TASCAM UH-7000
May 4, 2016 at 2:01 PM Post #16 of 37
    Use speakers, all recordings except binaural are meant for speakers, you'll never get correct sound stage using headphones for example.
The Tascam is the most accurate DAC I've came across.
Most high-end DAC is the opposite of accurate, they try to make things sound better than they are, I had a couple in the $4000+ range and sold them all.

 
Hi this is a great news !!!  thanks a lot. 
I should receive mine next week from Thomann.de   
What about the mentioned noise issues  ?  
have you experienced any noise/listening fatigue ?
I bought it mainly because i have a AD/DA converter ... actually 2,  
one very cheap from Behringer src2496 and another  Apogee Rosetta 200, both with AES/EBU in/out. 
So i badly needed an interface to connect the converters to the pc.   And i wanted to stay with balanced digital connections. 
But if i could do just with the Tascam ... maybe it is even better !
I have also heard that the HP out is quite good ?
It is also quite nice looking.  
I am getting excited 
Thanks a lot, gino 
 
May 5, 2016 at 11:05 AM Post #17 of 37
  Use speakers, all recordings except binaural are meant for speakers, you'll never get correct sound stage using headphones for example. The Tascam is the most accurate DAC I've came across. Most high-end DAC is the opposite of accurate, they try to make things sound better than they are, I had a couple in the $4000+ range and sold them all.

May I ask what kind of equipment you had in $4k range you sold ? just wondering what you compared the UH7000 with as listening experience.
 
May 5, 2016 at 1:31 PM Post #18 of 37
Hi Guys !
 
a very interesting and telling test here
 
http://www.computeraudiophile.com/f6-dac-digital-analog-conversion/tascam-uh-7000-power-supply-noise-28570/#post542090
 
The poster is a genius !
it seems that the noise comes from the smps not enough shielded  !!! 
eek.gif

Ripple wise instead the stock smps is quite excellent.  Very good indeed.
So the solution should be a relocation of the smps.
Another solution could be some kind of additional shielding on the parts more likely to radiate noise.
Does anyone knows of products suitable for the task ?
 
Ok ! i have an idea ...
 
i see there is already a barrier inside separating smps and circuits board
In case the idea is to cut two sheets of adesive copper foil and stick it to the barrier
and maybe another one to stick on the internal side of the covering lid ?
Above the copper foils adesive silicon sheet could be used to avoid short circuit with the copper foil.
I am amazed how cheap this interface is in USA considering the good overall quality
eek.gif

Luckily in Europe is much more expensive
wink.gif

 
Thanks a lot indeed.
Kind regards,
gino

 
 
 
 

 
Jun 24, 2016 at 3:39 PM Post #19 of 37

dam, this unit really got things DIY mode!,, gonna do some tests with shielding, but I doubt shielding between the two sides will help ufortunately. Its the auto switching power supply thats the issue. As you saw in the other thread on computeraudiophile using a linear power supply fixes the issue completely.
 
Another thing I noticed, is that once I pop the lid of the unit, and I disconnect some of the inputs and front panel, I get noise, once the hood is on, noise is less. and once I put my whole hand on top with hood open, noise is reduced. It made me think of a ground issue with the unit mabye something isnt grounded as it should ? So I tested it in my bathroom with proper grounded output, and I got exact same result.
 
Jun 24, 2016 at 4:32 PM Post #20 of 37
dam, this unit really got things DIY mode!,, gonna do some tests with shielding, but I doubt shielding between the two sides will help ufortunately.  

 
Hi !  yes i think you are very right.  
I tried moving the signal cables away from the smps but this had no effect on the noise floor.  The 20dB bumps are always there.
I am using the spectrum analyzer tool in the Arta software for my testing. I look for any change in the noise spectrum.  
So my guess is that is not a shielding problem.  But I have not extracted and tested the stock psu outside the case.
I have also disconnected all the signal cables going to the board.  Again the bumps above the noise floor persist.  The noise in coming from the smps. 
 
Its the auto switching power supply thats the issue. As you saw in the other thread on computeraudiophile using a linear power supply fixes the issue completely.  

 
Again i think you are right.  The issue is the smps injecting noise in the circuit.  
I have put a free dc connector to power the circuit board with an external psu.
 

 
Using a refurbished Dell smps (in the picture below)
 

 
this is what i got for the noise floor with mic volume to minimum using Arta software (a really great sw).
 

 
Much better.  Quite low and clean i would say.  
My take is that this unit is fundamentally great.  
I have tested other interfaces and they have all higher noise floor.  I do not know anything about distortion by the way.
 
For me extremely low noise and distortion are the keys to achieve very good sound. 
 
And from what i read with great AD and DA sections too. 
I tried many other smps (all very cheap) and also a bench lab power supply.   Almost all the psu give very similar results.
The ideal would be to find a new smps similar in size to the stock one and replace it.  
The unit is very very good but i really think they have slipped on the smps.  
 
Sadly i am not able to perform any kind of troubleshooting on the original smps to try to fix it.
 
But almost all 12VDC smps should work fine.  I do not understand why they have not changed the smps in Tascam.  This unit is intended for the professional sector.
Even if the noise is low in absolute term it is not nice. 
 Another thing I noticed, is that once I pop the lid of the unit, and I disconnect some of the inputs and front panel, I get noise, once the hood is on, noise is less. and once I put my whole hand on top with hood open, noise is reduced.
It made me think of a ground issue with the unit mabye something isnt grounded as it should ? So I tested it in my bathroom with proper grounded output, and I got exact same result.

 
With the external smps i do not hear any noise using it.  It is very quiet.
 

 
 
as the space inside is really tight i think i will put a dc socket on the rear panel and use one external smps.
Unfortunately i will not be able to switch the unit on with the frontal button. 
As i said the only one with noise issues is the original one  
frown.gif

A little more attention would have avoided any issue. 
A linear psu is not needed at all.   Just one decent power brick like one used for display or laptop with 12VDC output and some 2 Ampere.
This is what i have used to get the graph above. Not high end at all as you can see.   And no spikes at all above the noise floor.
I will post the final solution.   
Let me know your solution.
Thanks a lot again for the very valuable testing and advice. 
Kind regards,  gino  
 
Jun 24, 2016 at 4:39 PM Post #21 of 37
Thanks for a great post.
Did you reduce voltage to 11-11.3 or did you stay with 12v 2a ? just wondering if Ill do this test tonight.. hehe.. or if I need something to reduce voltage just slightly.
 
If you remove the original SMPS then you can use the power button and rewire the 12v signal.
 
Jun 24, 2016 at 4:52 PM Post #22 of 37
  Thanks for a great post.
Did you reduce voltage to 11-11.3 or did you stay with 12v 2a ? just wondering if Ill do this test tonight.. hehe.. or if I need something to reduce voltage just slightly. 

 
Thanks to you !  no i did not reduce the voltage.  
I just tried different 12V units and i did not get any apparent issue. 
In the specs the consumption is rated 15W that gives a little more than 1A of current draw.  
I think that any 2-3A  smps would be more than adequate and on the safe side. 
I am waiting other models i ordered (too many actually)
but i have already tried at least 6 o 7 different smps all with very similar and good results. No spikes at all. 
Maybe a good choice for long term reliability could be to buy something "genuine" from a good brand like Sony, Lenovo, Asus and so on.  
Something like this one 
 
http://www.ebay.com/itm/Genuine-Original-Sony-AC-Adapter-Power-Supply-Compatible-AC-ES1230K-12V-3A-36W-/301809046203?hash=item46453892bb:g:Ne4AAOSwD2pXF0Mt
 

 
As i said the main constraint is the little space available inside.  
Actually i have one 3A very small unit that could fit in.  
For unit up to 3A the size can be quite small.
Again you could find some smps small enough to fit in the box and be able to use the unit as intended from the frontal power button. 
And if you use a power brick you will have also already shielding that there was not in the original unit (another very strange design decision to use an open frame smps).
I am still thinking honestly about the best solution. 
Good to know that the stock psu is a problem.  
Let me know if you do anything.
Thanks a lot again.
Kind regards,  gino  
 
Jun 24, 2016 at 5:02 PM Post #23 of 37
Thats great. Ill do my tests for sure.. off we go with old power supply.. gonna rip that part straight out. hehehe.. I do notice that the old power supply is somehow same similar size as those power adapters we got laying around from old equipment. mabye one of them might fit! or purchase a china power supply that looks similar and has same bracket holes to reattach it.. these small power supplies from china cost like 20-30 dollars max.
 
Jun 24, 2016 at 5:09 PM Post #24 of 37
Thats great. Ill do my tests for sure.. off we go with old power supply.. gonna rip that part straight out. hehehe..
I do notice that the old power supply is somehow same similar size as those power adapters we got laying around from old equipment.  mabye one of them might fit!
or purchase a china power supply that looks similar and has same bracket holes to reattach it.. 
these small power supplies from china cost like 20-30 dollars max.

 
 
U P D A T E 
 
Hi again !
just out of curiosity i have opened some generic chinese smps ... of this type 
 

i am not an expert but i got a bad feeling.  The shield for instance is usually missing.
Cheap and very few parts ... uhmmm.   No good. 
 
Instead this in the picture below is the 12V/3A power brick by Dell i have used and mentioned above.
Much much better execution.   
 

 
Problem is the size.  
Every time i look inside the Tascam i get scared by the tiny space available for the psu.  
frown.gif

I would prefer a different unit size ... even a 1U standard rack case but with much more space for the power supply section.
I see many professional interfaces having standard sizes. 
 
Conclusion 
 
i am more and more thinking to go for the external psu solution.   
rolleyes.gif

This will allow me to try any 12V power supply around, switching or linear 
regular_smile .gif

My feeling is that this interface is a very sensitive tool for power supply quality. 
Moreover i am a dog at diy and the risk to damage something is just around the corner.
I am waiting for some parts i have ordered. 
I will post the result in any case 
redface.gif

Thanks a lot again
Kind regards,  gino  
 
P.S.  anyway ...  with a cheap chinese switching bench power supply
 
i see 450mA of current draw at 12VDC    
rolleyes.gif

 
always with inputs and outputs open and using only the headphones.
So i guess a 12VDC/1A psu should be enough to power the unit. 
 
In practical terms what i am going to do is to use the Tascam  ( it samples at 192kHz providing spectra up to 96kHz )  
with the Arta spectrum analyzer and see the noise using different power supplies.
Clearly the very interesting thing would be to check the noise coming directly from different psu to measure their ripple.
This is very very difficult to do wideband when noise is very low. 
But it should be done to get a precise measure of the power supply quality.
Then when i will have found a very silent psu i will be done.
As an example i have received today a new linear psu bought on ebay
This is what i get from Arta 
 

 
i really do not know if the graph will vary with the break-in of the psu 
But this is what i have in mind.
I hear often people talking about noise, sound of different psu.
my only guess is that the lower the noise floor the better
Like also distortion that i am not able to measure, for now.   
I do not know of any other way to test a psu. 
Actually it is the only test that i can do with what i have at hand   
redface.gif
  
Thanks a lot again for your very valuable advice and testing.
Your experience is what got me started.
Kind regards, gino
 
Mar 8, 2017 at 7:12 PM Post #25 of 37

 
I went and bought a new PSU for my Tascam unit. This noise floor is from my Schiit Yggrasil into the UH7000 to xlr3 input. No bumps whatsoever. The unit sounds amazing now compared to what it did sound like. Its a MeamWell switching PSU inside a metal cabinet. 12v, 2.5a 30w. It sits inside the Tascam cabinet just fine on top of the other electronics.
 
I suspect the bass part <200hz to be the power supply of my Yggdrasil, I have seen other yggy noise floors showing exactly the same rumbling bass in the noise floor. The noise floor is identical if I feed my Yggy with AES from a Gustard U12 or on usb with or without Intona usb galvanized unit.
 
I wish Uptone LPS-1 supported 12v, it only goes to 9v! then I could try the LPS-1 to feed the Tascam just for fun.
 
This was a big learning for me. Clean power is damn important. I cant believe Tascam equipped the UH7000 with that crappy PSU card that radiates noise into the circuitry of the board.
 

I bent the black plastic in 2, so it holds the center part of the PSU above the UH7000 electronics, and removed the old PSU, and rewired everything to the new PSU. the power connectors to the UH7000 board is red+white+   white- white-. Under the PSU I bought a thin plastic film that I glued under the PSU so just in case it would move a bit so nothing touch the cabinet. I super glued the PSU to the frame where it touches the frame.
 
As a AES device feeding my Yggdrasil it really is on par with USB input. I just cant tell the difference now. Before PSU change I could easily hear the difference.
As a DAC, the UH-7000 unit just went from a muddy soundstage to open up and alot of details appeared. For a few minutes I had to check that I had all cables correct playing from the UH-7000 as it was so damn good soundstage that I thought the sound came from my Yggdrasil. But after a few songs it was easy to tell the difference. But were not comparing a 2-300 dollar ADC/DAC against the Yggdrasil.
 
Now the fun part can start, measuring things.. Ive already done some square signals today trying to find a way to measure the Intona USB galvanized unit. I have some interesting recordings for sure.
 
I am waiting for a chinese linear PSU to appear in a week or two. I will try that PSU too, just to see if the bass section calms down a bit or not.
 
Mar 9, 2017 at 5:17 AM Post #26 of 37
Originally Posted by manpowre /img/forum/go_quote.gif
 
 
I went and bought a new PSU for my Tascam unit. This noise floor is from my Schiit Yggrasil into the UH7000 to xlr3 input. No bumps whatsoever.
The unit sounds amazing now compared to what it did sound like. Its a MeamWell switching PSU inside a metal cabinet. 12v, 2.5a 30w. It sits inside the Tascam cabinet just fine on top of the other electronics.
I suspect the bass part <200hz to be the power supply of my Yggdrasil, I have seen other yggy noise floors showing exactly the same rumbling bass in the noise floor.
The noise floor is identical if I feed my Yggy with AES from a Gustard U12 or on usb with or without Intona usb galvanized unit.
I wish Uptone LPS-1 supported 12v, it only goes to 9v! then I could try the LPS-1 to feed the Tascam just for fun.
This was a big learning for me. Clean power is damn important.  

 
Thank you very much indeed again for disclosing very interesting results.
I am just a schoolboy here and not educated also. 
In very trivial terms i think that the lower the noise and distortion the better.  And i tend to believe more in instruments than in my ears.
Mostly because lately i had the confirmation after a hearing test that they are not perfect but quite limited. 
It is true that a circuit has some level of power supply noise rejection.  
But with a very "silent" power supply there is almost no noise to reject.
As i said above even just decent 12V smps for pc give very flat and low noise spectra with Arta (with mic gain to min of course).
Below some point i think that the incremental improvement is negligible.  I paid those smps from 15-20 USD on ebay.com.
 
 
I cant believe Tascam equipped the UH7000 with that crappy PSU card that radiates noise into the circuitry of the board. 

 
I agree 100%.   Strange that they deliver a unit with these trivial glitches.  I have the feeling that they do not want to risk to make a very good unit for too little money.  It could challenge the more expensive units.
I am not an expert but i suspect that it could be improved with some more filtering and better grounding ? i saw something somewhere but i do not remember where. 
I see many parts missing on the boards like caps for instance.  Again just a 15 USD smps off the shelf can be better of the stock unit. 
 
 

I bent the black plastic in 2, so it holds the center part of the PSU above the UH7000 electronics, and removed the old PSU, and rewired everything to the new PSU.
the power connectors to the UH7000 board is red+white+   white- white-. Under the PSU I bought a thin plastic film that I glued under the PSU so just in case it would move a bit so nothing touch the cabinet.
I super glued the PSU to the frame where it touches the frame.  

 
I find it a little too big honestly.  Probably something smaller and decent quality will provide a very comparable noise floor.  
Below a certain level the differences are negligible i guess.  When you have a noise floor of -120dB flat … well it looks already quite low to me. 
Instead with the original psu the bumps were extending even 30dB above the noise floor.   Quite evident.
 
As a AES device feeding my Yggdrasil it really is on par with USB input. I just cant tell the difference now. Before PSU change I could easily hear the difference.
As a DAC, the UH-7000 unit just went from a muddy soundstage to open up and alot of details appeared. For a few minutes I had to check that I had all cables correct playing from the UH-7000 as it was so damn good soundstage that I thought the sound came from my Yggdrasil. But after a few songs it was easy to tell the difference. But were not comparing a 2-300 dollar ADC/DAC against the Yggdrasil.
Now the fun part can start, measuring things.. Ive already done some square signals today trying to find a way to measure the Intona USB galvanized unit.

 
A very important aspect, i think,  is the Total Harmonic Distortion.  I would like to be able to measure it one day.  With the loopback test it should be possible to test both the AD and DA sections of the unit with just one test.   
I think that THD is more influenced by the circuit design than by the psu noise.
I mean that an interface can have very low noise but also high THD and this is not good at all.
When I see some THD spectra I am amazed by how dirty they are with so many harmonics not present in the original testing signal.  This is not high fidelity at all.
 
I have some interesting recordings for sure.
I am waiting for a chinese linear PSU to appear in a week or two. I will try that PSU too, just to see if the bass section calms down a bit or not.

 
Have you noticed any difference using rca or xlr connections ? balanced connections should give better noise performance. 
Thank you very much again,
Kind regards, gino
 
Mar 9, 2017 at 7:14 PM Post #27 of 37
I wrote in the uh-7000 thread at computeraudiphile too.. 
 
But Houston <i have a problem.
 
The Meanwell PSU is a switching one.. here we have 50hz 230V AC current. And I get ripples with square wave every 5.5ms or so. my friend told me its either ground between the 2 units issue, or PSU ripple due to switching! The sales guy told me it is linear.. but I checked the product page on meanwells page. its switching.
 
So now I have to wait for a linear one.. I ordered 2 of them super cheap from china. And I cant measure anything more until I get this ripple effect fixed too!
 
I intend to start measure AES/coax/spdif/usb with and without galvanization and find ways to measure sound through my high end dac, especially to prove that galvanized USB improves !! I have seen improvements even with ripple effect that AES input on my yggdrasil is way better than the spdif for instance. 
 
The noise floor between the pictures were probably just me not having sound levels at 0 when I measured the original PSU. Noise floor is at -145db steady.
 
Last few days Ive been working on calibrating things, and started to adjust volume levels and test tones to same levels as stereophiles review of the DAC I own. And I did that, but then I measured the ripple effect. The good thing is that the UH-7000 can recreate almost perfect samples of square sound at 1khz until the switching mode kicks in every 5-6ms.
 
Until I get the linear power supply I will create methods for different measurements and learn more.
 
I did isolate this to the tascam unit only. so its not my high end audio system.
 
Here u can see the ripple effect:

 
I havent tried line in vs xlr3, I only work with XLR3 throughout my system. But on my list is to try to measure line in too. I have rightnark 6.3, and it basically tells me that there is too much distortion = ripple effect. I just didnt know 2-3 days ago what were making the distortion, now I know.
 
Mar 10, 2017 at 1:44 AM Post #28 of 37
Originally Posted by manpowre /img/forum/go_quote.gif...   So now I have to wait for a linear one.. I ordered 2 of them super cheap from china...

 
Hi ! and thanks again. 
I see that you have gone much deeper in the issue than myself
redface.gif
   i am always very superficial
Please post your findings here.  I am extremely interested in this thread. 
May i ask which psu have you ordered ? 
rolleyes.gif

i see many interesting offers of linear psus on ebay but never accompained by noise spectra (of course.  Not a very common test).
 
By the way, speaking of high Hz noise i wonder if noise out of the audio band (i.e. above 20kHz) can have a negative effect on a digital unit, like causing jitter or so on.
I can see easily the noise up to 30-40kHz with the beautiful Arta software ... but not beyond that.   (they told me that it depends on the sampling frquency of the interface.  Setting it to 96kHz i can see up to 96/2= 48kHz )
Thanks a lot again
Kind regards
gino
smily_headphones1.gif
 
 
Mar 10, 2017 at 11:32 AM Post #29 of 37
   
Hi ! and thanks again. 
I see that you have gone much deeper in the issue than myself
redface.gif
   i am always very superficial
Please post your findings here.  I am extremely interested in this thread. 
May i ask which psu have you ordered ? 
rolleyes.gif

i see many interesting offers of linear psus on ebay but never accompained by noise spectra (of course.  Not a very common test).
 
By the way, speaking of high Hz noise i wonder if noise out of the audio band (i.e. above 20kHz) can have a negative effect on a digital unit, like causing jitter or so on.
I can see easily the noise up to 30-40kHz with the beautiful Arta software ... but not beyond that.   (they told me that it depends on the sampling frquency of the interface.  Setting it to 96kHz i can see up to 96/2= 48kHz )
Thanks a lot again
Kind regards
gino
smily_headphones1.gif
 

 
Even though the switching power supply is creating 50hz ripples in my Tascam unit, the audio improvements were so much higher and got way more accurate with the switching PSU. The original square samples I measured before switching the PSU were all broken down and was the reason this unit were on the shelf for half a year. Mabye one day Ill see if I can mount the original PSU back and do exact same square tests to see the actual difference with a calibrated system (only difference will be switching between the PSU's while all software settings remain the same).
 
I bought this type as it looks similar to the ones the guys in the computeraudiophile thread mounted inside the Tascam unit:
http://www.ebay.com/itm/122120198083?ul_noapp=true
 
But I see the price has increased somehow I got it cheaper than that. And these things are 3 times the price here in Norway. Also I know there is possible to purchase DC-DC capacitor filters to clean up the power even more. I also intend to do that once the linear PSU's arrive and try to see if I can measure the difference.
 
Its a hobby afterall :)
 
Mar 10, 2017 at 11:35 AM Post #30 of 37
One thing that I have been thinking of, and that is to power the unit with a chargeable battery pack of 12v. But in that case I need a linear 12v dc- 12v dc regulator that is not of the switcheable type!
 

Users who are viewing this thread

Back
Top