OK, since you put it that way it looks as though I will have to build them!
Any chance for a link to a larger drawing of the original karlsonator to facilitate scaling?
Thanks!
The link to the reference drawing by GregB is in post #1 and also repeated below. It should be high enough resolution to build from.
https://lh3.googleusercontent.com/-IsPCPSK6_lQ/TtU92CD8zMI/AAAAAAAABik/txissf4J4PI/w678-h524-no/Karlsonator.png
You can download the 1/4 scale drawing from the 'more' in the menu at the top. Click "download photo":
Karlsonator.
There are scale verification bars if you want to print it out full size or whatever.
Karlsonator.
There are scale verification bars if you want to print it out full size or whatever.
A few weeks ago I put together some mini karsonators with the omnes bb4.al, a variation of the chr70. The build is a sandwich of foamcore and thin lauan ply that i had lying around, sealed with silicon, filled with old wool sweaters and polyfill, and exceedingly ugly. These are destined to stay in my studio.
I was not expecting much and that was pretty much what i got. After removing some stuffing they sounded better although still colored and a bit boomy. Today, however, i hooked them up to my YJ 3116 stock amp with matching looks and was given quite a shock. They sounded great! Still colored with a little extra punch in the bass but natural and really enjoyable to listen to. That was replacing a firstwatt F2 which sounds incredible with my fostex and my monacors! The bb4.al had been in my too small bib cabs built for the fe108es.
There must be some cheap and ugly synergy between the amp and drivers. I was ready to bin the cabs but it now looks as if a proper birch ply build is in the cards. And the amp which sounded ok but harsh with the fostex will be receiving some mods and perhaps even a case!
Thanks again X and Greg B.! Now I also need to build some tabaqs for my other pair to compare.
If you plan on keeping the YJ TPA3116D2 amp, there is a simple mod that will transform it 10x better. Swap out the two power rail caps with Panasonic OSCON's part 330uF 25V SEPF. Those cost $2ea but well worth their price. You will get deeper more powerful bass, cleaner and more articulate percussion and highs. That is the one must have mod.
If you are more advertursome, there is the bootstrap snubber mod - look in wiki. That will transform the HF's. Cheap parts wise but more soldering skill needed.
Those two mods will make the $13 YJ amp sound like a $400 amp.
If you like the sound of the CHR70, try the Dayton RS100-4 (aluninum cone) or Dayton PS95-8 - both have some top end sizzle.
Hi X, I've been meaning to do these mods but...
Are these the caps?
Isn't changing the coils also recommended?
LOL, that's the thing about the CHR70 sound I don't like. I'm experimenting with taming the ~10k breakup with some dammar at the moment. Maybe a bit too dull now, but it's not fully dry.
The whole cone or strategic spots?
I can't bring myself to Dammar my FF125K's.
Kinda hard to get a new pair if I mess-up or don't like it. These two peaks around 7-10kHz can get a bit annoying at times. Diluted rubber cement on the aluminum dustcap should be easily removable if I were to try anything.
Last edited:
The whole thing. Seeing as they are aluminum cones, it's reversible in this case. If they remain too dull after curing, I may pull the dammar off the main cones.
It's understandable that you don't want to muck up your now extinct FF125K. I did dammar mine, but used very thin varnish, and the change was subtle. Not sure about the rubber cement unless you just want to kill the highs. Sometimes the little felt disks used for bottoms of vases etc are useful. My old FF125 center-caps were obviously doped with something from the back; they made a dull thud when tapped rather than a ping.
It's understandable that you don't want to muck up your now extinct FF125K. I did dammar mine, but used very thin varnish, and the change was subtle. Not sure about the rubber cement unless you just want to kill the highs. Sometimes the little felt disks used for bottoms of vases etc are useful. My old FF125 center-caps were obviously doped with something from the back; they made a dull thud when tapped rather than a ping.
I am just testing my new FF105WK's and there are a couple of serious peaks and a small plateau even. It's all there in manufacturer's spec sheet - just did not realize how much it bothers me until I listen to it. It can be easily EQ'd out, but then why not just get a paper cone driver like the RS100P-4 that is smoother, or the Bery smooth fiberglass TG9FD, etc ? Or of you have more money, get the 10F/8424. The aluminum cone RS100-4 can be tamed with strips of tape, but I think rubber cement may work well and be reversible.
Hi X, have you tried a current source with the ff105wk? My fe108es and, to a lesser extent, my ff165k are unlistenable when not using my f2. I have tried using added resistance with my chip amp and 3116 but did not get anywhere near the same results. I suppose a tube amp would work as well.
Hi Greg, which revision chr70 do you have, not sure where the BB4.al slots in? I wasn't bothered by any sizzle or sibilance but just hf peakiness in general and lots of mid bass "boominess". Much better with the tpa3116 and if it improves with the mods I'll be very pleased. Will have to hook up my ACA's to the mini-karlsonators,very likely a nice match.
I'd expect most driver/enclosure combinations to sound terrible on current drive. The saner schemes would have to be low Qms drivers in sealed/aperiodic enclosures or huge horns, or whatever can produce an impedance profile with broad and low magnitude humps instead of the more usual shapr peaks. Basically, if the damping is not done electrically, it must be done mechanically.
Here is the result of a current-drive chip amp on TB W4-1320SIF in DCR and FF125K (no W, the old series) in 8L reflex:
It sounded as terrible as it looks in the mid/upper bass region.
The W4-1320SIF did fare better because of its relatively lower Qm versus the Fostex's high value. The added high-frequencies above ~10kHz due to voice-coil inductance was also not necessary in both cases IMO.Attachments
Last edited:
IG,
Where does one get a current drive chip amp? I thought the Pass F1 was one of the few current drive amps out there for DIY. I am really surprised that the current driver produces the huge mid bass peak at 100Hz and 120Hz. What causes that? You are getting the some serious 7kHz and 10kHz peaks on the FF125K that are reminiscent of the peaks on the FF105WK that I had to EQ out to make it listenable for more than 5 minutes. At first it sounds interesting and full of "detail" but it wears on me fast.
Where does one get a current drive chip amp? I thought the Pass F1 was one of the few current drive amps out there for DIY. I am really surprised that the current driver produces the huge mid bass peak at 100Hz and 120Hz. What causes that? You are getting the some serious 7kHz and 10kHz peaks on the FF125K that are reminiscent of the peaks on the FF105WK that I had to EQ out to make it listenable for more than 5 minutes. At first it sounds interesting and full of "detail" but it wears on me fast.
Last edited:
Wow, I'm surprised by your results with the little Fostex (I have no experience with Tang Band).
I don't have any measurements but my fe108es and ff165k, both in BIB cabs, sound MUCH better in all regards with my f2 amp. More bass, better highs, more natural sounding in general. It's not even close.
I don't have any measurements but my fe108es and ff165k, both in BIB cabs, sound MUCH better in all regards with my f2 amp. More bass, better highs, more natural sounding in general. It's not even close.
I would be interested in your take on all these drivers with a current source but shipping alone would cost more than most of the drivers. Any Virginia area DIYers willing to help out?
I would be interested in your take on all these drivers with a current source but shipping alone would cost more than most of the drivers. Any Virginia area DIYers willing to help out?
Here is my FF105WK in a FAST with XO at 225Hz. There is a -3dB baffle step compensation high-shelf Q=0.5 EQ applied at 750Hz. The blue is the non-EQ'd for peaks and the black curve is after EQ applied to tame the peaks. That is a 10dB peak at 7.8kHz - and the 5dB bump at 1.5kHz is very fatiguing with female vocals. Grey is TC9FD without EQ.
Here's a very advanced CAD rendering:
I just did that on an LMxxxx I have, I forget the exact part #. Not ideal as it picks up more noise, but works enough for the testing purpose. Mixed mode feedback would be better to produce a given output impedance value. Setting it at ~80ohm might be enough to simulate current drive at ~10X load impedance.
As voltage will track impedance at a constant current, impedance peaks will produce the most power. In the case of the Fostex, the Z peaks may be around 75-100ohm IIRC, so it figures there would be a huge rise in output.
http://sound.westhost.com/project56.htm
Attachments
Last edited: