F4 power amplifier

Jump to Latest
#5,042
Dennis Hui said:
Koos,

Increasing the bias will not have an effect on the volume you are getting. What it does, however, is increase your class A envelope.

Your current bias of 0.06V across the 0.47 ohm source resistors give you about 2.3W (RMS) class A into 8 ohms.
The standard bias of 0.2V across the 0.47 ohm resistors give you about 25W RMS class A into 8 ohms

In addition, increasing the bias lowers the distortion and changes the distortion profile. You can see these effects in Papa's article:

https://firstwatt.com/pdf/art_leaving_class_a_2019_redux.pdf

I don't know what output level you are currently using out of the amp, but I do suspect its distortion is somewhat higher than if it were at the standard bias level.

So disadvantage of higher bias? Heat. Right now you are running at less than 1/3 the dissipation of the standard bias.

I personally would make the change to allow for higher bias. If nothing else, you can then compare to see if you actually prefer one bias level over the other.

Cheers,
Dennis
Click to expand...
This is great information Dennis, thanks a lot!

Interesting article, which is exactly what I need in my learning curve with my non-electronics background. Great to build an amp this way and learn a lot.

I will share the output voltage of the preamp once measured.

Regards,

Koos
 
#5,045
william2001 said:
I agree with Dennis. I would increase the bias by stacking another 10k resistor on your current R9 to get the 5k value. No need to unmount pcb's and all that, in my opinion. Just a simple tack soldering job. Gain won't change, but the increased bias will allow the amp to function as designed and intended.
Click to expand...
Thanks for this tip of soldering an additional 10K resistor instead of removing the existing one. This saves me lots of work! Much appreciated 🙂

Regards,

Koos
 
  • Like
Reactions: william2001
#5,049
I have my pair of Silver 10’s in an open baffle with 15 inch helper woofers below them and they are powered by an adjustable class-d plate amp. I think this is best of both worlds.

There is a forum member who shared a baffle step compensation filter for the Silver 10’s (he has almost the same baffle dimensions as mine) which I will try once Lyric Audio has replaced one of my broken drivers. At the moment of writing I am waiting close to 8 weeks for this driver to arrive.

Anyway, with the actual situation in the world right now, and being safe where I am at this moment, we should not complain about long lead times, but enjoy other things that we already have. Peace.
 
Last edited:
#5,050
Dear Friends,

Finally the resistors arrived!

I soldered an additional 10K resistor to R9. Now, I can easily reach the 0.13 volt DC as mentioned in the F4 Build Guide.

However, now I am not able to get the zero offset. 11.4 Volt DC is the lowest I can get.

Am I missing something? Please find values attached.

Regards,

Koos
 

Attachments

  • D43C6D5F-EEE7-4E6C-8E67-8517BD78E486.jpeg
    D43C6D5F-EEE7-4E6C-8E67-8517BD78E486.jpeg
    505.4 KB · Views: 84
#5,052
Hello Sir,

Thanks for your quick reply!

I will post photos of several locations. I will start with the R9. Please find attached.

Is this done correctly, the additional 10K resistor in parallel with the original?
 

Attachments

  • 839AC648-016F-4D67-B1B0-D4ED9F4077AF.png
    839AC648-016F-4D67-B1B0-D4ED9F4077AF.png
    479.5 KB · Views: 116
#5,053
And another three photos attached.

These three photos cover one complete channel. Please let me know if any detailed photo from a specific component is needed.
 

Attachments

  • 2A0E294F-4977-4ABC-8AC3-0D8C9BDE7335.jpeg
    2A0E294F-4977-4ABC-8AC3-0D8C9BDE7335.jpeg
    346.4 KB · Views: 119
  • C71CD1D5-6D94-4BC3-8D5F-583BA2F4AEC4.jpeg
    C71CD1D5-6D94-4BC3-8D5F-583BA2F4AEC4.jpeg
    399.5 KB · Views: 110
  • 953614D4-E20A-4DB2-BF1B-188496A8128E.jpeg
    953614D4-E20A-4DB2-BF1B-188496A8128E.jpeg
    377.2 KB · Views: 112
#5,054
For your information, on one photo attached the the above post you can see the leads of my multimeter across a source resistor.

The strange thing is, I only added one 10K resistor to R9. Before this change I could easily set zero offset. However, I could not get the bias higher than 0.06.
 
#5,056
I am measuring the offset between the plus and minus of the output terminals. Please find a photo attached.

This offset is as good as it gets, I am “out of turns” with the Bourns resistor, I hear a clicking noise and the value doesn’t change anymore. Before I heard the clicking noise of the Bourns resistor I was able to change the offset. However, only until 11.40 Volt DC, then the clicking noise came. Probably out of turns / end of stroke.
 

Attachments

  • 41225A26-A785-4913-8C64-9D9345FC40FE.jpeg
    41225A26-A785-4913-8C64-9D9345FC40FE.jpeg
    318 KB · Views: 83
#5,057
Acoustic Emotion said:
I am measuring the offset between the plus and minus of the output terminals. Please find a photo attached.

This offset is as good as it gets, I am “out of turns” with the Bourns resistor, I hear a clicking noise and the value doesn’t change anymore. Before I heard the clicking noise of the Bourns resistor I was able to change the offset. However, only until 11.40 Volt DC, then the clicking noise came. Probably out of turns / end of stroke.
Click to expand...
Yea, out of turns.

your original issue was beibg unable to bias it up as specified.

I’d cut the 10k stacked resistor off, and start troubleshooting from there. Just snip it, you can clean it later.

problem is therefore as follows: Why are you unable to bias it up?

PS: both channels the same, or one of them OK?

it is a pretty simple circuit, you’ll get there.

regards,
Andy
 
#5,059
Hello Andy,

Thanks for your time.

It is not the same for both channels. One channels is at around 8 Volt DC offset, while the other channel is 11.4 Volt DC offset.

@6L6 , do you agree, should I cut the additional 10K resistor off and start faultdinding from there?

It is late night here in Holland, so I will go for a horizontal rest and start fresh tomorrow after work. So, if I am not responding anymore tonight I’m still alive but asleep 😎

Thanks all for your help, will continue tomorrow!

Regards,

Koos
 
#5,060
andynor said:
You have the red dc wire @ V-, and black @ V+. Why? How are they connected to the PSU?

please include overview pics, including wiring to and from PSU and boards and speaker outs, plus inputs.
Click to expand...
You are sharp 😃 and correct. Due to insufficient length of wires I had to swap red and black for V+ and V-.

I will post the photos that you are requesting tomorrow.

Thanks for your time!
 
  • Like
Reactions: andynor
You must log in or register to reply here.
Top Bottom