . Our team is still checking the exact cause (seems its something with DC/DC convertor and how we feed it). We will see how we can release a fix.


Meanwhile , you can either disable the analog stage (using the SW on uController board) or just remove the DC cables (instead of switching from AC side)


We tested with usbridge , digiones and revolution , none exibit this issue (of course the first 2 are digital outputs)


Yes USBridge/Audiopile SBC + Sig has only one "clean input" . Dirty input is direct from USbridge Sig/Audiophile SBC , DC input
 
Welcome back. We could have met. You on Corsica and me on Sardegna. :D

Just to make it clear.

The nasty sound from the speakers I reported first is something else then the nasty high-pitched-sucker-sound I hear right from the physical Katana "caps" right on the board.

And even if you do not experience these rough conditions on the other boards, it doesn't mean that these boards won't get tortured in a similar way.


And please. Consider to look at a switch for the Shanti 2.0 output that simply takes out the DC output-rails at a certain voltage or when the device gets turned off.

Keep also in mind. Not all people turn the devices off at the device power switch any more.
A 2019 power supply should allow to be switched off remotely (--> home automation <-- ). E.g. I do it with remote controlled power sockets.

Of course really nice would be a PS with some intelligence inside. ;) By e.g. using GPIO/serial IF, power rails could be enabled/disabled remotely.
 
Last edited:
The analog stage of any DAC is very sensitive to a imbalance between positive and negative rail and Shanti is gracefully going from 5V to 0 on the supercaps and while shutting down we measured an imbalance on DC/DC output stage.


On any digital system there is no such imbalance and LDOs cut off voltages output once it drops to a certain level



From all testing data , its safe to turn off AC switch on any digital system , if you have a analog system , its best to remove the DC connection (or in case of Katana , just use the SW on uController before switching off the AC)
 
...its best to remove the DC connection (or in case of Katana , just use the SW on uController before switching off the AC)

You gotta be kidding! :rolleyes:

You can do it this way on a lab workbench. BUT. You can't run a realworld system like this. Shanti is used as primary PS. You turn your system on/off via Shanti.
And not to forget. You'll have analog, digital and potentially a mix of both attached to it at the same time. Not to forget. Many "digital" also comes with analog rails.
E.g. Your Katana ucontroller. It's digital and provides the analog rails.
And it's the big cap on the "digital" ucontroller making these nasty sounds.

Cutting off the Shanti DC rails as soon as the AC goes done is what I've been suggesting as a potential solution to the issue.
That's probably the best and perhaps even the only solution. You simply can't tell people what devices to attach to Shanti.


However. Looking forward to hear more about your analysis. Let's hope there's a better solution to the subject. :xfingers: :D
 
Last edited:
Hi Chris


Nirvana transformer was found to heat much more than expected using thermal camera . Wurth (thransformer manufacturer) is checking the issue in HK , but due to certain events is taking longer than expected. I will update as soon as possible


Usbridge Sig/ Audiphile SBC is on track. New boxes will be sent Friday (one alu color and second black so we can judge the look and feel) . PCB is already ordered we should get it in max 10 days . I would expect USBridge to ship before Sep 10 (just an estimation)
 
Soundcheck



I think that you are seeing problems where there is none .



I understand you want to modify Shanti to solder direct wire to PCB and for you removing DC cable will not work . However most customers will have no issue doing just that


In your case , I am suggesting to use the uController sw and it will work fine for the few times that you want to switch off Katana. (but it will take you 10seconds more)



Your idea to use some type of mcu will not work in a ultra low noise environment .
 
Hi Chris


Nirvana transformer was found to heat much more than expected using thermal camera . Wurth (thransformer manufacturer) is checking the issue in HK , but due to certain events is taking longer than expected. I will update as soon as possible


Usbridge Sig/ Audiphile SBC is on track. New boxes will be sent Friday (one alu color and second black so we can judge the look and feel) . PCB is already ordered we should get it in max 10 days . I would expect USBridge to ship before Sep 10 (just an estimation)

Is it a question of replacing the transformer or using a larger ventilated case to allow for heat dissipation?