RJM Audio Sapphire Desktop Headphone Amplifier

This time the resistors values are set for -40db of total attenuation. I used only 1% resistors standard values, so it is easier to find the components and to replace them. The total resistance was randomly achieved. How does it look?
 

Attachments

  • AttenuatorV3.png
    AttenuatorV3.png
    9 KB · Views: 428

rjm

Member
Joined 2004
Paid Member
With only 11 steps, 1 being "-inf", half of those being above the usual listening volume position, it's inevitable that the steps will be larger than ideal.

The rest really comes down to knowing in advance what the listening range of attenuation is going to be. One way to check is to note the position on your existing volume control, then (on a similar unit) measure the resistance between wiper and ends, and calculate the dB loss.
 
With only 11 steps, 1 being "-inf", half of those being above the usual listening volume position, it's inevitable that the steps will be larger than ideal.

Perfect, thank you so much for the information, I didn't get it at the beginning.

At this point I think I will just try with the current setup, modifying and testing it until I don't get the values I prefer.

I will post the final result in case someone wants to try it.
 
Hi,

I would like to ask if it is feasible to use the Sapphire as both headphone amp and line stage at the same time / with the same configuration. I ask because I have a preamp which is basically just an input selector, volume pot and minidsp 2x4HD for eq and crossover for my biamped main speakers. This works really well, but I don't have quite enough gain from my sources to drive the speakers loud enough. At the same time I would also like to add a headphone out to my preamp so that I can use the same sources, selector and volume control when listening to headphones.

My idea was to put the sapphire in between the volume control and minidsp with an interrupting type headphone jack so that when the headphones are inserted the connection to the minidsp is cut. My headphones are Sennheiser HD700 which are 150 Ω and 105dB. The minidsp has in input impedance of 10k and switchable 2 or 4 V rms input sensitivity. I estimate that I need 12-15dB gain for both line stage and headphones.

Would there be any problems with this? many thanks in advance.
 
There are two considerations.

1. Do you need different voltage gain for the two applications. If so, you'll have to wire up the gain select switches to some convenient control knob on the front panel.

2. Is the Sapphire going to misbehave when there is a downstream component connected. Normally it just connects to headphones, so no issue with earthing and ground loops. With an active component instead, this is no longer a certainty. If there is, you'll have to have a switch that disconnects the output from the amp circuit when you connect headphones. There are headphone sockets with this switching function built in.

/R
 
Just another voice chiming here to praise the Sapphire. I built it 1.5 years ago and it was really nice. I have since move from pro audio to listening and now feed the Sapphire with a RME ADI-2 DAC. My headphones are DT-1990 Pro's (250ohm). C1 are Ampohm PIO Tin Foil caps.

The Sapphire GREATLY exceeds the clarity and presence of the already fantastic onboard HP amp on my DAC. The mids are slightly more forward and rounded. It is definitely not a reference sound, but not so far off that it cannot be used as a reference. I'd describe it as slightly warm but still extremely balanced.

Bravo and thank you Richard for your wonderful work on this circuit. I can't commit to buying more new toys in the future, but this amp will have a home on my desk forever.


A question- Can anyone try to compare the Mundorf Supreme vs PIO/Tin caps? I have some on order and want to experiment either way, but some opinions would be great!
 
There are two obvious issues. Without C1 you have a DC amplifier. Any DC offset on your input signal will end up on the output, multiplied by the gain. Second, the trim pot. which is used to adjust the DC output offset to zero no longer functions. So even if you don't have any DC on the input signal, you may likely have a large DC offset on the output regardless, and no way to counter that.
 
There are two obvious issues. Without C1 you have a DC amplifier. Any DC offset on your input signal will end up on the output, multiplied by the gain. Second, the trim pot. which is used to adjust the DC output offset to zero no longer functions. So even if you don't have any DC on the input signal, you may likely have a large DC offset on the output regardless, and no way to counter that.


Thanks Richard. You explained it beautifully. Now I know what to expect. I think I'm lucky, because the output offset without a connected source is 5mV, and with a connected 20mV. I don't use C1 because it soothes transients in music.
 
Thanks Richard. You explained it beautifully. Now I know what to expect. I think I'm lucky, because the output offset without a connected source is 5mV, and with a connected 20mV. I don't use C1 because it soothes transients in music.

You may need to carefully measure output DC transient while power up/ down also if your input source does not have output coupling capacitor.
 

rjm

Member
Joined 2004
Paid Member
I think I'm lucky, because the output offset without a connected source is 5mV, and with a connected 20mV.

I really can't recommend you operate it like this. The offset might have been 20 mV when you measured it, but you have no guarantee that it will stay at 20 mV. It will be dependent on the configuration of the output circuit of the component you have plugged into the inputs.

However, that said, there is little risk of actual damage to anything because the headphone impedance is high enough to withstand even very high offset voltages. To be absolutely clear though: That does not make it any less wrong.
 
I really can't recommend you operate it like this. The offset might have been 20 mV when you measured it, but you have no guarantee that it will stay at 20 mV. It will be dependent on the configuration of the output circuit of the component you have plugged into the inputs.

However, that said, there is little risk of actual damage to anything because the headphone impedance is high enough to withstand even very high offset voltages. To be absolutely clear though: That does not make it any less wrong.
I did it but it every hot at two Bd135 and two bd136. what wrong Mr RJM?. how to fix it? and when i try to adjust bias i don't know where is bias because i new member. thank you so much
 

rjm

Member
Joined 2004
Paid Member
The bias current is set by the resistors you use, it is not adjustable, nor should it need to be adjusted.

Without further information about what you did and what the measured voltages are, I can't provide any useful advice.

Those transistors will get a little warm in the normal version, and quite hot in the high bias current build. There are many mistakes you could have made that could cause the currents in the circuit to rise above normal. I suggest emailing me to discuss further.
 
Hi, Guys,

I finished the Sapphire headphone amp, Finally.:D
I order the kit of Sapphire, It is easy to build with very well designed board.
So, I successfully done the kit once, without any mistake.

My Sapphire is different from the original.

First, I change the PSU. What I use is the GR regulation that is designed by a Taiwanese. The GR regulation use a highly precise and ultra-low noise voltage reference IC and a op amp for error amplifier.

The last, the VR is Alps RK-50, it is located in another case.

The sound is awesome. With my ATH AWKT, it sound smooth and clear, with good punch. I like it.
 

Attachments

  • 8.jpg
    8.jpg
    586.9 KB · Views: 268
  • 7.jpg
    7.jpg
    51.1 KB · Views: 244
  • 6.jpg
    6.jpg
    94.5 KB · Views: 240
  • 5.jpg
    5.jpg
    615.3 KB · Views: 430
  • 4.jpg
    4.jpg
    978.3 KB · Views: 445
  • 3.jpg
    3.jpg
    856.8 KB · Views: 460
  • 2.jpg
    2.jpg
    471 KB · Views: 457
  • 1.jpg
    1.jpg
    479.2 KB · Views: 454
  • 9.jpg
    9.jpg
    985.4 KB · Views: 259