Frugal DOA twins ...

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

tv

Well-known member
Joined
Feb 22, 2006
Messages
1,459
PREFACE


Many moons ago, the old venerable Melcor 1731 had a secret love affair.

Fueled by pure love and passion, the two lived only for the moment, knowing there was no future for them. She was a Class-A lady after all, already promised to somebody else. And he knew that. Being just a simple rocker, old Melcor did his best to rock his lady as hard and as long as he possibly could. And he did so indeed, much to her pleasure. It was quite electric.

But then, a disaster happened. His lady got pregnant. Not knowing what to do, Melcor ran away, leaving his lady alone.

In total secrecy, she gave birth to two nearly-identical baby twins.

------------------------

Many years later, the twins learned who their real parents were.
Now they're here to claim what rightfully belongs to them.


Enter MelChiOr 1337 and MeJLHior 1337.


P.s.: Be careful when you meet them. Growing up without their parents, they developed a complicated, self-destructive personality. Just like kamikaze, they can hurt you badly. Handle with care!
 

Attachments

  • schematics_bw.png
    schematics_bw.png
    9.9 KB · Views: 473
FFT comparison ...
 

Attachments

  • fft_curves_simulated.png
    fft_curves_simulated.png
    16.1 KB · Views: 271
THD+N comparison ...
 

Attachments

  • thdn_comparison.txt
    2.7 KB · Views: 114
Perfboard LAYOUTS and TRACES ...
 

Attachments

  • layouts_traces.png
    layouts_traces.png
    30.3 KB · Views: 358
INSTRUCTIONS and EULA ...
 

Attachments

  • instructions_eula.txt
    7.4 KB · Views: 201
SQUARE WAVE simulation comparison ...

It shows ever-so-slightly improved and more symmetrical ramping response in both new circuits, meaning your transients will be nicely "ADSR-ed", transforming your signal into a nicely slewed, syruppy avocado smoothie.

Also shown is an affect of changing both 220pF capacitors, as mentioned in "instructions", on transient response. This means you could to some extent tune the overshoots for even more shaping.
 

Attachments

  • square_wave_simulation.png
    square_wave_simulation.png
    26.1 KB · Views: 154
Fantastic stuff, can't wait to try this out!

Has anyone found the time to plonk this on perf yet?
 
Very cool.  What wattage do you recommend for the parts?  The reason I ask is that this could also very easily be built using SMD components, assuming you could find SMD equivalents to the transistors.  I did some searching and couldn't find SMD packages for those transistors, but that doesn't mean they don't exist...
 
For all resistors but R11/12 or R12/R13 respectively you should be able to use standard 0805 packages with normally 1/10W or 1/8W.

Transistor equivalents for BC550/560 are BC850/860. The output transistors might be replaced using BCP53/BCP56 or MJD243/253 (but there are many others...). Care should be taken to get away the heat. To achieve the nominal maximum dissipation a copper pad has be placed under the metal tab.

Also, the SOT223 package needs actually way more space on the board than the TO126. So the through hole transistors not only have much better convection, but are more space friendly too. And for a DOA the third dimension really is your friend when laying out the pcb.
 
@millzners: have you read "instructions and eula"?

1) under normal conditions (f.e. nominal 600ohm load), these will work with standard components (0,25W resistors, transistors as specified), except output resistors, which should be 0,5W for longevity. Use of 1%metal-film resistors is recommended (usually they come rated at 0.6W), and is almost "selbst-verstandlich" for any DIY audio... Use small NP0/COG disk ceramics for "pF" caps and a multilayer ceramic for "nF" cap.

2) for SMD, use BD... TO126 as output (must be heatsinked) and BC850/860 as SOT23 variants of BC... devices. Resistors can be any SMD 0,25Watters. For output resistors, you can combine in parallel several 1603 resistors to acquire the power rating.


That's all there is to it EXCEPT if you foresee the possibility of getting the output SHORTED while running at full tilt ... because these opamps don't include the output current limiters (short-circuit protection) - which is somewhat counter-intuitive in "pure" Class-A circuits.

In such case, you have two options:
1) use 0,5W resistors as "fuses" (they will "smoke" ... )
2) use higher power rating for output resistors (f.e. some Caddocks) and use bigger heatsinks for output transistors.

In short: don't short the output and you'll be fine .. or use some buildout resistor.
 
DISSIPATION graphs ...

(simulation for MelChiOr 1337, the more "power hungry" of the two, but not much)


NOTE: both simulations imply the 14.77dBu output level (12Vp/24Vpp into 600 Ohm load), but the latter graphs show what would happen if the output was SHORTED at that same level AND 50/50 duty cycle (which is important if you think about it).

In other words: For output resistors, use at least 2W rating to survive AC short. For DC-output short, use at least 3W resistors. (Or limit the supply current @ 200mA).

For normal operation, any 0,5W resistor would be OK.
 

Attachments

  • dissipation.png
    dissipation.png
    62.7 KB · Views: 90
I have a 1:5 input tx sitting around waiting for someone to play with.... if I were to use one of the twins here in 312 configuration, what, if anything, would you recommend changing in that preamp design?
 
I can't think of any "necessary" modifications or changes to API circuits.

But, make sure you breadboard the opamp first and make sure that the quiescent current is setup like it should be, as explained in "instructions".

 
Sorry, probably another stupid question... in case I fry my only pair of 139/140s, can 135/136 handle this role? I have a bunch left over from other builds.  I noted the instructions saying a range of 200-250 hfe is needed, so the 135/136 are good there, but the max power dis is 1.25W compared to the 12.5W of the 139/140 pair...

Thanks again for your time with this! :)

EDIT: never mind, I see now that peterC's old design used the 135/136 pair...

sorry, should have looked before asking. :)
 
Back
Top