Volume level mismatch using MOSFET output device?

[zebra100]

Hi,

I have a Class A headphone amp which uses a power MOSFET (IRF513) as the output device and an LM317 as a constant current source for the output. The problem I have is that the right channel output is noticeably lower than the left channel. The sound quality is not compromised in any way, other than the volume.

Could it be the MOSFET at fault, or the LM317?

Thanks,

Chris.

[zebra100]

Well, I've ordered some new MOSFET's anyway. They should take a couple of days to get here. I'm hoping it's not the LM317's. The amp is a Szekeres design with a constant current source. It is fitted into my passive pre for my LM3875 Gainclone. I have a switch to select between either the input to the headphone amp or the RCA output's to the Gainclone. The headphone amp is supplied by a 13.8.v regulated PSU, which is separately switched. Volume is taken care of by an Alp's Blue 50k dual log pot with a 47k shunt resistor. All internal wiring is 22awg silver plated copper with teflon insulation.

[lineup]

could you point to the circuit

I suspect the mismatch is in the potentiometer.
At least if the 2 mosfets are reasonable same.

Expensive poteniometers can have dsual gang matched down to 1%
5% should be close enough.


One other possibility, as good and probable as anything,
is that your hearing is imperfect.
In fact almost no living person has same hearing on both ears!
Even if not damaged by loud sounds or at work, we can graduly get different curves in our ears.
I think you should test your ears.
Goto your nearest general doctors office.
Order one professional hearing test.

Me myself, always have to adjust my potentiometer, volume comtrol to one side
when listening in my headphones.

Regars from
and old man with both tinnitus and a certain hearing defects

[zebra100]

Hi Lineup,

I had my hearing checked not that long ago so I hope it's not that. I don't think it's the pot either. It's an Alps Blue 50k with a 47k metal film resistor shunt (1% tolerance). I rememeber reading somewhere that the MOSFET's should have a r(ds) 0.74 ohms (?)

I measure r(ds) of 5M ohms on both FET's, and r(gd) of 1.2k ohms on one FET and 0.73k ohms on the other. Something strange going on.

I have 4 FET's to try. I think what I'll do is measure all 4, then select the closest matching pair, see how I get on.

The amp design is from here.
http://www.headwize.com/projects/sho...e1_add_prj.htm

It's about 2 thirds down the page.

Regards,

Chris.

[zebra100]

Well,

None of the MOSFET measure the same r(gd), however no matter which cct you put the MOSFET in, it is still the right channel that is low. This leads me to believe that it could be the CCS (LM317) that is at fault?

I will swap them across later today.

Chris.

[zebra100]

Infact, I think I'll probably rejig the whole board layout and start again once I get the new MOSFETS. The current boards are a little cramped. That way, I can be sure everything else is ok. It also means that I can do away with the internal heat sinks and just use the case. If it is one of the LM317, I'll have to wait untill I can get a new one.


Chris.

[coloradosound]

Swap your LM317's and see if the prob. follows!

[AndrewT]

Hi,
measure the voltage across the current setting resistor.
The two channels should have ~1250mV and the same value of resistor.

Measure the output voltage AC & DC when the amp input is shorted to signal ground.

However, the gain of the stage is not affected by small or even medium changes in the CCS current.

The CCS current determines the maximum output current swing.
For headphone use that should never be a problem. They will deafen you before they demand too much current.