2stageEF high performance class AB power amp / 200W8R / 400W4R

[astx]

Dedicated to our DIYAUDIO electronic Guru @Dave Zan: Your Idea which helped to materialize the high end power beast SA2021-NCH rocks!
Wish you where here listening with me!

Have fun, Toni

 

[astx]

SA2021-NCH will get a LED-VU meter .... remote controlled ...

 

[astx]

There are 3 PWM channels which controls RGB led stripes in the inner to show the electronics.
The LED bar can be controlled to display classic VU meter, peak only, VU and peak hold, difference, difference and peak ...
The VU scale is from +6dB down to -32 dB in 2dB steps and showing the last 4 steps in 4 dB steps down to - 48 dB.
The LED's are multiplexed 2000 times per second in 6 groups which is about 166 Hz per LED.

 

[astx]

SA2021 LED VU: Looks like it would be soon ready for production ...
Remote control works. Mode switching works, changing amplifiers inside backlight color and brightness works (can be seen at the end of the video).

VU modes:

        LED_MODE_DARK,
        LED_MODE_BAR,
        LED_MODE_BAR_AND_PEAK,
        LED_MODE_PEAK_ONLY,
        LED_MODE_DOT_ONLY,
        LED_MODE_DOT_AND_PEAK,
        LED_MODE_OVERLOAD_ONLY,

For "LED_MODE_OVERLOAD_ONLY" it can be configured, that the display only works if a minimum dB level has been reached. So normally the VU meter is dark, but just be4 clipping occurs, the peak hold starts glowing.
The 2x24 VU LED's are multiplexed in 6 x 8 matrix with a 2000 ticks task which gives a comfortable 166Hz per LED. No ghosting and no flickering. The brightness of all LED's is equal during all modes.
The board drives a 3 color RGB LED stripe which lights the inner of SA2021 tower. The brightness is not adjusted using multiplexing, instead it uses 3 PWMs (160kHz base freq.) to generate a 0-5V "DAC" output to generate a 7 - 12V voltage for the LED stripes.

Have fun, Toni

 

[astx]

Attached a video from the original project from 2011.08.04. Made homebrew single sided pcb's. It was a good learning project.
For today this is now enough retro electronics ...
I'm glad, that this long running project is now finished (first prototype was from 2011 and the basic PIC software idea too).

BTW: both videos are not "lipsync" - in real life the VU has a very fast reaction time and a perfect peak detection.

Have fun, Toni

P.S.: pcb's for the new (!) project are on stock if one is crazy enough to build this VU 😉

 

[Reina]

Interesting project.

I see an electrolytic capacitor is used in the feedback path.

Is it really beneficial to use big and expensive polypropylene capacitors for just dc-blocking at the input then?
Can't a single and small polyester type be used here?

An electrolytic capacitor with the same capacity would improve the power-on dc behaviour...

 

[astx]

Dear @Reina,

To what post/schematic do you refer - post #1?

 

[Reina]

Hi @astx ,

yes. first post. main amp schematic.
i just discovered this thread...

 

[astx]

PP is lowest distortion. To achieve the same with a lytic we would need a 200-500uF lytic. But I wanted to limit the bandwith below 4Hz.

The 470uF lytic (sometimes bypassed with a PP) in the feedback path is for DC offset to be minimized.

 

[Reina]

Hi @astx ,

its actually the question: is this lowest distortion relevant here when there is zero volt ac across the capacitor within the audio range...

On the other side: the non-linearity of the ESR of the feedback electrolytic capacitor has an influence at low frequencies.

 

[astx]

These questions are already answered e.g. in the APAD 6th edition from D. Self.
Have a look at page 297 ff

 

[xrk971]

Transistors should last "forever" - only "electro migration" problem operating the parts near limits will kill them. Maybe the Voyager space probe has no lytics on board to stabilize the on board voltages (offered by nuclear batteries). This may explain the reliability of the electronics.

SA2014 may get problems after another 20000 - 50000 hours of operating when some lytics going the path to death (dry out). But thats maybe beyond my personal lifetime and it is very likely that I can never report this event to the DIY community 🤣

Have fun, Toni

P.S.: The EVVO A1837/C4793 are looking not to be a good replacement for Toshiba A1837/C4793. The Unisonic (UTC) or KEC variants are closer to the original

Spacecraft don’t use electrolytic capacitors because they aren’t able to be used in a vacuum due to out gassing and short lifetime. Tantalum caps or MLCC’s are used instead.

 

[astx]

Link to pictures for SA2014 and SA2021-NCH

SA2014 - 4 Channel - pictures
SA2021 - NCH - picture series 1
SA2021 - NCH - picture series 2
SA2021 - NCH - picture series 3
SA2021 - NCH - picture series 4
SA2021 - NCH - picture series 5