Q17 - an audiophile approach to perfect sound

Hello,

For this fall, a new version of the Q17-Turbo is in preparation.

New PCB design as the Q17-Mini, new input circuit with a HPF / LPF and various other improvements.

Regards,

Stef.
 

Attachments

  • Q17-TURBO-3D-VIEW.jpg
    Q17-TURBO-3D-VIEW.jpg
    182.5 KB · Views: 376
  • Q17-TURBO-schematic.jpg
    Q17-TURBO-schematic.jpg
    420.7 KB · Views: 411
Hello,

Some news about the PCB of the future Q17-Turbo v2.

I worked on the ground plane on the top. This new ground plane should allow to remove R32 from the original schematic and to improve the SNR. I also added a double footprint for C3 and C6. We find less and less 47uF with 5mm pitch. I added the 2.5mm pitch. I also optimized the placement of the components everywhere above and below. I also put all the components to the correct size on the 3D model.

More in the next episode.

Regards,

Stef.
 

Attachments

  • Q17-TURBO-PCB-FRONT.jpg
    Q17-TURBO-PCB-FRONT.jpg
    170.8 KB · Views: 147
  • Q17-TURBO-PCB-FRONT-green.jpg
    Q17-TURBO-PCB-FRONT-green.jpg
    261.7 KB · Views: 143
  • Q17-TURBO-3D-VIEW.jpg
    Q17-TURBO-3D-VIEW.jpg
    180.4 KB · Views: 147
Hello,

The other day, I told you about the existence of a 3uF MKP capacitor for C7. I finally cracked and ordered 2 for the future Turbo.

Cornell Dubilier - Film Capacitors POLY MET 100V 3uF - 935C1W3K-F
https://www.mouser.fr/ProductDetail/Cornell-Dubilier-CDE/935C1W3K-F?qs=BsPyPsHBXitmOy4GUpF/4Q==

They also fit in the Mini. Even better because they are shorter than the Vishay 2.2uF.

All that remains is to see what it will give in the circuit and when listening.

Stef.
 

Attachments

  • IMG_8979.jpeg
    IMG_8979.jpeg
    442.2 KB · Views: 166
  • IMG_8980.jpeg
    IMG_8980.jpeg
    520.8 KB · Views: 166
  • IMG_8977.jpeg
    IMG_8977.jpeg
    951.9 KB · Views: 138
  • IMG_8976.jpeg
    IMG_8976.jpeg
    570.8 KB · Views: 165
Hello,

I started testing the first Q17-Turbo v2 prototype.

This prototype allowed me to adjust the circuit configuration and optimize the PCB. I changed some component values after testing and added the measurements on the diagram.

The specs are good and the power with the circuit configuration used will be 150W under 8R (with 60Vdc at input and 1.2V sensibility). It clip at 35Vrms.

I still have a lot of testing to do and some optimizations to try but it seems to be going well.

Regards,

Stef.
 

Attachments

  • IMG_8999.jpeg
    IMG_8999.jpeg
    424.5 KB · Views: 249
  • Q17-Turbo-2.0_spectrum_1W_OPA828_SMD_Q1_Q4.jpg
    Q17-Turbo-2.0_spectrum_1W_OPA828_SMD_Q1_Q4.jpg
    383 KB · Views: 248
  • Q17-TURBO-schematic.jpg
    Q17-TURBO-schematic.jpg
    428.5 KB · Views: 262
Last edited:
Hello,

A quick call for advice.

To optimize the input ground plane, I am hesitating between these two versions.

Version A with a ground plan under the op-amp is the current prototype version.

Version B is a version without the ground plane above the op-amp.

Which version would be preferable to minimize op-amp disturbances?

Regards,

Stef.
 

Attachments

  • input-B.jpg
    input-B.jpg
    179.1 KB · Views: 86
  • input-A.jpg
    input-A.jpg
    131 KB · Views: 80
I'm trying to take maximum precautions against ground loops and op-amp pollution by ground coming from the input because I don't know yet what the Turbo v2 will give without R32. I haven't tested it in stereo yet with one PSU for both boards. Not before October. I'm missing parts to build two boards.

In the meantime, in mono, the Turbo works well. It's been running in real time for 2 days with background music on a speaker. Very good bass control. I think it should satisfy owners of speakers with a tortuous impedance.
 
  • Like
Reactions: inyashd
Hello!

I have published on my Github repository a maintenance version 3.0.3 of the Q17-Mini.

https://github.com/stefaweb/Q17-Amplifier/tree/main/Q17-Mini-3.0

Regards,

Stef.



Version [3.0.3] (14-09-2024)
  • Reduced hole size for R26, R27, R28 and R31.
  • Added a dual footprint with lead spacing of 2.5mm and 5.0mm for C3 and C6.
  • Changed R15 to 270R.
  • Modified ground plane connection for C2.
 
Hello,

I found the time to make a drilling plan for the Q17-Mini v3 on an aluminum plate in Type-T Aluminum angle or Type-F Aluminum angle.

The minimum length to be able to place a board is 80mm and the plan of the aluminum roof is at least 44mm wide.

Have fun!

Stef.
 

Attachments

  • Drilling diagram Q17-Mini v3.jpg
    Drilling diagram Q17-Mini v3.jpg
    25.9 KB · Views: 137
  • IMG_9027.jpeg
    IMG_9027.jpeg
    472 KB · Views: 132
Hello,

While waiting for the prototypes to progress, I prepared a mounting plate for the Q17-Turbo. It is 216 x 80 mm and 5 mm thick with 4 large holes for M4 screws (and 2 more holes in the middle if your plate bends a little). The other holes are tapped M3.

I made a .dxf file which is on my Github repository. I entered it on the site of a metal cutting seller. He swallowed it without flinching. 😉

To be tested.

Regards,

Stef.
 

Attachments

  • Mounting Plate-Q17-Turbo-216x80-1.0.png
    Mounting Plate-Q17-Turbo-216x80-1.0.png
    48.7 KB · Views: 100
  • IMG_9052.jpeg
    IMG_9052.jpeg
    285.1 KB · Views: 108
Last edited:
  • Like
Reactions: Yurik_V and inyashd
Hello,

I built both boards of the Q17-Turbo v2. It seems to work well in stereo.

I will let it run like this for a while. I still need to try it with a real power supply. For now, it is still connected to the lab power supply.

I made some modifications to the low level ground plane on the PCB to improve the 50Hz (or 60Hz) rejection. The 50Hz is now at -113dBu. The final version of the PCB will be v2.0.1.

Regards,

Stef.
 

Attachments

  • Q17-Turbo-2.0_spectrum-2.0.1.jpg
    Q17-Turbo-2.0_spectrum-2.0.1.jpg
    426.9 KB · Views: 117
  • Q17-TURBO-PCB.jpg
    Q17-TURBO-PCB.jpg
    667.1 KB · Views: 124
  • IMG_9070.jpeg
    IMG_9070.jpeg
    1.2 MB · Views: 120
Me again.

If you solder the mosfet Q1 and Q4 in SMD under the PCB, I advise you to test them immediately after soldering them. Be aware that if the smosfet are poorly soldered or non-functional, when switching on, you risk sending 50/55V directly into the opamp.

Another tip, always first test the board without the power transistors. If the board is functional, it draws around 11mA per rail. Then solder Q5 and Q6, the board should draw around 90/100mA per rail. With the 4 big mosfet, it should draw around 110/130mA per rail when warm. If when cold, you draw more than 250mA per rail, you have a problem with Q5/Q6 or the big mosfet.

Some pictures of my various tests on the proto 2.0.

Stef.


In guest start, one of the opamps that I fried. 😈
 

Attachments

  • IMG_8946.jpeg
    IMG_8946.jpeg
    966.2 KB · Views: 141
  • IMG_9005.jpeg
    IMG_9005.jpeg
    1.1 MB · Views: 130
  • IMG_9008.jpeg
    IMG_9008.jpeg
    1.2 MB · Views: 130
  • IMG_9062.jpeg
    IMG_9062.jpeg
    1.1 MB · Views: 138
  • sacrifice.jpg
    sacrifice.jpg
    300.2 KB · Views: 137
  • Like
Reactions: inyashd