The "Lovely Cube" Headphone Amp (Lehmann Black Cube Linear Clone)
Feb 15, 2011 at 2:51 AM Post #451 of 1,624


Quote:
Hmm, I think I learning what all this DC offset is all about, now. Thanks bla.
 
But what confuses me is one channel is 14.3 mV and the other is 4.2mV. Is something wrong? I tried every gain setting but it's always about ~10mV apart. 
 
P.s. I use 2107.
 
confused_face.gif


Nothing is wrong. Because of transistor mismatch, you have DC offset. In a circuit of regulators (LM317 and LM337), instead 1.1k resistor, you can put 1k resistor, and in series with that new 1k resistor - 200R trimpot. Then you should lower the Vout voltage of both regulators for about 0.5 volts and tune it to the lowest possible offset. "Sweet spot" in my setup is +14.2V and -14.25V. But I have BC109C and 179C regulators. You have to find "sweet spot" in your setup. It really does not matter if supply voltage is 15V or 14.5V or 14.2V.
 
But do not go under 14V because regulators could start dissipate too much wasted power (21V-14V = 7V, lets say 7Vx0.3A = 2.1W, 2.1W*27C°/W = 57C° above room temp)
Usually, for high demanding loads (headphones are not that kind of load) Vin - Vout is around 3V. In Lovely Cube Vin-Vout is 6V (21V-15V = 6V) that means better stability on Vout, but also bigger heat dissipation. Bigger Vin-Vout difference menans better stability but needs bigger heatsink. Smaller difference means lower stability but also smaller heatsink.
 
Feb 15, 2011 at 12:46 PM Post #452 of 1,624
Thanks Bla !
That really brought some light in the dark.
But one thing I'd really like to know is whether it i normal that DC offset goes up with more play time? I used the LC for maybe 10 hours or so and the DC offset rises slowly every with every time I'd put it on.
 
Feb 15, 2011 at 3:33 PM Post #453 of 1,624


Quote:
Nothing is wrong. Because of transistor mismatch, you have DC offset. In a circuit of regulators (LM317 and LM337), instead 1.1k resistor, you can put 1k resistor, and in series with that new 1k resistor - 200R trimpot. Then you should lower the Vout voltage of both regulators for about 0.5 volts and tune it to the lowest possible offset. "Sweet spot" in my setup is +14.2V and -14.25V. But I have BC109C and 179C regulators. You have to find "sweet spot" in your setup. It really does not matter if supply voltage is 15V or 14.5V or 14.2V.
 
But do not go under 14V because regulators could start dissipate too much wasted power (21V-14V = 7V, lets say 7Vx0.3A = 2.1W, 2.1W*27C°/W = 57C° above room temp)
Usually, for high demanding loads (headphones are not that kind of load) Vin - Vout is around 3V. In Lovely Cube Vin-Vout is 6V (21V-15V = 6V) that means better stability on Vout, but also bigger heat dissipation. Bigger Vin-Vout difference menans better stability but needs bigger heatsink. Smaller difference means lower stability but also smaller heatsink.



Bla, thanks for the insight!
 
Excuse my lack of knowledge as I'm a newwbie but would you mid posting a picture of the trimpot setup you have? I looked inside of the LC but the space is so small I don't know how I can achieve this.
 
Also, any perticular reason why did you replace the regulators? How do they desipate heat since they seem like metal cans?
 
Feb 15, 2011 at 5:44 PM Post #454 of 1,624


Quote:
Bla, thanks for the insight!
 
Excuse my lack of knowledge as I'm a newwbie but would you mid posting a picture of the trimpot setup you have? I looked inside of the LC but the space is so small I don't know how I can achieve this.
 
Also, any perticular reason why did you replace the regulators? How do they desipate heat since they seem like metal cans?


No problem. I will post pictures tomorrow.
 
Feb 16, 2011 at 1:32 AM Post #455 of 1,624
Hey bla,
 
I found a matched pair of Philips BC109C (rare) but I can't seem to find the BC179C anywhere! Do you know any place where I can buy them or is there a substitute for it?
 
Feb 16, 2011 at 3:12 AM Post #457 of 1,624


Quote:
Hey bla,
 
I found a matched pair of Philips BC109C (rare) but I can't seem to find the BC179C anywhere! Do you know any place where I can buy them or is there a substitute for it?


http://www.banzaimusic.com/BC179C.html

 
Here I bought BC109C and BC179C. BC179C is made by CDIL. Today, it seems that CDIL is the only one who makes old BC transistors. In my amp, transistors work fine, and have big hFE. According to my measurements, BC109Cs have 600-650, BC179Cs have 700-740. If you can, buy them 4 or 5 and try to find the closest pair.
 
Closest component that I found was BC550C with hFE around 560, and BC560C with hFE mostly below 600 (a few of them have hFE 600-620).
 
But BC transistors are hard to match. Their hFE is high, but vary a lot (150). Also I found that american 2N (2N5087/5088) has lower hFE (450) but also less fluctuation accross the range (50). The most tight variations have 2SA1015/2SC1815 (20), but hFE arround 400.
 
Feb 16, 2011 at 3:52 AM Post #459 of 1,624
So would you recommend going with the matched Philips BC109C i found and the BC179C from Banzai (hard to obtain) or something else like the BC5XX (easily available from USA)?
 
Feb 16, 2011 at 5:08 AM Post #460 of 1,624


Quote:
So would you recommend going with the matched Philips BC109C i found and the BC179C from Banzai (hard to obtain) or something else like the BC5XX (easily available from USA)?


It is really a metter of personal taste. I think BC109C/179 and 2N5087/88 will work perfectly. When you want to replace european transitor (BC) with american (2N) or japan (2SA/2SC), you have to pay attention on pins. They have different orientation. I would not suggest you to use japan transistor (in european or american topology) because their base pin is not centered like on american or european transistors... Always read technical documentation about transistor that you want to buy.
 
With BC109C and BC179C, you have to take care about voltages. They are not rated as high as BC550/560... Supply of LC is +15V/-15V. BC109C VCEO is rated for 20V. There are resistors that reduce voltage, but  always measure to be 100% sure.
 
BC5xx will do the job but you will have to buy them 20 and find good match. If you can choose (and find) use the one that have hFE above 500.
 
Feb 16, 2011 at 5:21 AM Post #461 of 1,624


Quote:
Bla, thanks for the insight!
 
Excuse my lack of knowledge as I'm a newwbie but would you mid posting a picture of the trimpot setup you have? I looked inside of the LC but the space is so small I don't know how I can achieve this.
 
Also, any perticular reason why did you replace the regulators? How do they desipate heat since they seem like metal cans?

 
Picture is here...
 

 
I do not have Lovely Cube at work. So I made a drawing.
 
http://img228.imageshack.us/i/94348873.jpg/
 
Feb 16, 2011 at 1:40 PM Post #462 of 1,624


Quote:
Bla, thanks for the insight!
 
Excuse my lack of knowledge as I'm a newwbie but would you mid posting a picture of the trimpot setup you have? I looked inside of the LC but the space is so small I don't know how I can achieve this.
 
Also, any perticular reason why did you replace the regulators? How do they desipate heat since they seem like metal cans?

 
Pictures are here:
 
http://www.freeuploadimages.org/images/77fd8d139u5gp9lcpbet.jpg
 
http://www.freeuploadimages.org/images/8izqfkmnx6534me8n4v6.jpg
 
 
Feb 16, 2011 at 2:37 PM Post #463 of 1,624
Thanks bla!
 
Great clean job!
 
I will add two trimpots to my mouser order! 
biggrin.gif

 
BUT, I have a question regarding the buffer stage.
 
I also want to replace the 2 BLUE tinfoil and 2 RED MKC caps around the Opamp. The statesd values are 100pF and .022uF, respectively. DO I have to replace them with the ame values or can I go up? If so, how much?
 
Feb 16, 2011 at 5:39 PM Post #465 of 1,624


Quote:
Thanks bla!
 
Great clean job!
 
I will add two trimpots to my mouser order! 
biggrin.gif

 
BUT, I have a question regarding the buffer stage.
 
I also want to replace the 2 BLUE tinfoil and 2 RED MKC caps around the Opamp. The statesd values are 100pF and .022uF, respectively. DO I have to replace them with the ame values or can I go up? If so, how much?

 
If you want to replace blue tinfoil capacitors I would suggest you to replace them only with polystyrene (Styroflex).  Stock Philips Tinfoil is not up to date capacitor, but has its own strength. It gives you more "in front" sound presentation.  If you replace Tinfoil with WIMA MKP (I would not do that at all costs) you will have more rounded sound. Too rounded. Personally I do not like WIMA capacitors and I avoid them.  ERO MKC is (maybe) slow and old but it has "something". (But it is up to you.) Polystyrene is excellent choice here. It gives very wide (and almost magically "air") sound stage. And they have excellent precision and dynamic over entire frequency range. I would not change capacity because if you put there bigger cap then you might have high frequency roll off. If you put there lower capacity then OPAMP will amplify very high frequencies that  might put transistors in to oscillation. 100pF represents frequency filter that does not let to high frequencies to pass through OPAMP. If you do not notice any roll off, then 100pF is good.
 
Capacitors for OPAMP coupling is just another story. 22nF is optimum between too slow and to too little. If you want to change something then you should choose better cap. Personally I do not know if MKC are good caps for coupling, because I put there blue ERO KP 1830 - 1% (some people say that they are "the best" for coupling). Also tried WIMA, but that crap I will avoid for a long time. As I said before - too round and artificial. If you put bigger cap that means bigger capacity (reserve for OPAMP) but also means slower reaction. Philosophy always goes toward this: "If you have two capacitors in parallel connection, then bigger one always represents "power tank", and smaller one always represents "speed". " I tried bigger and slower cap. Smaller cap (10nF) polystyrene has more impact on sound that bigger (47nF). But in my case, 10nF polystyrene gave me too much sound stage and too weak vocals. It was not good combination with quite fast Panasonic FM. 47nF polystyrene was better in vocals but it was like entire frequency range went up. Too bright. Again not good with Panasonic FM.
 
ERO KP capacitor has flawless positioning and precision. It is good for opamp coupling. For DC blocking ERO MKC is better. MKC has sparkle and posses power in mid range (vocals) that KP and polystyrene do not have. MKC has in front presentation. KP has better separation and positioning (slightly back than MKC) but less power in vocals. Polystyrene is wide, laid back and "air" and has the least power in the mid range.
 
I think that for small caps, polystyrene are excellent. Polystyrene is also good for DC blocking but it has to be in a combination with something that has strong mid frequency range. ERO KP is good for coupling, and MKC is very good for DC blocking. (In my case.)
 
But it all depends on your headphones, and type of music that you like.
 
 

Users who are viewing this thread

Back
Top