Hi, I managed to detect the problem, it is due to bad quality switch for the gain switch that make them sound weird. I bought the better gain switch know and get the fanstastic result. I would love to know you guys opinion for upgrading the sound. Would it get good result with replacement of two input cap, right now I am using M cap Mundorf 4.7uf. Or is there anyone here experiment on changing transistor bc337 to another and better transistor.
Hi RJM, I have one question. I have Sapphire 3.0f, can be used as preamplifier? If I remove on integrated amplifier jumper Pre Out-Main and connect output of Sapphire to Main? Gain is not problem, I have Sapphire at 1x gain, and can be modified if will be needing. My Sapphire have much better parts and construction, integrated amplifier have 3 electrolytic capacitors, worst OpAmp, potentiometer, resistors and many switches in preamp path, one transformer for both stages etc..
Edit: Signal Path of Integrated amplifier is: Input - 100uF - 100ohm - OpAmp - 150ohm - 100uF - Source Direct bypass over correction - Volume potentiometer - 100uF - 470ohm - Main stage.
Edit: Signal Path of Integrated amplifier is: Input - 100uF - 100ohm - OpAmp - 150ohm - 100uF - Source Direct bypass over correction - Volume potentiometer - 100uF - 470ohm - Main stage.
Last edited:
How does the Sapphire v4 handle low impedance planar magentic headphones? (not the latest high current option)
I've used my Sapphire with mostly HD650/HD800S and was looking to buy a MrSpeakers Aeon Flow.
The specs there are 14ohm and ~92db/mW..
Gain calculator shows 14db which is close to what I have already.
Are there any concerns with the current capabilities?
I've used my Sapphire with mostly HD650/HD800S and was looking to buy a MrSpeakers Aeon Flow.
The specs there are 14ohm and ~92db/mW..
Gain calculator shows 14db which is close to what I have already.
Are there any concerns with the current capabilities?
14 ohms is quite punishing, even if the efficiency is reasonable. You can see from the attached image that the output power of the Sapphire in the standard 37.5 mA bias configuration drops off considerably below 60 ohms.
There is nothing about the circuit that makes it incapable of driving these loads with low distortion however, it's just that raising the bias current would be optimal.
There is nothing about the circuit that makes it incapable of driving these loads with low distortion however, it's just that raising the bias current would be optimal.
Attachments
OK I took a look at the notes in the spreadsheet and seems like I need to change R11, R12.
A few questions:
1) Based on your graph, 75mA seems insufficient still. Since it's hard to see where 14ohm lands on the graph, what would be the optimal value? I assume somewhere between 100-150 ?
2) I assume that even if I go for 150mA, the default R17/R18 of 4.75 should remain?
3) Is the gain calculation the same after changing R11/R12?
EDIT:
Just stumbled upon #5443967 and saw the values in the updated BOM.
I will follow that for 150mA...would appreciate any other thoughts worth considering.
Thanks!
Last edited:
My answer I tried to post earlier got zapped as diyaudio went offline for a while.
Anyway ... the latest BOM you can download from my website gives more info on the bias current mods. There is no change between it and the 4.1t BOM, other than the electrolytic capacitors have been reduced in number and increased in size.
But, no, no one "needs" more than a couple of mW of headphone output power. Much like solid state power amplifiers, its a question of headroom and dynamic reserve rather than clipping distortion. So it's a matter of will it sound better, which in turn is a question of whether the additional distortion and noise from running at higher bias current will or will not balance the reduction in distortion achieved by the increased headroom.
My recommendation stands: for 60-300 ohms, use the standard configuration. for 16*-60 ohms, use 75 mA.
*14 or 15 ohms, also. Close enough.
pcb-sapphire-42b2-bom
New BOM for review. I've changed the Power Options sheet to remove the 150 mA version, keeping the 37.5 mA and 75 mA version.
I've heard the story that the burger restaurants offer a choice of three sizes as a psychological ploy in order to upsell customers to at least buying the medium-size burger.
For the Sapphire amp, the same psychology was operating in a negative sense. The additional choice was causing indecision and confusion. It would be better to settle on just "regular" and "large" options, rather than "small" "medium" and "large". That way most people would go with the normal bias mode, and anyone who didn't would have a clear single choice remaining.
Also, the scaling was wrong anyway. As usual I was overthinking the problem. And 12 W heat dissipation for the stereo circuit at 150 mA is kinda nuts. 75 mA provides plenty of power into low impedance loads. Let's leave it at that.
New BOM for review. I've changed the Power Options sheet to remove the 150 mA version, keeping the 37.5 mA and 75 mA version.
I've heard the story that the burger restaurants offer a choice of three sizes as a psychological ploy in order to upsell customers to at least buying the medium-size burger.
For the Sapphire amp, the same psychology was operating in a negative sense. The additional choice was causing indecision and confusion. It would be better to settle on just "regular" and "large" options, rather than "small" "medium" and "large". That way most people would go with the normal bias mode, and anyone who didn't would have a clear single choice remaining.
Also, the scaling was wrong anyway. As usual I was overthinking the problem. And 12 W heat dissipation for the stereo circuit at 150 mA is kinda nuts. 75 mA provides plenty of power into low impedance loads. Let's leave it at that.
Thanks for the update Richard.. the latest BOM is clearer.
Reading the power graph posted earlier, I was under the impression that both 37.5 and 75mA version would perform the same way for loads above 60ohm, thus the 75mA would be a more versatile choice for any headphone from 16-300ohms.
Does the 37.5mA perform "better" for high impedance cans because the extra headroom is not needed and distortion is added?
Reading the power graph posted earlier, I was under the impression that both 37.5 and 75mA version would perform the same way for loads above 60ohm, thus the 75mA would be a more versatile choice for any headphone from 16-300ohms.
Does the 37.5mA perform "better" for high impedance cans because the extra headroom is not needed and distortion is added?
@itsikhefez
1. That's a good way of looking at it: increasing the bias current widens the "useable" range of headphone impedance to include smaller values. While above 100 ohms or so the clipping level is unchanged with bias current, the distortion levels at lower output power does change with bias current. The relationship between distortion and current and load is complicated enough that I don't feel confident in trying to map it out, but essentially the optimal point is moved to larger output power as the bias current increases.
2. Formally only in the input coupling cap, C1, is in the signal path. I don't count signal return currents for this, so that's not to say that the others are unimportant ...
1. That's a good way of looking at it: increasing the bias current widens the "useable" range of headphone impedance to include smaller values. While above 100 ohms or so the clipping level is unchanged with bias current, the distortion levels at lower output power does change with bias current. The relationship between distortion and current and load is complicated enough that I don't feel confident in trying to map it out, but essentially the optimal point is moved to larger output power as the bias current increases.
2. Formally only in the input coupling cap, C1, is in the signal path. I don't count signal return currents for this, so that's not to say that the others are unimportant ...
Linear supplies are much less problematic re. RFI and sound good. I wouldn't recommend switching supplies, but you can do it if you want. Just remove the Z-reg section and use the main amplifier circuit powered directly from +/-12 V rails.