The "Lovely Cube" Headphone Amp (Lehmann Black Cube Linear Clone)
Oct 1, 2014 at 4:26 AM Post #1,173 of 1,624
Sorry about taking so long to get back.  I ran into an issue.  I can use my shunt PS on either the positive or negative but not both at the same time.  Let me explain!  When I hook up the positive or negative alone it works fine on that channel but if I hook both up at the same time it blows a 1.25 A fuse.  I'm not sure why it does this as the original design only uses around 145ma per channel and and I've been testing the PS with over a 200ma load on each channel without any issues.
 
I haven't give-up on this idea.  I ordered a new pcb for about $4 and I will build it up and play with it.  My current set-up is working good and I don't want to tear it apart and break it.
 
I just built a Millet Maxed with BJT's and it's sounding very good and have The Wire and Marsh Headphone amp waiting.  Half of the fun is the building for me.  But I do love listening as well.
 
 I will post the results.  If someone has a theory of why it's drawing so much current with both the positive and negative supplies hooked up I would like to understand this problem.
 
Oct 3, 2014 at 7:42 AM Post #1,175 of 1,624
I removed the large power supply caps C7 and C8 the two regulators and R9 and R10 I put the + and - voltages on the corresponding pins on the empty regulator pins and then put the grounds on the grounds.  Did I miss something?  I wanted to try it after pulling all the power supply parts but thought it would be easier just to try it on the new board I ordered.  Thanks for any input!
 
Oct 9, 2014 at 12:06 AM Post #1,176 of 1,624
I received the new PCB and started populating already.  I need to go tomorrow and pick-up a few things.  The new board is a bit smaller but seems fine other than that. It cost about six dollars shipped.
I have damaged builds by doing mods in the past and given myself extra work.  I will play it safe this time.  I'm also going to pick-up some fuses that are slightly bigger.  I have a feeling that everything is okay and just draws more current on start-up than I am calculating.  I need to work this weekend so it may take a little while to finish.
 
Oct 10, 2014 at 11:45 PM Post #1,177 of 1,624
I built the new pcb and tested it and it has the same issue of over current.  I believe when the PS grounds are tied together on the PCB it causes a direct short on the transformer. The transformer was shaking a bit. I will buy a couple of cheap 15v transformers and I believe all will be fine.  When I hooked up first the neg supply it worked fine. Then I hooked up the positive supply side and it was fine. But together its a no go situation.  I'm just wondering if uses a capacitor and or resistor as a ground lift would work? 
 
Oct 12, 2014 at 4:30 PM Post #1,178 of 1,624
Saw someone selling the Littlecube Premium in a local forum for MYR500 (about 180USD) and got interested, found this thread. Color me interested, subbed.
 
**edit
Some questions, how does it fare with sensitive cans? I have the JVC HA-S500 that's rated at 32ohm, 106db/1mw and upcoming Rockjaw Kommand rated at 16ohm, 110db/1mw. I have another Goldring DR150 but that's not much of a concern.
 
Oct 14, 2014 at 4:42 AM Post #1,179 of 1,624
I learned something today.  I picked-up a 0-15, 0-15 transformer and replaced thee 15-0-15 and the problem has been corrected. The two sides are truly separate. On the 15-0-15 it had three wires.
 
Now I need to leave for two straight weeks of work.  I will try to  replace the power supply on the lovely cube again and do some listening tests.
 
Penmaker- It drives my 35 ohm Hifiman cans fine, but I still prefer the Sennheiser HD-600's with it.
 
Oct 17, 2014 at 3:16 PM Post #1,180 of 1,624
I learned something today.  I picked-up a 0-15, 0-15 transformer and replaced thee 15-0-15 and the problem has been corrected. The two sides are truly separate. On the 15-0-15 it had three wires.

Now I need to leave for two straight weeks of work.  I will try to  replace the power supply on the lovely cube again and do some listening tests.

Penmaker- It drives my 35 ohm Hifiman cans fine, but I still prefer the Sennheiser HD-600's with it.
that's great!

I pulled from trigger on a pair of HD650 with a dead left channel, going to have a little fun with the lovely cube once I've checked what's wrong with them.
 
Nov 4, 2014 at 5:36 AM Post #1,181 of 1,624
Well I finally had a day off and finished my experiment with the Salas power supplies.  I tried it on the original board and had just a bit of noise.  I didn't want to deal with it so I finished the
new board and swapped some of the parts over.
 
I've been listening for almost six hours and I don't want to stop.  I wasn't expecting much from the power supply change.  The improvements were more pronounced than I would have guessed. 
 
Using Flac files with Foobar on WASPI=PUSH output and an Arcam rdac.
 
I starting hearing more information on my music.  Extension is improved on both ends. It just sounds more natural.
 
The last couple of hours has been with my Hifiman HE-400's.  I always found them to be dark and therefore the have found duty on my wife's computer with the Objective 2.  They are sounding better than they ever did before.
 
This amp sounds very good in it's original form and most people would never want to venture down this path.  For you more ambitious types you will be pleasantly surprised.
 
Nov 5, 2014 at 10:38 AM Post #1,182 of 1,624
Hi
Just wanted to share my build of a lovely cube.  I'm trying to get as close as possible to the linear SE original version.
I have ordered this board from ebay seller minishow0328 for as little as 38 dollars (free shipping)
http://stores.ebay.com/minishow0328/

I received a beautiful, well made build with the exact layout of lovely cube (and original black cube)
The pot was immediately changed to ALPS 50k and NE5532 to OPA2134 (asked the seller)
I was very happy with the build quality, but as expected, a lot of components just demanded to be upgraded.
 
At first testing I noticed immediately a volume imbalance between L and R.  Measured power supply output (at the LEDS): 14.71V and -15.06V.
I have immediately changed the voltage regulators (probably fakes) with some new ones (together with all transistors and all PSU caps) and now voltages are 14.80 and -14.82 at the leds. Volume imbalance seems to be gone and I must say this thing sounds pretty god damn amazing out of the box.
Allthough lots of caps had wrong values. (4x 0.1uF at rectifying section and 22nF opamp coupling and 22nF input cap bypass were all the same Rifa 0.047 630V - all the white ones in the picture) All transistors looked like fakes and even the opamp I'm not sure.
Still I was completely blown away by the VAST improvement - this is my first dedicated headphone amp and previously I have been using the headphone outputs on my audio interfaces and mixer.  The soundstage!! The separation!! The fast bass response!!! And especially the non-fatiguing experience and airyness!!!
 
Here are some pictures of my build as it is now, with case:
board as received:

Some components already replaced, case built (case is an old Pioneer SA-420 amplifier cut down to size)

Will make better front plate for case later.  Notice I have built a wall outlet into the back (fused 2A) to charge portable sources directly from the unit.
There are some benefits to a slightly oversized case :)
 
This week, I'm waiting for:
 
-2x Mundorf Mlytic 4700uF 40V PSU reservoir caps (to replace Nover audio grade 10000uF 35V)
-2x Mundorf MCap MKP 2,2 uF 250V (to replace both the 1,2 uF EPCOS 1000V and the 0,022uF WIMA MKP10 bypass that I have put there)
-6x Panasonic FC-series 470uF 35V (to replace KME caps in PSU section and Opamp (RC filter?)
-6x Panasonic FC-series 470uF 25V (to replace the former as experiment)
-PRP metal film resistors all over (except PSU section)
-2x LCR polysterene 22nF 63V (to replace wima MKP10 22nF that I have put there as Opamp coupling)
-2x LCR polysterene 100pF 160V (to replace unknown capacitor in Opamp Low pass filter)
-4x vishay roederstein kp1830 22nF 63V (to replace wima MKP10 22nF that I have put there as Opamp coupling)(or to try as input cap bypass)
-4x vishay roederstein mkp1837 22nF 160V (to try as input bypass or opamp coupling)
-2x vishay roederstein mkt1813 22nF 250V (to try as opamp coupling if polysterene gets too much heated)
-Several opamps: LME4986, genuine OPA2134, LT1364
 
I guess this should lift the performance to devine level :)
Any hints how to make the best use of these components are VERY welcome (espescially for roederstein caps)
 
I would like to thank BlaBlaBla, Flukell, francisdemarte, ringer, aspenx and others VERY MUCH for sharing so much great info.
 

Users who are viewing this thread

Back
Top