The "Lovely Cube" Headphone Amp (Lehmann Black Cube Linear Clone) - Page 59

Quote:

Originally Posted by miky2011 

 

 

   Francisdemarte,   DIP are working right. I have tested with multimeter the on/off function.

 

   I'm in Spain. Barcelona.

 

   You think that MKP1837 (22nF) capacitors round the opamp can have any problem? Its the only one that I have not change. 

 

   

Too much guesswork, if you ask me. It is a very ineffective method. And why would 2 capacitors of good quality be defective? As far as I know nobody has complained about unreliable component delivered with the LC:

You did everything right in the beginning, measuring the supply voltage and other important voltages, plus checked the orientation of components etc. Just continue with simple and logical measuring, and you will find the fault very soon.

For example:
1..Set your multimeter to measure small AC voltages. 2. Send a signal to the head-amp, preferably a sinus tone (you can download test tones, or a free tone-generator for PC like ’Sigjenny’). 3. Now measure, say, at the input and the output of the op-amp etc. Is there amplification?

Remember to set the dip switch to amplification, and to turn up the volume..

AC measuring of the signal will not destroy the amp. (one must be more careful when measuring resistances, because a current then is sent through the circuit).



BTW:
You are aware that electrostatic discharge easily can destroy the sensitive input section of an op-amp? Both your op-amps may be defective for the same reason, if the problem boils down to the op-amp. At least it can not be excluded. (Never mount an op-amp without touching the ground of the amplifier with the same hand before the pins are making contact, and moist skin is an advantage).

Olaf
 

EDIT: NVM. I figured it out.

I'm loving Elna Silmic's filtering the Opamp! Can't wait to pair it with some Polystyrenes!

Would the Lovely Cube Premium be a good amp for My ultrasone 750?

Stock, Dual OPA627 or OPA2111AM?

FYI,

The new Lehmann Linear SE was recently shown in Australia with an RRP of $3,199 Australian. That means it will be over $3,000 in the United States. The biggest change advertised is a switch to Mundorff capacitors. Wow!

I know it is insane.

Quote:

Originally Posted by francisdemarte 

 

The new Linear SE, equipped with Mundorf caps in its power supply and signal path as well as a gold-plated AHP microfuse, rests vibration neutral on custom-designed SSC feet. For the interior wiring we used products from the Japanese maker Mogami. Housings made of cool, satin mat aluminium in six homely, noble real wood veneer versions (designed by Guido Gutenstein) protect the superior Lehmannaudio technology and represent a high aesthetical added value.

 

So we have:

 

1. Mundorf Caps (You can buy them here http://www.madisoundspeakerstore.com/supreme-caps/)

2. Fancy Fuse

3. Rubber feet

4. Mogami wire (if you notice the picture there not a lot of wiring since everything is on the board. You may just need a foot or so (http://www.studiowiring.com/mogami.html)

5. New case design

 

Somehow all that adds up to $2000 extra?!

 

 

But..but..what about the R&D costs? 

In other news, I finally got started modding my LC since I have some time on my hand and with each step the sound is becoming more and more sonic orgasm. Literally.

So far I've done the follwoing:

Bypass input caps.

Silmic II 470uf for the Opamp supply.

NOS .027uf Philips Polystyrene Caps bypassing the above.

Panasonic FM swapped for those four 470uf PSU electrolyics. 

Still to do:

Change output resistors (FlukeII method)

Change the 47ohm resistors on pin 8 of opamp.

Change the 4700uf caps.

Add 220uf Muse ES caps to the Silmic's (worth it?...Got plenty sitting here)

Bypass the four PSU eletrolytics (worth it?...Got plenty sitting here also)

Change the ADJ caps for 10uf tantalums.

Moreover, I was reading the Jaycar DIY amp forum over at RockGrotto for some ideas and I constantly came across the name JHL Ripple Eater project which seems like a interesting idea. Think it might help us?

Quote:

Originally Posted by BlackstoneJD 

FYI,

 

The new Lehmann Linear SE was recently shown in Australia with an RRP of $3,199 Australian. That means it will be over $3,000 in the United States. The biggest change advertised is a switch to Mundorff capacitors. Wow!

Quote:

Originally Posted by francisdemarte 

 

The new Linear SE, equipped with Mundorf caps in its power supply and signal path as well as a gold-plated AHP microfuse, rests vibration neutral on custom-designed SSC feet. For the interior wiring we used products from the Japanese maker Mogami. Housings made of cool, satin mat aluminium in six homely, noble real wood veneer versions (designed by Guido Gutenstein) protect the superior Lehmannaudio technology and represent a high aesthetical added value.

 

So we have:

 

1. Mundorf Caps (You can buy them here http://www.madisoundspeakerstore.com/supreme-caps/)

2. Fancy Fuse

3. Rubber feet

4. Mogami wire (if you notice the picture there not a lot of wiring since everything is on the board. You may just need a foot or so (http://www.studiowiring.com/mogami.html)

5. New case design

 

Somehow all that adds up to $2000 extra?!

 

 

I had started to laugh so hard that my tears were running in streams. After few minutes I ended up in hospital. Final diagnosis was - "jaw dislocation and dehydration".

Ten years ago, I used to watch Benny Hill. Today I read advertisements about HIFI equipment.

BUT MOGAMI WIRE! Everyone knows it's the best thing since sliced bread.

Quote:

Originally Posted by JiggaD369 

 

But..but..what about the R&D costs? 

 

In other news, I finally got started modding my LC since I have some time on my hand and with each step the sound is becoming more and more sonic orgasm. Literally.

 

So far I've done the follwoing:

 

Bypass input caps.

Silmic II 470uf for the Opamp supply.

NOS .027uf Philips Polystyrene Caps bypassing the above.

Panasonic FM swapped for those four 470uf PSU electrolyics. 

 

Still to do:

 

Change output resistors (FlukeII method)

Change the 47ohm resistors on pin 8 of opamp.

Change the 4700uf caps.

Add 220uf Muse ES caps to the Silmic's (worth it?...Got plenty sitting here)

Bypass the four PSU eletrolytics (worth it?...Got plenty sitting here also)

Change the ADJ caps for 10uf tantalums.

 

Moreover, I was reading the Jaycar DIY amp forum over at RockGrotto for some ideas and I constantly came across the name JHL Ripple Eater project which seems like a interesting idea. Think it might help us?

 

 

 

 

Pin "4" and pin "8". Those are in power supply. You can change stock 47R resistors with low inductance metal film type, but that is not good chioce. You will get more "grainy" sound (more noise) because low inductance resistor will alow high frequency from LM regulators to pass thorugh (and come to opamp +/- supply). But resistor in output (pin "1" and pin "7") must be the best what you can get because through them signal goes to transistors. Here inductance (and capacitance) must be zero.

Don't go to low with ESR after LM regulatros. They will start to oscillate. In datasheet is written allowed output resistance (usually between 0.01R and 0.001R) that depends on frequency. This means that you have know ESR curve of electrolytic capacitor and be sure that ESR will never go below allowed resistance.

Most people claim that Tantalum capacitors are bad in circuits, but in LM regulators (Cadj value) they are miles ahead over electrolytic capacitor (low leakage current). If you can put MKT/MKC or MKP in Cadj - fine. But it will cost you a lot. (10uF 63V MKT is big and expensive).

Cadj and R2 resistor have big influence on noise. LM regulator likes when R2 is very small. The smaller the R2 is, the bigger effect will have Cadj capacitor. But because you have to devide something (1+R2/R)*1,25V= Vout, you can't go lower than R2=1k1. There is a trick to have lower resistance but the same output voltage - zener diode. Zener diode has 10 times smaller resistance and it will give you something between 14.2 and 14.5 volts on output. Use two (13V, 1W) zener diodes in a parallel connection (not because resistance - this rule does not work here, but for safety) in each LM regulator. Try to find good quality diodes because difference between + and - supply will be less. As far as Cadj goes, Tantalum capacitor are small and have "punch" in sound. Better are MKT/MKC/MKP but also bigger (won't fit here). Electrolytic are too mellow and you have to deal with safety diodes that have to limit output voltage (in LM regulator Vout mustn't go above Vin).

You can change input capacitors (4700 uF) with 2x2200 uF or something like that. Response might be better (lower resistance).

Quote:

Originally Posted by BlaBlaBla 

 

Pin "4" and pin "8". Those are in power supply. You can change stock 47R resistors with low inductance metal film type, but that is not good chioce. You will get more "grainy" sound (more noise) because low inductance resistor will alow high frequency from LM regulators to pass thorugh (and come to opamp +/- supply). But resistor in output (pin "1" and pin "7") must be the best what you can get because through them signal goes to transistors. Here inductance (and capacitance) must be zero.

 

Don't go to low with ESR after LM regulatros. They will start to oscillate. In datasheet is written allowed output resistance (usually between 0.01R and 0.001R) that depends on frequency. This means that you have know ESR curve of electrolytic capacitor and be sure that ESR will never go below allowed resistance.

 

Most people claim that Tantalum capacitors are bad in circuits, but in LM regulators (Cadj value) they are miles ahead over electrolytic capacitor (low leakage current). If you can put MKT/MKC or MKP in Cadj - fine. But it will cost you a lot. (10uF 63V MKT is big and expensive).

 

Cadj and R2 resistor have big influence on noise. LM regulator likes when R2 is very small. The smaller the R2 is, the bigger effect will have Cadj capacitor. But because you have to devide something (1+R2/R)*1,25V= Vout, you can't go lower than R2=1k1. There is a trick to have lower resistance but the same output voltage - zener diode. Zener diode has 10 times smaller resistance and it will give you something between 14.2 and 14.5 volts on output. Use two (13V, 1W) zener diodes in a parallel connection (not because resistance - this rule does not work here, but for safety) in each LM regulator. Try to find good quality diodes because difference between + and - supply will be less. As far as Cadj goes, Tantalum capacitor are small and have "punch" in sound. Better are MKT/MKC/MKP but also bigger (won't fit here). Electrolytic are too mellow and you have to deal with safety diodes that have to limit output voltage (in LM regulator Vout mustn't go above Vin).

 

You can change input capacitors (4700 uF) with 2x2200 uF or something like that. Response might be better (lower resistance).

 

Hey bla, I ordered:

1. CMF55 1%/500mw/50ppm 47.5ohm resistors for pin 1 and 7. Good choice or throw them in the trash?

2. Is BZX85C13 fine for the two zener diodes in parallel for R2?

3. Vishay Sprague 199D 10uf/25V tantalum for the ADJ caps.

4. Vishay BYV27-200-TR for rectifiers.

5. 470uf Panasonic FC to change out for the four FM I put in yesterday.

5. EKMH500VSN472MQ35T for 4700uf swap.

6. 100pf Polystyrene.

7. MBB02070C3019FRP00 (3x30 ohm in parallel) for FlukeII method of the 10hm output resistor.

8. Vishay CMF/RN 1%/50ppm for all signal path resistors. The only reason I ordered these was because I checked the resistor values on the amp yesterday and some of them were +/- 10%.

I know alot of it is overkill but I'm very comfortable with DIY'ing now and my soldering has become very good. So why not? 

Quote:

Originally Posted by JiggaD369 

Hey bla, I ordered:

 

CMF55 1%/500mw/50ppm 47.5ohm resistors for pin 1 and 7. Good choice or throw them in the trash?

Vishay Sprague 199D 10uf/25V tantalum for the ADJ caps.

Is BZX85C13 fine for the two zener diodes in parallel for R2?

Vishay BYV27-200-TR

470uf Panasonic FC to change out for the four FM I put in yesterday.

EKMH500VSN472MQ35T for 4700uf swap.

100pf Polystyrene.

MBB02070C3019FRP00 (3x30 ohm in parallel) for FlukeII method of the 10hm output resistor.

Vishay CMF/RN 1%/50ppm for all signal path resistors. The only reason I ordered these was because I checked the resistor values on the amp yesterday and some of them were +/- 10%.

 

I know alot of it is overkill but I'm very comfortable with DIY'ing now and my soldering has become very good. So why not? 

CMF 55 (Dale) is excellent resistor (actually I don't know anything from Vishay that is less than excellent). Dale is excellent choice.

In signal path always goes metal film. Dale is neutral. PRP is more smooth and airy. Holco is very transparent. Takman is more closed - I didn't like it too much.

But in output path if you put metal film (3xMBB Vishay), you might find that sometimes sound is too harsh or metallic. Then you can put old but excellent Allen Bradley, that will smooth harsh edges and warm up all things little bit... Or you can go towards very fast and detailed Kiwame/KOA Speer resistor. I must say that I have never seen so detailed and neutral carbon film resistor like Kiwame. It behaves almost like metal film. But without metallic sound signature. For comparison (vs Kiwame) Riken Ohm is smoother and fluid, and Bradley more warm and has "body".

Quote:

Originally Posted by BlaBlaBla 

CMF 55 (Dale) is excellent resistor (actually I don't know anything from Vishay that is less than excellent). Dale is excellent choice.

 

In signal path always goes metal film. Dale is neutral. PRP is more smooth and airy. Holco is very transparent. Takman is more closed - I didn't like it too much.

 

But in output path if you put metal film (3xMBB Vishay), you might find that sometimes sound is too harsh or metallic. Then you can put old but excellent Allen Bradley, that will smooth harsh edges and warm up all things little bit... Or you can go towards very fast and detailed Kiwame/KOA Speer resistor. I must say that I have never seen so detailed and neutral carbon film resistor like Kiwame. It behaves almost like metal film. But without metallic sound signature. For comparison (vs Kiwame) Riken Ohm is smoother and fluid, and Bradley more warm and has "body".

 

HI Bla,

Thanks for the input. IF the MBB don't sound too good than I'll order some Allen Bradley from ebay.

But more importantly, Is BZX85C13 fine for the two zener diodes in parallel for R2?
 

Quote:

Originally Posted by JiggaD369 

 

HI Bla,

 

Thanks for the input. IF the MBB don't sound too good than I'll order some Allen Bradley from ebay.

 

But more importantly, Is BZX85C13 fine for the two zener diodes in parallel for R2?
 

 

Yes. It is.