The "Lovely Cube" Headphone Amp (Lehmann Black Cube Linear Clone)

[JiggaD369]

Is there anyone who can match these parts for us?

 

[BlaBlaBla]

FlukeII, I was on internet page that you had sended to us...
That workshop is better equipped than my university. Multimeter costs twice than Lovely Cube alone....
 
Man, in that demand for perfection, you have gone too far. 

 

[FlukeII]

 
Quote:

FlukeII, I was on internet page that you had sended to us...
That workshop is better equipped than my university. Multimeter costs twice than Lovely Cube alone....
 
Man, in that demand for perfection, you have gone too far. 

Hi BlaBlaBla,
Electrical Engineering is my livelihood and I learnt early on having good test instruments can save a lot of hassle and bother – they don’t have to be new, just good and reliable.
 
I guess if you want to make Electronics a long-term hobby, investment in a reasonably good Digital MultiMeter (DMM), for example, will always hold you in good stead.
 
Sorry good test equipment isn’t that cheap; But it will last, the first Fluke DMM I bought (FLUKE 89) cost me a lot after 14 years I gave it away to relative when I bought the FLUEK 189, now 20 years on that FLUKE 89 is still going strong. Also, when you do something dumb when measuring a circuit, like trying to measure a voltage with the Amps range, you want to know that the only thing that has blown is the fuse in the DMM! I’ve done this a few times with my FLUKE, carelessly not thinking and/or tired, and was relieved when the internal fuse was replaced everything worked OK – One of my friends bought a much cheaper DMM and had the same misfortune, the only problem was that after he replaced the internal fuse in his DMM, the DMM still did not work!
 
This isn't a sales promotion on FLUKE by the way, I've seen some great DMM's from DIGITECH and SANWA that cost half what a FLUKE110 would cost.
 
Personally, I would really like an Audio Precision 2722 Analyzer for screwing around with with Audio Circuits: Check out: http://ap.com/products/2700 Only trouble is these cost around USD$20K! I know I'm dreaming, but it gives you and idea of the relative cost of a good DMM to an industry benchmark like the 2722 Audio Analyzer. I know someone who works in the US for a large corporate who makes Audio products and he mentioned to me once that at their factory in Texas they had several of these 2722 analyzers as well as about 6 of the older AP Two (This is a discontinued version: It's commercial names was "System Two" from Audio Precision: They do come up on Ebay from time to time; if you buy one be prepared to send if back to Audio Precision to get is aligned and tested - it will probably need updating. Also make sure you get the PC Card Interface - if the model you acquire needs it - these are not cheap from Audio Precision!!) versions, these still cost around USD$18K a piece. Where I work we have an AP Two - I use it from time to time for my projects.
 
I'm just trying to set the scene - this is the kind of investment some manufacturers will go to.

 

[JiggaD369]

Hey Flukell,
 
Are you still going to post the spreadsheet with the caps you tried around the opamp?

 

[FlukeII]

Quote:

Hey Flukell,
 
Are you still going to post the spreadsheet with the caps you tried around the opamp?

There is more to come, I've just been busy with other things i.e. work.
 
Other things:

• User a Super Beta filter to fed the OpAmp
• OpAmp changes: I like the LME49720, but friends prefer the AD823, NE5532 even! Personally, I think the OPA2134P is a little dull sounding, it's polite, but to me lacks punch, on the other hand the AD823 is more in your face presentation...
• Power Supply mods
• Triming the LM337/LM317 voltage regulators to get the +ive and -ive voltages equal

 

[FlukeII]

And the different capacitors for the power supply stage; as well as alternatives to the United Chemicon LXZ 470uF 25V Electrolytic used to decouple the various stages in the amp...

 

[BlaBlaBla]

Yesterday I ordered LME49720HA... I will try it in combination with 2N5087/5088 and matched MJE243/MJE253.
 
Hi FlukeII, what do you think about stock TO220 heatsinks (in LC)? I think that amplifier is too warm. I mean, I do not believe in sentence "it is A class so it has to be warm". Problem is there is not enough place for bigger heatsink...
 
Quote:

Quote:

Hey Flukell,
 
Are you still going to post the spreadsheet with the caps you tried around the opamp?

There is more to come, I've just been busy with other things i.e. work.
 
Other things:

• User a Super Beta filter to fed the OpAmp
• OpAmp changes: I like the LME49720, but friends prefer the AD823, NE5532 even! Personally, I think the OPA2134P is a little dull sounding, it's polite, but to me lacks punch, on the other hand the AD823 is more in your face presentation...
• Power Supply mods
• Triming the LM337/LM317 voltage regulators to get the +ive and -ive voltages equal

 

[ringer]

Quote:

Yesterday I ordered LME49720HA... I will try it in combination with 2N5087/5088 and matched MJE243/MJE253.
 
Hi FlukeII, what do you think about stock TO220 heatsinks (in LC)? I think that amplifier is too warm. I mean, I do not believe in sentence "it is A class so it has to be warm". Problem is there is not enough place for bigger heatsink...
 
Quote:

Quote:

Hey Flukell,
 
Are you still going to post the spreadsheet with the caps you tried around the opamp?

There is more to come, I've just been busy with other things i.e. work.
 
Other things:

• User a Super Beta filter to fed the OpAmp
• OpAmp changes: I like the LME49720, but friends prefer the AD823, NE5532 even! Personally, I think the OPA2134P is a little dull sounding, it's polite, but to me lacks punch, on the other hand the AD823 is more in your face presentation...
• Power Supply mods
• Triming the LM337/LM317 voltage regulators to get the +ive and -ive voltages equal

It has to be warm. Because in class A transistors are always open.

 

[BlaBlaBla]

Quote:

Quote:

Yesterday I ordered LME49720HA... I will try it in combination with 2N5087/5088 and matched MJE243/MJE253.
 
Hi FlukeII, what do you think about stock TO220 heatsinks (in LC)? I think that amplifier is too warm. I mean, I do not believe in sentence "it is A class so it has to be warm". Problem is there is not enough place for bigger heatsink...
 
Quote:

Quote:

Hey Flukell,
 
Are you still going to post the spreadsheet with the caps you tried around the opamp?

There is more to come, I've just been busy with other things i.e. work.
 
Other things:

• User a Super Beta filter to fed the OpAmp
• OpAmp changes: I like the LME49720, but friends prefer the AD823, NE5532 even! Personally, I think the OPA2134P is a little dull sounding, it's polite, but to me lacks punch, on the other hand the AD823 is more in your face presentation...
• Power Supply mods
• Triming the LM337/LM317 voltage regulators to get the +ive and -ive voltages equal

It has to be warm. Because in class A transistors are always open.

But I really do not like when it becomes warm! I do not like when transistors are hot (over 60°C) and I do not care about commercial trick "it works best when it is warm". I am sure that class A amplifier can be cold and still operate in A class. Best amlifiers behave in that way, because designers are not trying to save money on the most important thing - cooling. But this minimalistic design just sucks. Small design involves high noise, too much heating and all components in it have shorter life.
 
I bought a few of these heatsinks: http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=280498961521&ssPageName=STRK:MEWNX:IT I also bought a few heatsinks from local store that are higher than those supplied with LC. Because of mimimalistic design I will have to improvise (drill extra holes, etc)...
 
Output resistors (1W) become brown (I do not like that because they were white as new) and also know that good resistors do not change color, unless they are overloaded), so I concluded that 1W is not enough. Besides, I can not find who made them. Maybe some "Middle of nowhere" company from China? Mills is too big (5W). Riken, Shinkoh are to expensive (5-10 USD each) Vishay is sick (10-15 USD each), Takman (2 USD each) I like more in the signal circuit, Alan Bradley I have never tried... Kiwame 2W are the best choice...

 

[BlaBlaBla]

Quote:

Quote:

Yesterday I ordered LME49720HA... I will try it in combination with 2N5087/5088 and matched MJE243/MJE253.
 
Hi FlukeII, what do you think about stock TO220 heatsinks (in LC)? I think that amplifier is too warm. I mean, I do not believe in sentence "it is A class so it has to be warm". Problem is there is not enough place for bigger heatsink...
 
Quote:

Quote:

Hey Flukell,
 
Are you still going to post the spreadsheet with the caps you tried around the opamp?

There is more to come, I've just been busy with other things i.e. work.
 
Other things:

• User a Super Beta filter to fed the OpAmp
• OpAmp changes: I like the LME49720, but friends prefer the AD823, NE5532 even! Personally, I think the OPA2134P is a little dull sounding, it's polite, but to me lacks punch, on the other hand the AD823 is more in your face presentation...
• Power Supply mods
• Triming the LM337/LM317 voltage regulators to get the +ive and -ive voltages equal

It has to be warm. Because in class A transistors are always open.

But I really do not like when it becomes warm! I do not like when transistors are hot (over 60°C) and I do not care about commercial trick "it works best when it is warm". I am sure that class A amplifier can be cold and still operate in A class. Best amlifiers behave in that way, because designers are not trying to save money on the most important thing - cooling. But this minimalistic design just sucks. Small design involves high noise, too much heating and all components in it have shorter life.
 
I bought a few of these heatsinks: http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=280498961521&ssPageName=STRK:MEWNX:IT I also bought a few heatsinks from local store that are higher than those supplied with LC. Because of mimimalistic design I will have to improvise (drill extra holes, etc)...
 
Output resistors (1W) become brown (I do not like that because they were white as new) and also know that good resistors do not change color, unless they are overloaded), so I concluded that 1W is not enough. Besides, I can not find who made them. Maybe some "Middle of nowhere" company from China? Mills is too big (5W). Riken, Shinkoh are to expensive (5-10 USD each) Vishay is sick (10-15 USD each), Takman (2 USD each) I like in the signal circuit, Alan Bradley I have never tried... Kiwame 2W are the best choice...

 

[ringer]

Quote:

But I really do not like when it becomes warm! I do not like when transistors are hot (over 60°C) and I do not care about commercial trick "it works best when it is warm". I am sure that class A amplifier can be cold and still operate in A class. Best amlifiers behave in that way, because designers are not trying to save money on the most important thing - cooling. But this minimalistic design just sucks. Small design involves high noise, too much heating and all components in it have shorter life.
 
I bought a few of these heatsinks: http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=280498961521&ssPageName=STRK:MEWNX:IT I also bought a few heatsinks from local store that are higher than those supplied with LC. Because of mimimalistic design I will have to improvise (drill extra holes, etc)...
 
Output resistors (1W) become brown (I do not like that because they were white as new) and also know that good resistors do not change color, unless they are overloaded), so I concluded that 1W is not enough. Besides, I can not find who made them. Maybe some "Middle of nowhere" company from China? Mills is too big (5W). Riken, Shinkoh are to expensive (5-10 USD each) Vishay is sick (10-15 USD each), Takman (2 USD each) I like more in the signal circuit, Alan Bradley I have never tried... Kiwame 2W are the best choice...

 
That radiators will conduct worm from transistors. But the case will not cool down while radiators inside. If you want to cool down all stuff in the case there are two options:
1. Make active cooling
2. Pull out radiators from the case

 

[BlaBlaBla]

I found old Riken Ohm RM resistors. But... Very expensive.

 

[JiggaD369]

Quote:

I found old Riken Ohm RM resistors. But... Very expensive.

How would Kiwame 2W resistors work for the 1W resistors?

 

[BlaBlaBla]

 
Quote:

Quote:

I found old Riken Ohm RM resistors. But... Very expensive.

How would Kiwame 2W resistors work for the 1W resistors?

I heard that there is something written like P=U2/R and P=I2*R where I - current through resistor, U - voltage across resistor, R - resistance
 
When I calculate few things I get something close to 1W for 10R resistance.
 
But to be honest, I give a crap about books.
 
I always use my intuition. My sixth sense. First I take part into my hands, and then I decide what to do. I have never decided wrong. Probably because I do not waste time calculating too much (because nothing is ideal like it is stated in book) and because I like to exaggerate ("old school way of thinking" where everything was twice as big and oversized, but has lasted few decades). So I will put there 2W. And if I could I would put there 5W Mills.
 
Moreover, I like green color of Kiwame resistor and also like how word Kiwame" (Ki-Wa-Me) sounds. It is simple word (name) and easier to remember than phrase "some crappy no-name metal film, made in Middle of nowhere".
 
Same goes for next brands:
 
Word Riken (Ri-ken) reminds me on car plates in my town (RI-xxx-xx). 
 
PRP reminds me on something strong, effective. P - R - P
 
Mills has mellow "M" and beautiful "sssss".
 
Caddock has big "D" (in croatian language). Something "Deep", "Dark", "Dreadful". Wow.
 
Shinkoh has heavenly "Shhhh" and it is beautifully crafted. So simple and so stylish (you see - "sh").
 
There is some art, in those words above...
 
 
People, please no offense... I was joking!  
 
JiggaD369, use ohms law and do not bother with small digits. Everything abowe 1W (2W and more) will be ok. It depends only about free space in your amplifier. It is very uncommon using 1W resistors in output stage. I mean, output stage is not place for saving money...

 

[Whitefox]

Quote:

That radiators will conduct worm from transistors. But the case will not cool down while radiators inside. If you want to cool down all stuff in the case there are two options:
1. Make active cooling
2. Pull out radiators from the case

There are still way use the enclosure as a radiator.
But it's best if the aluminum case...