With those amplifiers, you might want to look at something a bit less picky about the amps they use than the Fostex (and that can benefit from the power you have (will have) on tap... don't limit yourself to what Wilmslow sells, lots of good Euro suppliers, and it seems a fair amount of kit arrives into the UK from North America.
A current fave seems to be the EL70 Met (i know 1 guy who has built 3 sets of bipoles so far)
dave
A current fave seems to be the EL70 Met (i know 1 guy who has built 3 sets of bipoles so far)
dave
Hi Dave
EL70 seem to be pitched at a good price and design detail available in the metronome table. Thanks for that. just need to find a supplier.
I'm curious about what makes fostex speakers "picky" about the amps that drive them.
BTW fired up my TDA2050 amp project for the first time last night. sprung into life at the first attempt. very rewarding. Only tested it on some cheapo technics speakers sofar. looking forward to seeing how it drives my rogers. Not expecting miracles but will be interesting anyway.
Regards G.
EL70 seem to be pitched at a good price and design detail available in the metronome table. Thanks for that. just need to find a supplier.
I'm curious about what makes fostex speakers "picky" about the amps that drive them.
BTW fired up my TDA2050 amp project for the first time last night. sprung into life at the first attempt. very rewarding. Only tested it on some cheapo technics speakers sofar. looking forward to seeing how it drives my rogers. Not expecting miracles but will be interesting anyway.
Regards G.
Others will pipe in of course, but I'd not describe as picky, as much as "truthy"
For example, the dynamics, forward presentation and rising response of the higher sensitivity FExx6 series (whose measured FR can cause some folks anxiety depending on the degree of smoothing) can be less than forgiving of high noise floors /hum levels etc., and revealing of dynamic compression or constriction of "soundstage dimensionality" (if you believe in that sort of subjective mumbo jumbo 
) I particularly found the FE126E to be fatiguing on certain budget class T amps, but far more relaxed on lower power SETs ("euphonic distortions" and all)I was using a 120w Sansui (rebuilt and recapped in the past year) while building a SET for my Saburos with FE126eN drivers. The Sansui sounded very good but when I recently replaced it with the SET, the music sounded so much better. I don't have the audiophile vocabulary of Dave and Chris, but I would agree with Chris that the dynamics were improved, but there is something else, maybe better control, more detail...
hmm, a lot of information happening at the deci-watt or lower level, and of course just as with modest to arc-welder SS amps of any vintage, not all low powered tube amps will necessarily have the synergy you've "stumbled upon"?
EL70 drawings got done because of the enthusiastic builder in California.
If you just want to build a monopolar Met poke me, i'm almost finished the drawings for those.
If you just want stock drivers, Henkjan in the Netherlands is probably your best bet for a supplier.
What Chris & GM said (and Flemming observed), but also the efficiency leading to typical class AB amps working where they are at their worst, right near the cross-over point.
dave
not all low powered tube amps will necessarily have the synergy you've "stumbled upon"?
Well it wasn't quite as accidental as all that, Dave told me to do it
Well it wasn't quite as accidental as all that, Dave told me to do it
Well not quite. You were already leaning that way, i just gave a push.
dave
Metronome BSC and Zobel Values
I built a pair of Metronomes for the Fostex FE108Esigma drivers recently and used the default values from the tables for the BSC and Zobel circuits. I was not pleased with the sound, way to bright, and the bass was a bit flat sounding.
So using an online BSC and Zobel calculator I came up with these values that I think sound much better: I just wanted to share these in the event they might be useful to someone else:
Rbs 1,8 ohms
Lbs 0.56 mH
Rz 8.5 ohms
Cz 0.5 uF
I built a pair of Metronomes for the Fostex FE108Esigma drivers recently and used the default values from the tables for the BSC and Zobel circuits. I was not pleased with the sound, way to bright, and the bass was a bit flat sounding.
So using an online BSC and Zobel calculator I came up with these values that I think sound much better: I just wanted to share these in the event they might be useful to someone else:
Rbs 1,8 ohms
Lbs 0.56 mH
Rz 8.5 ohms
Cz 0.5 uF
Thanks Dave, that explains a lot. I always felt that adherents to the “Tubes are best” philosophy simply enjoyed the warm colouration from the square law distortion. Cos, after all, if you could get modern amps with didly squat distortion (THD or TIMD) then what was the tube group really saying? Of course - cross over distortion! They don’t like it. So, class A bipolars being non starters with their exponential distortion Class A tube amps win out every time. But what about power MOSFETS i believe they are square law devices so are there any audiophile class A power MOSFET amps?
Probably not the right forum for this type of question so my next post will get back onto a more MET / MLTL related subject. Currently working on some images.
Regards G.
Probably not the right forum for this type of question so my next post will get back onto a more MET / MLTL related subject. Currently working on some images.
Regards G.
I remember coming across a post that discussed the idea of folding a MET. Hmm I thought, may well save you height but at the expense of a much more complicated build. Then i started thinking, what if you maintained the area of the sharp pointy end but change its aspect ratio so that its longest dimension was the same as the mating dimension of the fat blunt end. This would significantly simplify construction see pictures.
I also remember reading that TL designs that are open at the fat end are longer than straight TLs and these are themselves longer than TLs where the open end is the thin end. If this is the case and you want a shorter MET then why not box off the fat end and mount the port at the pointy end? Not having the necessary calculator i cant work out how much this would actually save you. Anyone prepared to do the sums?
The idea of changing aspect ratios but maintaining areas made me think – if there is any advantage of straight TLs (other than the simple issue of construction) and there is the advantage of the MET that it has no parallel surfaces to set up standing waves then why not combine the two. Start with a rectangular base, longest axis running from front to back and arrive at the top with a rectangle of same area and aspect ratio, longest axis running side to side. Simple construction i.e. four side panels similar to the MET but assembled with front and back panels pointing down and the two side panels pointing up. No parallel sides but a “straight design”. Ii think it would look elegant as well. What do you think see picture.
Being new to this i don’t know if someone has already done it. If so, then i apologise most sincerely, plagiarism not intended.
Regards Graham
I also remember reading that TL designs that are open at the fat end are longer than straight TLs and these are themselves longer than TLs where the open end is the thin end. If this is the case and you want a shorter MET then why not box off the fat end and mount the port at the pointy end? Not having the necessary calculator i cant work out how much this would actually save you. Anyone prepared to do the sums?
The idea of changing aspect ratios but maintaining areas made me think – if there is any advantage of straight TLs (other than the simple issue of construction) and there is the advantage of the MET that it has no parallel surfaces to set up standing waves then why not combine the two. Start with a rectangular base, longest axis running from front to back and arrive at the top with a rectangle of same area and aspect ratio, longest axis running side to side. Simple construction i.e. four side panels similar to the MET but assembled with front and back panels pointing down and the two side panels pointing up. No parallel sides but a “straight design”. Ii think it would look elegant as well. What do you think see picture.
Being new to this i don’t know if someone has already done it. If so, then i apologise most sincerely, plagiarism not intended.
Regards Graham
Graham,
One must keep in mind that single number THD and IMD numbers are completely irrelevant to evaluating amplifier sonics. One needs to look at both the quanity, and the order of individual harmonic components and the shape of the distribution to have any hope of correlating distortion to sonics. Harder yet is to evaluate these components under dynamic load. And unless you are driving a load identical to the speaker under question, any results obtained are very limited (and sometimes very misleading). Further, distortion at the levels they will be used at is more important than anywhere else. Also very important, is how an amplifier clips, and how well it recovers from clipping.
Given the difficulty of adequately quantizing all these things, and then correlating the measure set to sonics, comes the advice "to just listen to it"
dave
The planet10 monoliths (Mileva & Demetri) are what happens when you run with that idea.
Each taper ratio has its advantages and disadvantages and what works best is dependent not only on the choosen driver, but goals, aesthetics and more.
dave
To follow up on that note; having experimented with both the above - FE167 in Demetri and FE127 (& FF125K?) in Mileva, one observation was that the wide/shallow aspect ratio of the enclosures makes it very tempting to place near a back wall. For my taste the results, particularly with the 167, was excessive boundary reinforcement and somewhat boomy bass. Pulling the boxes 2-3 feet from the wall was all it took restore the balance.
Thanks Dave, I'll reserve any future comments on sonics, distortion and amps untill i'v had the chance to hear a tube setup
Your comments on the idea for a folded MET are confusing me. The Met is a dual taper design, front to back as well as side to side. If i'm not mistaken the mileva and demetri designs both have parallel front and back panels. The ideas I presented in my last post both tried to retain the dual taper concept.
WRT to my "The straight MLTL with a twist" i was obviously not wearing my maths head when i said it maintained its cross sectional area like a straight TL. It turns out that the cross sectional area at the half way point, for an aspect ratio of a (where a>1) is (a^2 + 2a + 1)/4. Does Martin Kings TL design tool cater for diverging and converging cross sectional area designs? Having produced a rough outline drawing i'm feeling inclined to having a go but help with numbers would be appreciated.
Regards G.
Your comments on the idea for a folded MET are confusing me. The Met is a dual taper design, front to back as well as side to side. If i'm not mistaken the mileva and demetri designs both have parallel front and back panels. The ideas I presented in my last post both tried to retain the dual taper concept.
WRT to my "The straight MLTL with a twist" i was obviously not wearing my maths head when i said it maintained its cross sectional area like a straight TL. It turns out that the cross sectional area at the half way point, for an aspect ratio of a (where a>1) is (a^2 + 2a + 1)/4. Does Martin Kings TL design tool cater for diverging and converging cross sectional area designs? Having produced a rough outline drawing i'm feeling inclined to having a go but help with numbers would be appreciated.
Regards G.