Hiraga "Le Monstre"

So I have finally got the hum/buzz down to about 1mV which is just about acceptable for general use in my system.
I rebuilt the power supply to have a single star ground which shaved a few mV of hum off.
I redressed the earth return wires to take them further away from the power transformer.
I changed the earthing at the RCA socket such that the socket earth went to the star ground and the internal signal wires shield terminated at the circuit board earth - helped a little.
I placed snubbers on the secondary of the transformer and the diodes which cleaned up the supply lines a lot - but not much impact on the buzz.
All of these changes cleaned up the hash on the output a lot, with each modification producing a flatter sharper scope trace - but there was still a significant first quadrant spike of about 3mV which would not go away. It turned out that this was power transformer ringing produced by charging spikes. I decided to place 0.5R resistors on each of the secondary legs which killed about 2mV of hum. So I have about 1mV of spike left, and I feel that this can only be cured by taking the transformer out of the case, which I am unwilling to do. I may try placing a ceramic snubber on the first cap to see if it bleeds off any more spike - but I doubt it will.

Sounding fine and clean and extended, with a nice rounded presentation. Its now become a keeper.

Shoog
You can put the transformer in steel box. : D
 
Have you tried moving/rotating the transformer?
Few times it helped me to get rid of small hum residuals - the field it radiates is often not uniform so by rotating the toroid we can direct the strongest EM emission away from sensitive parts of the circuit.

Can try, but surprisingly difficult with the stiff secondaries.
Starting to think that this fetish for overspec'd transformers is a fundamentally wrong headed approach. If the transformer was about 1/3 - 1/2th the size the internal resistance would present an inbuilt choke to the ringing and the supply would be naturally quiet.

Shoog
 
Can try, but surprisingly difficult with the stiff secondaries.
Starting to think that this fetish for overspec'd transformers is a fundamentally wrong headed approach. If the transformer was about 1/3 - 1/2th the size the internal resistance would present an inbuilt choke to the ringing and the supply would be naturally quiet.

Shoog

Le Monstre 8W Test Board. (maximum self noises level -110 db, small 0,18F filter, transformer 2*18V*4,4A, Resistors 0,25W 0,1% 5 ppm, ECAP on board - Nichicon VZ 470uF*25V+Epcos MKT 1,5uF*250V and mica 560 pF on input RC, bias 0,7A).
 

Attachments

  • Le Monstre 8W EVAL-BOARD v1-0.jpg
    Le Monstre 8W EVAL-BOARD v1-0.jpg
    432.6 KB · Views: 807
  • Le Monstre 8W EB2 2.JPG
    Le Monstre 8W EB2 2.JPG
    433.9 KB · Views: 760
  • _MG_2609.JPG
    _MG_2609.JPG
    524.2 KB · Views: 695
  • _MG_2610.JPG
    _MG_2610.JPG
    598.6 KB · Views: 702
Starting to think that this fetish for overspec'd transformers is a fundamentally wrong headed approach. If the transformer was about 1/3 - 1/2th the size the internal resistance would present an inbuilt choke to the ringing and the supply would be naturally quiet.

+1

As soon as I get some time I'm going to be building some choke input supplies. Same mass of iron should provide a much better result: more regulation; less switching noise and less stress all round
 
I built a DC blocking box and just added 10R to the primary side AC. This coupled with 0.5R of resistance in each leg of the secondary has got the hum down to a liveable level.

I am still a bit disappointed that I haven't got this a quiet as my valve amps, but the performance is good enough. I may look into replacing the power transformer at some point.

Shoog
 
I built a DC blocking box and just added 10R to the primary side AC. This coupled with 0.5R of resistance in each leg of the secondary has got the hum down to a liveable level.

I am still a bit disappointed that I haven't got this a quiet as my valve amps, but the performance is good enough. I may look into replacing the power transformer at some point.

Shoog

Resistor in CRC must be 1 om minimum.
I'll make 1,32F or 1,65F summary filter with buffer's batteries 12V 4,5A*h. Transformer will be 320W 2*15V*10,7A.
 
+1

As soon as I get some time I'm going to be building some choke input supplies. Same mass of iron should provide a much better result: more regulation; less switching noise and less stress all round

+1

Would the attached schematic work ? It works great (CLC) in reducing hum with tube amps and it did the trick when I built the De-Lite amp.

With a bias current of 1A/ch would the amount of capacitance be enough ?

Thanks,
Eric
 

Attachments

  • 8w PS  for 2 channels.gif
    8w PS for 2 channels.gif
    13 KB · Views: 608
+1

Would the attached schematic work ? It works great (CLC) in reducing hum with tube amps and it did the trick when I built the De-Lite amp.

With a bias current of 1A/ch would the amount of capacitance be enough ?

Thanks,
Eric

Use PSU2 to model it and see what you get at the output. You want the ripple to be less than 1mV which is a tall order indeed. You probably need more capacitance - but the good news is capacitance at these voltages is relatively cheap.

Shoog
 
+1

Would the attached schematic work ? It works great (CLC) in reducing hum with tube amps and it did the trick when I built the De-Lite amp.

With a bias current of 1A/ch would the amount of capacitance be enough ?

Thanks,
Eric
Having an inductor with a DC resistance of 4 ohms is way way too much. (is't it 0.4 ohms? Can read the schema good enough)
The CLC values look good although for a first cell a lower C would be good (reduce inrush). having CLCRC with R=0,2 ohm is also good imho.
Of course it is small but the L does a good job.
Hiraga had in the base concept something like 3x33.000 uF but these were the big screw can types - nowadays everyone has smallish units that even boast a good low ESR on paper - but in practice they cannot sustain a high ripple (for instance in a first cell).
 
Just bought a pair of LeMonstre kits off ebay (JIM audio). While searching for info about building these amps, I ran across a post saying the input FET's could be easily damaged due to electrostatic discharge. :mad: And the ebay picture shows the "matched pair" fets in a clear poly bag....(ESD?! :eek:)

These FET's are not commonly available any more, do you guys think I should get a couple of sets as replacement, just in case? Though these replacements will also be from JimAudio and will probably arrive in the same clear poly-bags as well....:(
 
Thanks Eric, that made me feel much better. (However I still wonder if they just "removed the FET's from an old polybag and wrapped them in aluminum foil"......)

A question about thermo management: do 2SB716/2SD756 get hot? should I add small heatsinks just in case?

Heatsinks for 2SD756+2SB716 not need if voltage not bigger than +-16V. Cold drivers lead to worse quality sound. Normal driver's temperature must be +35...+50 C (+95...+122 F).
 
Here is a PCB layout created my friend(also a member here).Can anyone verify this layout.I am going to order few PCBs.Also I could not find reasonably priced authentic source for matched 2sk170bl/2sj74bl



Regards,
Sachin

1) Using SQP series resistors not good, better to use 5*MFR-2 1% 100 ppm (or better ppm)
2) Power lines must be go from one point.
3) You must to set LPF (for example 1k2+470 pF[WimaFKP or Mica], where capacitor).
4) 2k (potentials), 10om (feedback), 47k (input resist) resistors and cap of LPF go to one point of signal ground.
5) Power ground from bypass go to one point, this point to do under input point (signal ground) on 3 mm and connect them by thin line (width 0,4 mm)
6) Caps of bypass must be bigger than 47uF (220-1000 uF) and very good quality (as example Nichicon KZ, FG, KA or Elna Silmic II, ROA).
7) All resistors must be not worse than 0,5% 50 ppm (better to use 0,1% 15 ppm).