holland
Headphoneus Supremus
- Joined
- Jun 15, 2007
- Posts
- 2,210
- Likes
- 22
Gain too high (hisssss)? Turn on pop annoying? Easy fix.
First, I replaced the batteries from 2 2600mAh AA to 3 800mAh AAA, to get the voltage up a bit (not related to title) from 2x1.2 to 3x1.2. Even with that change, I've been running the amp in unattended mode since Sunday afternoon non-stop, and it's still not dead. The charge LED will become useless though. There's not enough voltage variance with the charger to turn on the LED when the batteries aren't dead.
I recapped the electrolytics and lowered the gain on my amp. No more turn on pop and I have more usable volume control.
Original items:
(2) 1000uf 6.3V Nichicon VR(M)
(2) 2.2uf 50V RSA
(1) 10uf 50V
(2) 100uf 10V caps
(2) 40.2K SMD resistors
I replaced it with Panasonic FM/FC and Panasonic 0805 SMD resistors.
(2) 1000uf 16V FM (doesn't quite fit into the case so I did a naughty and bent the leads and stuck them out). I'll look for better a fitting cap some other time.
(2) 330uf 6.3V FM or 220uf 10V FM (you'll see that changing the gain would allow for a smaller bypass cap).
(1) 10uf 50V FC
(2) 100uf 10V FM
(2) 24K SMD resistors for approximately 5x gain.
Why did I change the 2.2uf to 330uf?
The turn on pop is directly related to
10/(Cbypass * 250K-Ohm) <= 1/ ((Rfeedback + Rinput) * Cinput). This means the Cinput is charged before Cbypass is charged. With the original values it wasn't. With the higher value cap though, the startup time is faster as well, I believe (and definitely seems that way).
Original
Cbypass = 2.2uf
Rfeedback = 40.2K
Rinput = approx 4410 (measured)
Cinput = 100uf
New
Cbypass = 330uf
Rfeedback = 24K
Rinput = approx 4410 (measured) -- untouched
Cinput = 100uf
Why didn't I change Cinput.
Rinput and Cinput forms a high pass filter for the input, so I didn't want to mess with that. The center frequency is 1/(2*pi*R*C).
The 1000uf on output forms a lowpass and is a DC blocking cap. It is mandatory. The chip has a default DC offset of Vdd/2.
Again, same center frequency calculation. 1000uf is fine there for me, R is your headphone R, but that changes with frequency.
All the electrolytics listed, except for the 10uf is in the signal path. The 10uf is a power bypass cap. I had a tant on hand, but I didn't replace it with that.
What's it sound like? I don't know, it seems somewhat similar to pre-mod, but I must admit I didn't use the amp much, the hiss and pop was too annoying. It seriously needs a grounded metal/conductive plastic case to reduce RFI. I have to stick the amp into an anti-static bag at work to be anywhere useful. I think the lowered gain would help there.
I just did the mod and everything seems to be OK. My primary goals were to eliminate the pop and add one more battery for a voltage increase. Needless to say, any warranties are null and void, but that's insignificant to me.
Gain is Rfeedback/Rinput.
Original is 9x:
Rfeedback = 40.2K
Rinput = approximately 4410
New is approximately 5-5.5x
Rfeedback = 24K
Rinput = approximately 4410
First, I replaced the batteries from 2 2600mAh AA to 3 800mAh AAA, to get the voltage up a bit (not related to title) from 2x1.2 to 3x1.2. Even with that change, I've been running the amp in unattended mode since Sunday afternoon non-stop, and it's still not dead. The charge LED will become useless though. There's not enough voltage variance with the charger to turn on the LED when the batteries aren't dead.
I recapped the electrolytics and lowered the gain on my amp. No more turn on pop and I have more usable volume control.
Original items:
(2) 1000uf 6.3V Nichicon VR(M)
(2) 2.2uf 50V RSA
(1) 10uf 50V
(2) 100uf 10V caps
(2) 40.2K SMD resistors
I replaced it with Panasonic FM/FC and Panasonic 0805 SMD resistors.
(2) 1000uf 16V FM (doesn't quite fit into the case so I did a naughty and bent the leads and stuck them out). I'll look for better a fitting cap some other time.
(2) 330uf 6.3V FM or 220uf 10V FM (you'll see that changing the gain would allow for a smaller bypass cap).
(1) 10uf 50V FC
(2) 100uf 10V FM
(2) 24K SMD resistors for approximately 5x gain.
Why did I change the 2.2uf to 330uf?
The turn on pop is directly related to
10/(Cbypass * 250K-Ohm) <= 1/ ((Rfeedback + Rinput) * Cinput). This means the Cinput is charged before Cbypass is charged. With the original values it wasn't. With the higher value cap though, the startup time is faster as well, I believe (and definitely seems that way).
Original
Cbypass = 2.2uf
Rfeedback = 40.2K
Rinput = approx 4410 (measured)
Cinput = 100uf
New
Cbypass = 330uf
Rfeedback = 24K
Rinput = approx 4410 (measured) -- untouched
Cinput = 100uf
Why didn't I change Cinput.
Rinput and Cinput forms a high pass filter for the input, so I didn't want to mess with that. The center frequency is 1/(2*pi*R*C).
The 1000uf on output forms a lowpass and is a DC blocking cap. It is mandatory. The chip has a default DC offset of Vdd/2.
All the electrolytics listed, except for the 10uf is in the signal path. The 10uf is a power bypass cap. I had a tant on hand, but I didn't replace it with that.
What's it sound like? I don't know, it seems somewhat similar to pre-mod, but I must admit I didn't use the amp much, the hiss and pop was too annoying. It seriously needs a grounded metal/conductive plastic case to reduce RFI. I have to stick the amp into an anti-static bag at work to be anywhere useful. I think the lowered gain would help there.
I just did the mod and everything seems to be OK. My primary goals were to eliminate the pop and add one more battery for a voltage increase. Needless to say, any warranties are null and void, but that's insignificant to me.
Gain is Rfeedback/Rinput.
Original is 9x:
Rfeedback = 40.2K
Rinput = approximately 4410
New is approximately 5-5.5x
Rfeedback = 24K
Rinput = approximately 4410