LKS Audio MH-DA003
Feb 24, 2017 at 4:25 PM Post #511 of 838
  Hi b0bb,
 
Thank you very much for your help.
I will immediately place order for the PS board, the Antek transformer and all the parts to be swapped.Some

Some of the other cheaper Ebay listings do not have the Crystek clocks, you will have to spend another 75USD if you have to put them in after purchase.
 
Feb 26, 2017 at 7:12 AM Post #512 of 838
 
 
I think you should discontinue the use of the V5D as it is not able to handle a common mode voltage of 0.004V
Burson did not supply this info on the V5 datasheet.

So are you suggest to change output opa to NewClassD DEXA Special Edition?
What will be optimal DC value beetween 1-3 and 2-3 RCA pins?
 
I have problem with remove PCB from the chassis. This small lamina in the AES/EBU socket make it impossible.
I have tried to remove this lamina form the AES/EBU socket with no effect.

How to deal with it?
 
Feb 26, 2017 at 4:43 PM Post #513 of 838
  So are you suggest to change output opa to NewClassD DEXA Special Edition?
What will be optimal DC value beetween 1-3 and 2-3 RCA pins?
 
I have problem with remove PCB from the chassis. This small lamina in the AES/EBU socket make it impossible.
I have tried to remove this lamina form the AES/EBU socket with no effect.

How to deal with it?

 
I have no solid recommendation for the OPA going to the RCA sockets. The DC offset adjustment range did not change with the DEXA in the unbalanced converter position going to the RCA sockets.
 
I have only used the DEXA for a few weeks, like the Burson, the documentation is missing many OPA parameters. Burson started failing after 4 months, the Burson in my later photos are the replacement set.
 
In that position I have also tried the NJR MUSES01, Sparkos SS3602 and the Ticha994. The Ticha is the best but an expensive option, the DEXA is the second best.
 
I paid 40USD on a sale, I will not pay the normal asking price.
 
The back panel can remain attached to the PCB, I rework the board with the back panel still attached, it separates from the chassis cleanly once you remove the power socket and power switch.
 
Feb 26, 2017 at 9:15 PM Post #514 of 838
Hi b0bb,
 
Can you reconfirm that all the parts as listed in the Bill of Materials for Amanero board powersupply mod in your post # 114 are still good to mod with the Amanero USB PS from Shenzhen Audio and sold on Ebay as indicated in your post # 507?
Thank you
 
Feb 26, 2017 at 10:13 PM Post #515 of 838
  Hi b0bb,
 
Can you reconfirm that all the parts as listed in the Bill of Materials for Amanero board powersupply mod in your post # 114 are still good to mod with the Amanero USB PS from Shenzhen Audio and sold on Ebay as indicated in your post # 507?
Thank you


Post a few pictures after you receive the card, the Ebay vendors sometimes change the board layout.
 
Do not mod until the card is verified as working in the DAC.
 
Feb 27, 2017 at 9:44 PM Post #516 of 838
Hi b0bb,
 
I will pur
 
Post a few pictures after you receive the card, the Ebay vendors sometimes change the board layout.
 
Do not mod until the card is verified as working in the DAC.


Hi b0bb,
 
I will post the pictures upon receipt of the card as per your recommendation.
Thank you for your valuable hep.
 
Mar 1, 2017 at 1:52 AM Post #517 of 838
   
A couple of comments
1)The left one goes to the DAC, the other two are optional.
 
2)The heatsinks have to be removed to properly mount the new regulator and to check for short circuits.
 
3)Get the LM317 version not the 78XX version. The LM317 version did not exist when I did my mod, so there is no need to add the wires like I did
http://tekdevice.com/chapter2/index.php?route=product/product&path=25_54&product_id=142
 
4)The voltage adjustment resistors and bypass cap have to be removed from the motherboard. Bypass cap terminals shorted

 
 
5)The Tekdevice board needs a 10uF tantalum capacitor soldered between  the IN pin on the regulator and ground, a suitable point is the bypass cap location, the one with the soldered jumper.
I mounted the the 10uF cap on the other side of the board.
 
NOTE: +ve electrode to the IN terminal and -ve to ground, do not connect the tantalum cap in reverse
 
Here is a suitable cap
http://www.mouser.com/ProductDetail/AVX/TAP106K020SCS/?qs=sGAEpiMZZMtZ1n0r9vR22d%252b8XmbM9QM8iVuE3tFomWE%3d
 
6)Add heat sink paste between the regulator and the heatsink. Do not use the stuff  containing metal, like like Arctic silver.
I use Dow 340
http://www.ebay.com/itm/DOW-CORNING-340-DC340-Silicone-Heat-Sink-Compound-350-F-/280620305862
 
7)The layout of the current Tekdevice boards are different, the mounting hole is further up, you may need to drill a new hole in the heat sink

 
8)This is a newer design that might be better from a mounting consideration but there is additional work to be done

http://www.ebay.com/itm/TPS7A4700-Ultralow-noise-4-VRMS-low-dropout-linear-regulators-3-5PIN-1-4V-20-5V-/252348749037?hash=item3ac12854ed:g:rRMAAOSwJkJWjBAt

Hi b0bb,
 
After the installation of Tekdevice board is completed, one needs to reinstall the output filter cap. Am I right?
Thanking you
 
Mar 1, 2017 at 2:10 AM Post #518 of 838
  Latest addition, this is the NewClassD DEXA Special Edition
http://www.newclassd.com/index.php?page=123&hv=1
 
This is a high bias (20mA) Class A opamp

 
This has replaced the Burson V5, I was using previously in the balanced to single-ended converter
 
The soundstage got considerably wider and the transients (snare drum, kickdrum, hihat, bells, triangles)  were clearer.
 
Bass output extended down by almost 1/2 octave, the double bass rendition was deep and tight, electric bass guitar sounds very much like a very big stringed instrument compared to the indistinct blubbery sound that was the previous norm.
 
With the DEXA the difference between the unbalanced and balanced outputs was non existent.
 
In terms of the opamps I have used in this role, the pecking order in terms of decreasing performance
 
Dexa SE, Burson V5, Sparkos SS3602, NJR Muses 01, Stock Opamp.
 
The DEXA was about the equivalent of the Ticha 994 in this role.
 
Further closeup

 
The 3 LEDs are quite pretty, picture without flash.

 
 
The last 2 pictures show the DEXA in the I/V role.
 
Performance was mediocre at best, despite the high Class A like bias.
 
Compared to the Ticha 994 the soundstage underwent a dramatic collapse with a great loss in transparency and the absence of the fast transients.
 
The flip side was a much warmer presentation with emphasis on the midrange, particularly in the vocal ranges.
 
Does not look like the DEXA can replace the Ticha 994.

Hi b0bb,
 
Is the latest version of Dexa opamp better in terms of sonic performance compared to the former version as offered by partsconnexion? the price is double thus is it worth to get the latest version?
Thanking you for your kind explanation.
 
Mar 2, 2017 at 12:34 AM Post #519 of 838
  Hi b0bb,
 
Is the latest version of Dexa opamp better in terms of sonic performance compared to the former version as offered by partsconnexion? the price is double thus is it worth to get the latest version?
Thanking you for your kind explanation.


This is the only version I have tried, don't know about the older version, have not quite decided if I want to keep this DEXA in there long term.
 
Mar 2, 2017 at 1:49 AM Post #520 of 838


​somebody trial this LKS stuff ???
 
Mar 3, 2017 at 4:26 AM Post #521 of 838

 

 
lt3042,tps7a4700 on Taobao,42,35 Rmb
 
Mar 5, 2017 at 4:47 AM Post #522 of 838
 
 

 
lt3042,tps7a4700 on Taobao,42,35 Rmb

 
Very nice find!
 
 
Can you please provide a link of those?
 
Cheers 
beerchug.gif

 
Alex
 
Mar 5, 2017 at 5:50 PM Post #523 of 838
   

 
The pins on the XO socket are not coming out on the other side, if they are too short you may not have full contact on every layer of the PCB.
You may have a hidden dry solder joint in there.
 
Notice the pins for the opamp socket is clearly visible
 
Alternatively buy or borrow an oscilloscope of at least 500MHz bandwidth and see if there is a clock pulse train coming out of the XO.
(Some of the old analog scopes might be had for cheap if you look around)
 
Basically if there is no XO signal, there will be no lock.


Hi b0bb, 
 
I think I may have solved the original problem relating to the XO. I have an old analogue 'scope, I will use this to test. But first I must solve another problem.
 
As mentioned in an earlier post, the aftermarket regulators that came with my unit have both broken away from the board and caused some damage. 
 

 
 
These regulators replaced the 1763 IC. I say 'replaced', but from the pristine condition of the solder where the IC should be, it looks like my DAC never had the 1763 in this position in the first place. 
 
As you can see I have temporarily removed the caps with my rework tool in order to facilitate resoldering of the regulator. There are, however, a couple of problems.
 
1) In coming away from the board, the aftermarket regulator has taken with it - and broken - the very small resistor(?) just to the right of the leftmost removed capacitor (near what I think would be Pin 8 of the missing IC).
 
2) The copper trace leading to the +ve of the rightmost removed cap has peeled away (not an insurmountable problem)
 
For information purposes, this is a picture of one of the aftermarket regulators.
 


If I remember correctly, you wrote earlier that there could be RF hum issues associated with such a design.
 
I'm chancing my arm here, but maybe you could answer some or all of the following questions:
 
1) Should I try to restore these aftermarket regulators or, instead, buy a couple of LT1763's? If the latter, then I think it could be right on the limit of my soldering skills.
 
2) How can I replace the missing 'resistor'? I'm assuming it is a resistor, part of an RC combo, but it was actually so small that I couldn't tell what it was.
 
3) The copper trace is slightly damaged at one end of this little resistor, so it would be helpful to know to what it connects, just in case I cannot repair the trace.
 
 
I will admit that I have been very close to jettisoning this now damaged DAC and buying a new 'clean' DAC, but the masochist in me keeps bringing me back ...
 
Thanks as always b0bb.
 
Mar 6, 2017 at 9:55 PM Post #524 of 838
  I'm chancing my arm here, but maybe you could answer some or all of the following questions:
 
1) Should I try to restore these aftermarket regulators or, instead, buy a couple of LT1763's? If the latter, then I think it could be right on the limit of my soldering skills.
 
2) How can I replace the missing 'resistor'? I'm assuming it is a resistor, part of an RC combo, but it was actually so small that I couldn't tell what it was.
 
3) The copper trace is slightly damaged at one end of this little resistor, so it would be helpful to know to what it connects, just in case I cannot repair the trace.
 
 
I will admit that I have been very close to jettisoning this now damaged DAC and buying a new 'clean' DAC, but the masochist in me keeps bringing me back ...
 
Thanks as always b0bb.


 
A couple of comments
1)The melted cap must be replaced, there is a substantial danger it will fail in the future and short out the power supply rail.
2)The missing 0.01uF ceramic cap has to go back in to achieve the 20uV noise performance, it looks like 0805 or 1206 package (US measurement), measure the dimensions and find something suitable on Mouser, X7R or better dielectric.
3)Torn trace can be replaced by a short segment of fine copper wire.
4)Original I/V resistors are still in place, these have to be removed when using the Vishay resistors. Looks like you removed the I/V integrating cap, that needs to go back in.
 
Here is the picture on mine, the integrating cap is the translucent white polystrene cap is on the left

 
The LT1763 must be soldered in using solder paste and hot air, using a soldering iron is not recommended given the extent of the damage to the board.
 
Mar 7, 2017 at 5:48 PM Post #525 of 838

Many thanks as always, b0bb.
 
1) I'll get the damaged cap replaced, great spot. 
 
2) Thanks for identifying the 0.01uF cap, I will source a suitable replacement. In the picture, the northernmost terminal of this (missing) cap appears to connect to Pin 8 of the regulator. Are you able to determine what the southernmost terminal connects to? I ask as I think the copper trace may be damaged here, and it may no longer be a suitable soldering point for terminating the 0.01uF cap.
 
3) I've used the fine copper wire trick elsewhere on the board where the trace is damaged, I will do this again as instructed.
 
4) Yes, I eased out the I/V cap with my rework tool, however this was only as a temporary measure in order to facilitate some awkward soldering around where the aftermarket regulator was located.
 
 
I'll get hold of some soldering paste asap for the 1763 regulator insert. I didn't realise such material existed.
 
As an aside, I think I'll turn down the temperature of both my soldering iron and the rework tool as my hand isn't the steadiest and I seem to be causing collateral damage as I try to 'improve' things.
 
 
As already stated, if I can't recover the damage to this board, then I may acquire a new MH-DA003. If you had the choice, b0bb, would you try out the MkII version (the one with the ES9028 chips) or stick with the older model? Both are still available, it seems.
 

Users who are viewing this thread

Back
Top