F5 with 2SK2013/2SJ313

[juma]

What's your hands-on experience with Keratherm or similar heat conductors?

Keratherm is good but you don't need it with plastic cased 2sk2013/2sj313. Any insulator pad (no matter how good) is just another thermal resistance that impedes heat transfer between silicon and the heatsink.
As for my experience, I easily get 5 degrees C difference between power device and heatsink with thermal paste and mica (where needed).

Is that graph "inconsistency" visible above 40-50kHz, the result of audio analyser quantization noise?

Anchan uses SMPS as PSU.

 

[Extreme_Boky]

Thanks Juma, much appreciated.

 

[Anchan]

Is that graph "inconsistency" visible above 40-50kHz, the result of audio analyser quantization noise?

I'm not really sure. Those were my first ever readings right out of the box. I was so excited to get my first readings, but I realize now there are all kinds of reasons for all the spikes.

• I had the analyzer sitting right on top of the amp- USB cable draped right on top of the power supply practically
• I also had all kinds of electrical stuff going on very near proximity - window AC, electronic dimmer in the light right next to me

I'm gonna redo it under a little closer to "lab conditions." I'll also zoom in on some other frequencies, especially lower down near 60 and above.
I also realize that this unit is cool for learning and to get an idea, but in the end is a $200 instrument.

>>>
Quote:
Originally Posted by Extreme_Boky View Post
Is that graph "inconsistency" visible above 40-50kHz, the result of audio analyser quantization noise?
Anchan uses SMPS as PSU.
<<<<

I know there is a heavy bias against switchers, but this is a pretty unique model and others have used them here with success. I put the unit on a regular scope and couldn't see anything below 200khz. Nonetheless, it warrants further investigation

 

[Anchan]

Measurements part 2

So, I redid the measurements- this time without having the distortion analyzer sitting directly on top of the amp. Also the USB cable is not draped over the power supply lol.

Attached see before and after.

 

[Anchan]

I'm not sure what some of those small peaks are below 1K.
Also, the 2H and 3H are a little higher than what's in the FW F5 manual. I'm using just one 100R feedback resistor rather than the two paralleled. I wonder if this has something to do with it?

The noise floor above 1K @ 125 - 130db looks to be mostly the distortion analyzer. Even a loopback without the amplifier connected, there's a fair amount of increasing hash up high.

Anyway, interesting to do measurements, and learn something. Curiosity satisfied!

 

[Extreme_Boky]

I'm not sure what some of those small peaks are below 1K.
Also, the 2H and 3H are a little higher than what's in the FW F5 manual. I'm using just one 100R feedback resistor rather than the two paralleled. I wonder if this has something to do with it?

The noise floor above 1K @ 125 - 130db looks to be mostly the distortion analyzer. Even a loopback without the amplifier connected, there's a fair amount of increasing hash up high.

Anyway, interesting to do measurements, and learn something. Curiosity satisfied!

"I'm using just one 100R feedback resistor rather than the two paralleled" ...so the gain of your amplifier might be 3-4dB higher? This may explain the "2H and 3H are a little higher than what's in the FW F5 manual" you are seeing...