Mini Karlsonator (0.53X) with Dual TC9FDs

Founder of XSA-Labs
Joined 2012
Paid Member
So I was listening to Gretchen Parlato on my 0.40x's (3FE25) powered by my Silicon Harmony SE Class A *headphone* amp, and it sounded just glorious. I was curious so I put my measurement rig in place with a 8w power resistor and the Focusrite 2i4 as the interface and here is the FFT I get in REW (at the same volume level I listen at - about 0.71vrms). So one look at the FFT shows that it is dominant second harmonic, with a tiny bit of third, and nothing else, and the noise floor is flat and clean.

Silicon Harmony is the one in the case with fins:
attachment.php


707mV rms into 8ohms:
attachment.php


I agree with Bonjonno, these make excellent computer monitors.
 

Attachments

  • SH-0.707vrms-8ohms-FFT.jpg
    SH-0.707vrms-8ohms-FFT.jpg
    128.7 KB · Views: 379
  • SH-0.40x-MK.jpg
    SH-0.40x-MK.jpg
    237 KB · Views: 371
xrk971- I finally got the parts and assembled Juma's Easy Peasy Cap Multiplier. My hands were shaking through the whole soldering process. Luckily none of the delicate parts smoked. I'm running an HP 130 watt 19V laptop SMPS into the cap multiplier into a TDA7297 into MK 0.41X's w/ TC9FD18's. Like you said it would, the bass has improved. It's actually hard to believe the sound that is coming out of a few dollars worth of parts and 3.5" speakers. Just listened to Jaco playing on Joni Mitchell's version of Goodbye Pork Pie Hat; it was a visceral, life affirming experience and I was grinning like a kid with an ice cream cone. Two cones. You mentioned also a block of 20 or so 1000 uF cheap caps for added zest. Are they just wired in parallel? And then where do they go in the circuit exactly? Thanks again for the great guidance.

Jon, do you have the schematic for your cap multiplier? I think the schematic on the first page of that thread looks different than the one you built? Thank you.
 
Founder of XSA-Labs
Joined 2012
Paid Member
I don't think it is different - just use the critical part of the cap mx: the 220uF cap, 10k resistor, 1n400x diode, 220R gate stopper, 1uF and 1R5 snubber. The 4700uF output cap is optional - use whatever you have or the rail cap on the actual amp. Also, any N-channel enhancement mode MOSFET will work (IRFP240, IRF610, IRF510, etc). Even a little ZVN4306GTA and all SMT components can be used to make one that fits on a dime:

633425d1504437434-xrk971-pocket-class-headamp-gb-img_7677-jpg


This one is courtesy of RaptoLightning who provides Gerbers for ordering your own - these are cap mx for my little pocket class A amp.
 
that's right. I followed xrk971's instructions as best I could. Just seven parts altogether. It would be the top right quadrant of the original schematic, minus the 4700uF cap. I used an IRF610 and a 1N4007. Put in my order for a bag of 50 1000uF caps and when that gets here I can build the cap array to complete the trifecta.
 
I see your aperture goes a lot higher than the one I have, following the sketches I found here for 0.4x. (unless I completely miscalculated it... which has been happening in the past! I'm far from fool-proof!)

Could it be the reason why I think it sounds boxy and muffled?
 

Attachments

  • IMG_20171003_131305.jpg
    IMG_20171003_131305.jpg
    287.7 KB · Views: 117
Here's the FR and distortion for the miniK I built.
Not super clean.

I include the TABAQ FR and distortion as comparison. It needs a BSC to bring the top end down, but as you can see, it still is as loud as a miniK 0.4, even when adding a BSC.
 

Attachments

  • miniK FR+dist.jpg
    miniK FR+dist.jpg
    110.7 KB · Views: 359
  • tabaq FR+dist.jpg
    tabaq FR+dist.jpg
    110.9 KB · Views: 70
Founder of XSA-Labs
Joined 2012
Paid Member
That looks nice Perceval. If you can show us a straight frontal perspective so we can see how big the aperture is relative to the driver cone. The height mainly controls the bass extension and power and the width of the aperture controls the mids and HF transmission. Try to make the aperture big enough so that it intersects the driver surround as the curved wing profile crosses through the driver face. That should open it up - it is sensitive so maybe 5mm to 10mm more removed from the aperture should do the trick.

attachment.php


It looks like you may have some rattles/buzz on your box, are these two measurements same drive voltage and mic at same distance away?
 
Last edited:
I want to thank X for submitting this design.
Many people do indeed seem to like it very much, but, unless there's something I can change, it is not for my own ears.

I have been enjoying the open sound that OB provides, but the miniK 0.4 with TC9 sounds stuffed, boxy and muffled to my ears. Plus, I guess the aperture "projects" the sound more than regular boxes without the aperture, seems good for sounds reaching farther away.

But for me, one of the reason why I prefer OB is I like the sound to be ambient, not "projected".

It's ok. I'm glad I built it and tried something else. I've been wanting to try this miniK for a long time.

In the end, I will keep the TABAQ, as I feel they are more balanced, and my, wonderful to my ears, OBs.
 
It looks like you may have some rattles/buzz on your box, are these two measurements same drive voltage and mic at same distance away?

Yes, forgot to mention... measurements were with the exact same setup, minutes apart with nothing touched besides swapping the speakers to place them at the same place, and moving the mic up and down a bit to be in front of the driver. Mic is at 1m.

I'll try opening the aperture a bit more, see if that takes care of the boxy sound.

You don't get that kind of distortion with your miniKs? I thought it might me the design, as nothing rattles in the box.
 
Founder of XSA-Labs
Joined 2012
Paid Member
So a thought just occurred to me. People who make this speaker out of foam core or cardboard seem to have a much better sound. It may be that wood is the culprit with the boxy sound. The foam core actually re-radiates some of the sound, possibly contributing to more passive radiation like a guitar body. Just a thought. If you have some cardboard - try making a quick and dirty one and see how it sounds. The foam core also seems to not have as much of an issue with the W notches you have with wood. Although Jim Shearer reported excellent results from his wooden 0.53x Karlsonator with dual TC9FDs earlier in the thread. I just notice both you and emuMannen reported closed boxy sound and both your builds are wood.
 
Last edited:
I don't think it is different - just use the critical part of the cap mx: the 220uF cap, 10k resistor, 1n400x diode, 220R gate stopper, 1uF and 1R5 snubber. The 4700uF output cap is optional - use whatever you have or the rail cap on the actual amp. Also, any N-channel enhancement mode MOSFET will work (IRFP240, IRF610, IRF510, etc). Even a little ZVN4306GTA and all SMT components can be used to make one that fits on a dime:

This one is courtesy of RaptoLightning who provides Gerbers for ordering your own - these are cap mx for my little pocket class A amp.

that's right. I followed xrk971's instructions as best I could. Just seven parts altogether. It would be the top right quadrant of the original schematic, minus the 4700uF cap. I used an IRF610 and a 1N4007. Put in my order for a bag of 50 1000uF caps and when that gets here I can build the cap array to complete the trifecta.
Gotcha. The schematic is for dual mono dual rail, our TDA is stereo single rail. What is the watt rating for the resistor? Is 1uF cap critical? Can I go a little bigger or lower? I kinda like the idea of using TO220 like Jon did, is that enough heatsink for IRF610? What other MOSFET can I use? Or, teach me how to find correct equivalent for the MOSFET in that position. Thanks so much.
 
Last edited:
So a thought just occurred to me. People who make this speaker out of foam core or cardboard seem to have a much better sound. It may be that wood is the culprit with the boxy sound. The foam core actually re-radiates some of the sound, possibly contributing to more passive radiation like a guitar body. Just a thought. If you have some cardboard - try making a quick and dirty one and see how it sounds. The foam core also seems to not have as much of an issue with the W notches you have with wood. Although Jim Shearer reported excellent results from his wooden 0.53x Karlsonator with dual TC9FDs earlier in the thread. I just notice both you and emuMannen reported closed boxy sound and both your builds are wood.

Hi, for what it's worth, I build mine out of wood, 0.4x with one TC9. Not boxy at all, no funny resonances or terrible peak that I notice. The inside is lined pretty heavily with 8mm felt.