to MJK and all the others that had the luck to have both whizzered and not whizzered cones OB, like lowthers and FF85K:
apart from the classic difference one can expects from having a whizzer, how does the back wave from the not whizzered drivers contribute to the sound? shouldn't the high frequencies firing backward be controlled some way? thing that the whizzer already does? acting more like an OB mid driver with only front firing highs?
i found was a discussion of some years ago on this thing, i understand it's a matter of recipes and tastes, that is indeed why i think a description from who has tested both can be helpfull
apart from the classic difference one can expects from having a whizzer, how does the back wave from the not whizzered drivers contribute to the sound? shouldn't the high frequencies firing backward be controlled some way? thing that the whizzer already does? acting more like an OB mid driver with only front firing highs?
i found was a discussion of some years ago on this thing, i understand it's a matter of recipes and tastes, that is indeed why i think a description from who has tested both can be helpfull
OK, I'm fully satisfied with the OB using the FF85k. I sprang for the Solen caps and Erse inductors for the HP side of the XO, but tried using a plate amp for the LP side and to drive the woofers. The plate amp has an inherent hum which is objectionable, so now I'm looking to purchase the parts for the LP side of the XO. I can't go cheap on the inductor--it has to be the Erse as specified by MJK in order to maintain the correct resistance (and there's no way around cost on inductors anyway.)
But for the caps, I'm wondering if I can cut cost without destroying the sound. Parts Express has the 68 uF 400 V Solen cap for $24.31; they also have a Dayton DMPC-68 68 uF 250 V cap for $15.37 and a cheapie non-polarized 68 uF 100 V cap for $1.47. So how low can I go before I'd seriously regret it.
Cheers, Jim
But for the caps, I'm wondering if I can cut cost without destroying the sound. Parts Express has the 68 uF 400 V Solen cap for $24.31; they also have a Dayton DMPC-68 68 uF 250 V cap for $15.37 and a cheapie non-polarized 68 uF 100 V cap for $1.47. So how low can I go before I'd seriously regret it.
Cheers, Jim
Jim, I am satisfied with Erse inductors and caps in my OB Bass setup with Alphas.
The caps are I think even better tolerance. I am using a pair in my phono preamp now-not bad!
If it were me, I would make sure the crossoversound is correct with BP electroyltics before spending much on PolyPro caps.
Erse makes bipolars as well, of course there is also Mundorf ones (not too expensive from partsconnexion) and of course blackgate, but for bass, I would not go overkill.
The caps are I think even better tolerance. I am using a pair in my phono preamp now-not bad!
If it were me, I would make sure the crossoversound is correct with BP electroyltics before spending much on PolyPro caps.
Erse makes bipolars as well, of course there is also Mundorf ones (not too expensive from partsconnexion) and of course blackgate, but for bass, I would not go overkill.
Tea-Bag,
Thanks for the response. I knew I didn't want to go cheap on the HP side, but wanted to check out more closely before spending almost another $100 for the LP side. This since I may or may not stay w/ this design, I am trying to keep costs down. Next summer I will be trying out an H-baffle for the Alpha 15's with the FF85k on OB; since the HP side of the XO should be usable there, I didn't question spending there, but I will obviously need a new LP XO--the more costly side, of course.
Cheers, Jim
Thanks for the response. I knew I didn't want to go cheap on the HP side, but wanted to check out more closely before spending almost another $100 for the LP side. This since I may or may not stay w/ this design, I am trying to keep costs down. Next summer I will be trying out an H-baffle for the Alpha 15's with the FF85k on OB; since the HP side of the XO should be usable there, I didn't question spending there, but I will obviously need a new LP XO--the more costly side, of course.
Cheers, Jim
I guess it's a good thing I waited on the LP side of the XO. It has only just now occurred to me that what I built isn't exactly the straight OB MJK laid out. It's more of a U-baffle. 
The sides extend 14.5" at the bottom and are 6" deep at the top of the woofer area; over the top of the woofer is a 2x4 (which is 3.5" deep.) So upon reflection, I think I will try for a passive XO closer to 150-160 Hz. I also looked at the XO setting on the plate amp--that is, actually thought about where I had it set; when tuning the XO and the volume on the plate amp, I wasn't thinking about WHERE it was set, just moving it around until I got the freqs balanced. Perhaps procrastination can beneficial sometimes. 
The sides extend 14.5" at the bottom and are 6" deep at the top of the woofer area; over the top of the woofer is a 2x4 (which is 3.5" deep.) So upon reflection, I think I will try for a passive XO closer to 150-160 Hz. I also looked at the XO setting on the plate amp--that is, actually thought about where I had it set; when tuning the XO and the volume on the plate amp, I wasn't thinking about WHERE it was set, just moving it around until I got the freqs balanced. Perhaps procrastination can beneficial sometimes.
Final comments about my OB build:
I have finished the LP side of the XO using a 12 mH inductor and an 80 uF (cheap, non-polarized) cap. These values are different from Martin’s original plans. Due to the addition of the side pieces, what I made is more like a U-baffle than the straight OB. Thus the XO needed to shift down toward 150 Hz from the original target of 200 Hz. I didn’t need to make any adjustment of level between the Alpha 15a and the FF85k.
I tried switching the polarity of the FF85k back and forth between normal and reversed. Textbook recommendation is to reverse polarity for a second order XO, but to try both ways to see which you prefer. I prefer Martin’s recommendation of matching polarity of the Alpha and the FF85k. With polarity reversed there seemed to be a bit of scratchiness in the lower end of the vocal range. My wife agreed, so I don’t think I’m imagining it.
And I must say that these are the best sounding speakers I have build thus far.
WOW! (I just had to say it one more time.)
Cheers, Jim
I have finished the LP side of the XO using a 12 mH inductor and an 80 uF (cheap, non-polarized) cap. These values are different from Martin’s original plans. Due to the addition of the side pieces, what I made is more like a U-baffle than the straight OB. Thus the XO needed to shift down toward 150 Hz from the original target of 200 Hz. I didn’t need to make any adjustment of level between the Alpha 15a and the FF85k.
I tried switching the polarity of the FF85k back and forth between normal and reversed. Textbook recommendation is to reverse polarity for a second order XO, but to try both ways to see which you prefer. I prefer Martin’s recommendation of matching polarity of the Alpha and the FF85k. With polarity reversed there seemed to be a bit of scratchiness in the lower end of the vocal range. My wife agreed, so I don’t think I’m imagining it.
And I must say that these are the best sounding speakers I have build thus far.
WOW!
(I just had to say it one more time.)Cheers, Jim
ping planet 10
I have built MJK's OB. While I liked it very much, I cannot help but think
that it could be improved. Not reinvent the wheel, but upgrade from pressed
steel with plastic hubcaps to some nice alloys.
I would like to model with the Beta 15 higher up on the baffle with wings.
This would give better integration between the drivers as they would act as
a point source. Because of the 12/db/octave xovers and the distance between
the drivers, and because the two drivers neatly divide the bass from the
mids/treble info., one really gets the bass from down low and the highs from above, and I did not enjoy this aspect of the sound. Likie Lampizator, I do not like sound coming to me at my knees.
the freq. response was
incredibly smooth though.
I have a 92.5 db efficiency mid/high driver that i want to use (8 ohm, qts.1)
Another reason I want to use the Beta, besides efficiency, is to make the
speaker more tube friendly. The low damping of the tube amp effectively
adds .15 to the qts of the driver, making the beta more OB suitable. I would not mind
24" baffles (up 4 inches from the org. design)
I have only 12-15 watts, and the Alpha 15 design did not QUITE cut it. Almost
though.
Another idea I am keen upon, and do not know if Martin's program will model, is to
use a 1st order xover on the fullranger. Since baffle rolloff is 6 db/octave, and 1st order
rolloff is 6 db/octave, it seems that with the upper portion of the baffle trimmed down
correctly one would have the correct 12 db./octave rolloff to match the rolloff of the
woofer.
I have a mac. and so cannot use Mathcad. Anyone willing to model some baffles for me? I will Pay Martin the $25 gladly.
I have built MJK's OB. While I liked it very much, I cannot help but think
that it could be improved. Not reinvent the wheel, but upgrade from pressed
steel with plastic hubcaps to some nice alloys.
I would like to model with the Beta 15 higher up on the baffle with wings.
This would give better integration between the drivers as they would act as
a point source. Because of the 12/db/octave xovers and the distance between
the drivers, and because the two drivers neatly divide the bass from the
mids/treble info., one really gets the bass from down low and the highs from above, and I did not enjoy this aspect of the sound. Likie Lampizator, I do not like sound coming to me at my knees.
the freq. response was
incredibly smooth though.
I have a 92.5 db efficiency mid/high driver that i want to use (8 ohm, qts.1)
Another reason I want to use the Beta, besides efficiency, is to make the
speaker more tube friendly. The low damping of the tube amp effectively
adds .15 to the qts of the driver, making the beta more OB suitable. I would not mind
24" baffles (up 4 inches from the org. design)
I have only 12-15 watts, and the Alpha 15 design did not QUITE cut it. Almost
though.
Another idea I am keen upon, and do not know if Martin's program will model, is to
use a 1st order xover on the fullranger. Since baffle rolloff is 6 db/octave, and 1st order
rolloff is 6 db/octave, it seems that with the upper portion of the baffle trimmed down
correctly one would have the correct 12 db./octave rolloff to match the rolloff of the
woofer.
I have a mac. and so cannot use Mathcad. Anyone willing to model some baffles for me? I will Pay Martin the $25 gladly.
Last edited:
Another driver worth modeling would be the Delta LFA 15". Why are people
not using this driver on OB? Although it cost $40 more than the Alpha it goes
very low and is ruler flat up to @1000 hz. But the efficiency is much higher (95 av.)
Is it because this is a much heavier driver and so detail will be lost in the bass?
Once again, if anyone can simulate for me I would be glad to pay Martin. You can use
my experience, having already built the Alpha/fe103 version.
I have no interest in using two woofers, like the Emerald Physics. I just stacked two
Dyanco a25 speakers (with the tweeter disconnected on one), and although the bass did come up,
it became boomy and the clarity suffered. I know, these cross over much higher (1600 hz.)
than the Alpha/fe103, but even so ------
not using this driver on OB? Although it cost $40 more than the Alpha it goes
very low and is ruler flat up to @1000 hz. But the efficiency is much higher (95 av.)
Is it because this is a much heavier driver and so detail will be lost in the bass?
Once again, if anyone can simulate for me I would be glad to pay Martin. You can use
my experience, having already built the Alpha/fe103 version.
I have no interest in using two woofers, like the Emerald Physics. I just stacked two
Dyanco a25 speakers (with the tweeter disconnected on one), and although the bass did come up,
it became boomy and the clarity suffered. I know, these cross over much higher (1600 hz.)
than the Alpha/fe103, but even so ------
Last edited:
ping planet 10
Another driver I would love to model if I get a PC, or would
love to have someone model for me, is the Delta LFA15. This driver averages
95 db. effiency, goes as low as the Alpha, but costs $40 more. It is ruler flat up to 1000 hz. too.
Why is no one using this driver? Is it because it is heavier, so some low level
detail would be lost in the bass?
Specs on the fullrange I want to use: Fs 47, qts 1, vas 58 L., efficiency maybe 91 or 92, perfect
for the Beta 15.
Another driver I would love to model if I get a PC, or would
love to have someone model for me, is the Delta LFA15. This driver averages
95 db. effiency, goes as low as the Alpha, but costs $40 more. It is ruler flat up to 1000 hz. too.
Why is no one using this driver? Is it because it is heavier, so some low level
detail would be lost in the bass?
Specs on the fullrange I want to use: Fs 47, qts 1, vas 58 L., efficiency maybe 91 or 92, perfect
for the Beta 15.
Last edited:
ping planet 10
Thanks for the replies.
Martin uses high damping SS amps to control the small magnet Alphas.
He uses the Alphas to minimize the width of the baffles. And he uses them
to provide deeper bass. By his own words, these speakers are designed for
SS amps. With a tube amp the bass is looser and "rolling".
But the Beta goes almost as low on the same baffle according to his sims.
And will go lower still with a low damping factor tube amp. And will go lower still
with a slightly wider baffle and some small wings.
This can be done.
Thanks for the replies.
Martin uses high damping SS amps to control the small magnet Alphas.
He uses the Alphas to minimize the width of the baffles. And he uses them
to provide deeper bass. By his own words, these speakers are designed for
SS amps. With a tube amp the bass is looser and "rolling".
But the Beta goes almost as low on the same baffle according to his sims.
And will go lower still with a low damping factor tube amp. And will go lower still
with a slightly wider baffle and some small wings.
This can be done.
Last edited:
Dave,
The OX on the FF85k is exactly as Martin recommends in his paper: 24 uF and 4 mH. I did ask him about the XO, as the FF85k has a higher F(s) than the FE103e in his original design. The 500 Hz XO point is sufficiently above F(s) that the impedance rise doesn’t cause a problem.
I might also mention that the room where I have the OBs is a bit bright, so I have the treble control on the Yamaha CR-1000 set for -1 with a turn over point of 5 KHz--this probably translates to -1 dB at 10 Khz.
If you find a better solution for the LP side of the XO, I would appreciate hearing about it. What I did sounds good to me, but I didn’t want to buy pricey extra parts to play around with it. And in fact I went with the cheap cap for it ($2), figuring it would have minimal impact on the sound quality. (I know that on the HP side, the quality of the cap will be audible, so I didn’t cheap out on that one.)
For those who may be interested but have not yet checked current prices, this is what they cost from Parts Express :
4 mH Erse Super-Q $19.14
12 mH Erse Super-Q $22.38
24 uF Solen cap $11.33
82 uF Solen Cap $28.05
Which is about $81 per side for the XO. The drivers cost me about $36 each for the FF85k and $58 each for the Alpha 15a. So the XO is a major expense!
And thanks again Dave for bringing the FF85k to my attention. That is one SWEET driver!
Cheers, Jim
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.