CFH7 Amp

xrk971 that is one impressive Class A mono block amp. :)
I think you should build that to the market.
 

Attachments

  • image_2021-11-22_175031.jpg
    image_2021-11-22_175031.jpg
    174.9 KB · Views: 307
next try ...

which one would you prefer ?
plane or no plane

with an additional plane on TOP layer i have to place tons of vias to connect the planes together

regarding skrews ...
yes the 3 drivers are a bit painfull

the mosfets are fixed with PCB to heatsink ... there is no problem i think.

to make it easier -
1: place parts below PCB wihout soldering ,
2: mount to heatsink
3: solder all parts
4: cut the pins



with 68Vpp @ ~72mA per Fet ( x3 )

Harmonic Frequency Fourier Normalized Phase Normalized
Number [Hz] Component Component [degree] Phase [deg]
1 1.000e+03 3.445e+01 1.000e+00 -0.21° 0.00°
2 2.000e+03 9.526e-05 2.765e-06 17.87° 18.09°
3 3.000e+03 1.291e-04 3.746e-06 92.28° 92.50°
4 4.000e+03 6.962e-05 2.021e-06 177.26° 177.47°
5 5.000e+03 1.093e-04 3.172e-06 -96.03° -95.81°
6 6.000e+03 2.433e-05 7.063e-07 -179.17° -178.96°
7 7.000e+03 2.511e-05 7.289e-07 -101.71° -101.49°
8 8.000e+03 2.619e-05 7.601e-07 -177.53° -177.32°
9 9.000e+03 1.906e-05 5.532e-07 -99.87° -99.66°
Total Harmonic Distortion: 0.000614%(0.000000%)
 

Attachments

  • CHR11_woPlane.png
    CHR11_woPlane.png
    227.8 KB · Views: 288
  • CHR11_Plane.png
    CHR11_Plane.png
    274.6 KB · Views: 243
Founder of XSA-Labs
Joined 2012
Paid Member
I like the SMT layout but not sure if everyone appreciates it as much here. I find it quicker to build actually and makes for a compact layout. If going SMT maybe use SMT SOT24 for the TO92 and use SOT223 for the drivers with PCB as heatsink. The VBe multiplier temp sensor still needs to be TO126. Can you post a JPG of PDF of the schematic with easy to read component symbols so we don’t have to go into LTSpice to check? I work on DIYA threads mostly from my phone screen.
 
Last edited:
That was also an Idea to change the Transistors to SMD variants
But which types

Yes .. with 0402/0603 SMD parts of IS much smaller
But i would leave the components Like original.


Resistors are 1206 .. so quite big ..
Caps are FKP or MKP WIMA types.

X5R oder X7R are Not so good for audio.
NP0 types are quite expensive or only OK for very small values.




Red is bottom layer and blue ist top layer

Yes i will Post a Picture or PDF later...
 
Yes.. bc856/846 .. or similar
Better would be a dual Part Like
BC846BPDW1..



5401 /5551 .. i will Check.
An additional Copper or Aluminium Part between heatsink and SMD Part on bottom Side for thermal coupling.

Or an Second PCB Like an PCB Module on PCB.
Depends on height of the 5401/5551 parts


For me .. Sounds Like a good Plan
 

Attachments

  • BC846BPDW1T1-D.PDF
    210.5 KB · Views: 79
Last edited:
dual BC846BDW is still available ...
seems that is a very good replacement for typical 546/556

simulation is ok .. not optimized
too tired today ^^

see attached files... i hope .models are ok

replacement for BD139 is BCP56 ... but how to connect to heatsink...

PBHV8540/PBHV9040 are not so good ...
DZT5551/DZT5401 are ok ..
but not as good as KSA/KCS

PZTA42/PZTA92 ( MPSA SMD types )


i learned today:
these drivers are important for audio ...
 

Attachments

  • CHF11_v2_PCB.PDF
    197.5 KB · Views: 108
  • CFH11-Cx-xrk971_v3_SMD.asc
    19.6 KB · Views: 79
Last edited:
Founder of XSA-Labs
Joined 2012
Paid Member
I think the radial bulk metal power resistors you had before are fine (like KOA BPR series) It’s what I prefer since it can radiate heat and is very low distortion. You don’t want metal oxide film for MOSFET source resistors. They add third harmonic distortion. The SMTs also have to conduct heat to the PCB as heatsink.

Rest looks good. I agree Wima no caps are fine for bypass. Although power bypass with X7R 0.1uF 100v is really actually very good as they are low ESR. It’s audio path that I avoid X7R. Like compensation caps in the pF range then use NP0/C0G ceramic.
 
Member
Joined 2014
Paid Member
Nice work Saddevil! I like the usage of the dual BJT transistor, stock might be an issue in these times though :(
I'd tend to agree with X about using radial/axial leaded source resistors instead of SMD's.
Maybe include a few voltage test point locations on the pcb to check bias current, post capMx voltage, etc..?
:cheers: