well, more luck to have a good result with a Full range than making a two way with a large overlap and a 6 db slope (the 2.2 uf cap). Remember you have just a little IC to power that and you are near field... you don't really need a tweet ! At least that's what I think, other may have experienced different. But imho from a only listening view, better to stay FR with the ad'hoc 4" unit.
I'm working my way through a double cask matured Tamnavulin - another Speyside single malt whisky.btw , curently drinking a 14 yo Tormore
like to putt a little neutral water in the double cask's W. -especially if the second barel had some cherry in it (no more thn 20% water, it levels the flavors)
Forgott to say : in the Peereless datasheet I linked there is the size measurment as I have no idea if it fits ! But that's the spirit : no too much expensive and very good sound, anyway you won't have bass in a 1 liter cabinet...
Forgott to say : in the Peereless datasheet I linked there is the size measurment as I have no idea if it fits ! But that's the spirit : no too much expensive and very good sound, anyway you won't have bass in a 1 liter cabinet...
Last edited:
The amp under that heatsink is the old Toshiba TA8227P...what might look like an eight-pin package...but the center two pins are a one piece tab.
The TA8227P was designed to run off a 9V supply, it's rated at 2.5W...but according to the THD curves, it's more like 1W....There are far far better all-in-one amp chips currently running our white plastic PC "monitors".
So far, the best PC set I've torn into & repurposed was running the TDA1517, running at six watts, & it made my large home speakers sit up & roar.
----------------------------------------------------------------------------Rick........
The TA8227P was designed to run off a 9V supply, it's rated at 2.5W...but according to the THD curves, it's more like 1W....There are far far better all-in-one amp chips currently running our white plastic PC "monitors".
So far, the best PC set I've torn into & repurposed was running the TDA1517, running at six watts, & it made my large home speakers sit up & roar.
----------------------------------------------------------------------------Rick........
Yes, as pointless as it might be, my mind is fixated on filling that empty spot! Just knowing that the tweeter is a dummy is the kind of thing to bug me, even if it's no detriment to sound quality. Welcome to my mind
I will take your suggestion for the capacitor value though (and your scotch serving suggestions)
Is the TA8227P the kind of thing you can just swap with a drop-in replacement? And if so, would it be worth it in my case, and what chip should I use?
The classic FE103 basket has tha hole circle, but the large hole size (the blue circle in the picture) would make it hard to seal.
Are you sure 104mm is the cut-out size? The pre-2K FE103 looks to be very close to your driver — manipulating your picture to line up as best i can.
Here are the dimensions (magnet varies alot across iterations).
dave
Attachments
So finally finished my speakers. I had to wait for a few bits and pieces that I ordered to arrive, and then christmas got in the way, so it took a bit longer than expected. But I'm pretty happy with the results. They're not perfect, but they definitely have the surprising and unexpectedly high audio quality out of crummy looking beige desktop boxes that I was going for!
As you'll see, I did replace all the caps in the end. As I said, I had to order a few things anyway, and I just decided to throw the caps in to the order since they're cheap and I could probably do with some practice soldering. Besides, this way I know they're not going to die anytime soon. (And testing the old caps that came out of the board, they mostly seemed to have drifted out of spec)
However I think it's clear that the parts aren't necessarily designed to go together, as the speakers sound a little unbalanced, even to my untrained ear - I have to put the treble way up and the bass way down, and even then it sounds like there's not enough mid-tones. But I think I can make it tolerable with the mixer on my computer.
At first I thought it lacked dynamic range too (the big dramatic hits in the Tron Legacy soundtrack sounded flat), but now I'm thinking that was just because of the crummy Vibra16 soundcard I was using. I do have an AWE64 on the way though, which should mitigate that somewhat. Either way, they sound a damn sight better than they did (even if the right driver hadn't been broken).
So yeah, thanks again to everyone for your tips! I've learned a lot, and ended up with a pretty decent set of speakers for a retro computer!
As you'll see, I did replace all the caps in the end. As I said, I had to order a few things anyway, and I just decided to throw the caps in to the order since they're cheap and I could probably do with some practice soldering. Besides, this way I know they're not going to die anytime soon. (And testing the old caps that came out of the board, they mostly seemed to have drifted out of spec)
However I think it's clear that the parts aren't necessarily designed to go together, as the speakers sound a little unbalanced, even to my untrained ear - I have to put the treble way up and the bass way down, and even then it sounds like there's not enough mid-tones. But I think I can make it tolerable with the mixer on my computer.
At first I thought it lacked dynamic range too (the big dramatic hits in the Tron Legacy soundtrack sounded flat), but now I'm thinking that was just because of the crummy Vibra16 soundcard I was using. I do have an AWE64 on the way though, which should mitigate that somewhat. Either way, they sound a damn sight better than they did (even if the right driver hadn't been broken).
So yeah, thanks again to everyone for your tips! I've learned a lot, and ended up with a pretty decent set of speakers for a retro computer!
Thanks for the update. That looks like a quality job!
If you used a 1.5uF capacitor in series with the 8 ohm tweeter, you could experiment with a larger value (up to a maximum of 4.7uF) that will pass a wider band of treble energy.
And for a quick and easy experiment, you could try to tame the over-generous bass by blocking the reflex ports with bath sponge type foam or even an old pair of socks!
If you used a 1.5uF capacitor in series with the 8 ohm tweeter, you could experiment with a larger value (up to a maximum of 4.7uF) that will pass a wider band of treble energy.
And for a quick and easy experiment, you could try to tame the over-generous bass by blocking the reflex ports with bath sponge type foam or even an old pair of socks!
Thanks! I think it's a reasonable effort for a first time speaker resto. Definitely good enough for their purpose.
After more testing I'm starting to think the problem is almost like an uncanny valley effect - the sound is too good to be one thing, but not quite good enough to be another. I've tried a few different audio samples on different systems, and I'm finding that ideal levels tend to differ a lot.
(In particular I was constantly fiddling with the levels in Star Wars Episode IV every time it switched between music, dialogue and action)
I guess I'm just used to cheap desktop speakers that flatten everything out to an inoffensive beige meh. These just show up every imperfection and imbalance!
I'll see how I go with those other suggestions though (reversing leads, trying a different capacitor for the tweeters, and sticking a sponge in the reflex ports)
After more testing I'm starting to think the problem is almost like an uncanny valley effect - the sound is too good to be one thing, but not quite good enough to be another. I've tried a few different audio samples on different systems, and I'm finding that ideal levels tend to differ a lot.
(In particular I was constantly fiddling with the levels in Star Wars Episode IV every time it switched between music, dialogue and action)
I guess I'm just used to cheap desktop speakers that flatten everything out to an inoffensive beige meh. These just show up every imperfection and imbalance!
I'll see how I go with those other suggestions though (reversing leads, trying a different capacitor for the tweeters, and sticking a sponge in the reflex ports)
If you have anything floating around that even remotely behaves like a measurement microphone, that would definitely help. I have even used a fixed-pattern cardioid LDC with a well-extended and balanced response - takes some extra distance to fend off proximity effect but actually worked quite well.
Also, the 1-capacitor "crossover" is very crude and only ever works properly if the bass driver rolls off where you want in a well-behaved manner with well-controlled breakup modes (chances of that are pretty slim), and both drivers' sensitivities sort of match. That generally means adding some inductance in series (when then provides a much-needed 90° phase shift at crossover, complementing the -90° - 180° of the highpass-filtered and phase-reversed tweeter), and probably a series resistor or resistor divider for the tweeter which tends to be the more sensitive of the two. And then that still isn't great as such a 1st order XO has to be fairly high to avoid overloading the tweeter, often running afoul of possible geometric driver spacing.
You can sort of coax a 1-capacitor XO into working by aiming for a rising treble response and then pulling down the treble electronically to compensate. That's what e.g. the JBL 104s do. The problem is, this generally doesn't give nearly enough suppression of breakup modes unless the bass driver is extremely well-behaved, giving distortion peaks and erratic directivity.
By the time you are making the jump from one to two drivers with its obvious benefits in terms of reduced IM distortion and not requiring an expensive one-size-fits-all driver, you are also confronted with a number of new problems to be solved, no way around that.
Given your listening impressions, I'll assume the original drivers are fairly high-fs jobs to get some half-decent efficient out of what I would guess is not the strongest magnet system and narrowest air gap in the world, which is to say they would play decently loud with the little amplifier provided but not have that much bass, and possibly still be bass deficient despite support from the port. The new driver may even be filling in the low end too much, hence the suggestion of plugging the port. Hopefully it also features some half-decent (linear) excursion, as Xlin requirements quadruple per each octave added.
Also, the 1-capacitor "crossover" is very crude and only ever works properly if the bass driver rolls off where you want in a well-behaved manner with well-controlled breakup modes (chances of that are pretty slim), and both drivers' sensitivities sort of match. That generally means adding some inductance in series (when then provides a much-needed 90° phase shift at crossover, complementing the -90° - 180° of the highpass-filtered and phase-reversed tweeter), and probably a series resistor or resistor divider for the tweeter which tends to be the more sensitive of the two. And then that still isn't great as such a 1st order XO has to be fairly high to avoid overloading the tweeter, often running afoul of possible geometric driver spacing.
You can sort of coax a 1-capacitor XO into working by aiming for a rising treble response and then pulling down the treble electronically to compensate. That's what e.g. the JBL 104s do. The problem is, this generally doesn't give nearly enough suppression of breakup modes unless the bass driver is extremely well-behaved, giving distortion peaks and erratic directivity.
By the time you are making the jump from one to two drivers with its obvious benefits in terms of reduced IM distortion and not requiring an expensive one-size-fits-all driver, you are also confronted with a number of new problems to be solved, no way around that.
Given your listening impressions, I'll assume the original drivers are fairly high-fs jobs to get some half-decent efficient out of what I would guess is not the strongest magnet system and narrowest air gap in the world, which is to say they would play decently loud with the little amplifier provided but not have that much bass, and possibly still be bass deficient despite support from the port. The new driver may even be filling in the low end too much, hence the suggestion of plugging the port. Hopefully it also features some half-decent (linear) excursion, as Xlin requirements quadruple per each octave added.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.