Holy irons, batman! Don't know if I could afford all of that inductance. If there's a schematic available I'd be very interested in having a look.
Why ditch the parafeed, though? Seems a good counter to the PSRR issues that horns can cause.
Regardless, very neat looking builds!
Why ditch the parafeed, though? Seems a good counter to the PSRR issues that horns can cause.
Regardless, very neat looking builds!
Driver excursion is maxed at ~1.5W, so I don't need all that much beef. Though I suspect a decent power margin couldn't hurt.
Just a small remark about EQ, dips vs boosts. In many situations it amounts to the same thing. If you pull down midrange, you will turn volume up. If you boost bass, you will not turn up the volume. Either way, you will feed in more power to get that improved balance in the end. What does not work, is boosting cancellation dips. You would just be pumping in power with little return.
The practical truth is somewhere in the middle, of course.
To be honest I have most experience with digital EQ and have managed to replicate every mistake in the book. If you are simply aware of how much boost or cut you are applying, and how that interacts with your driver's power handling and the available power of your amp etc, it's not that hard to do a decent job. Very good is hard, but isn't that with all things we only pursue as a hobby?
The practical truth is somewhere in the middle, of course.
To be honest I have most experience with digital EQ and have managed to replicate every mistake in the book. If you are simply aware of how much boost or cut you are applying, and how that interacts with your driver's power handling and the available power of your amp etc, it's not that hard to do a decent job. Very good is hard, but isn't that with all things we only pursue as a hobby?
In the upstart 2 of the pieces of iron are two of the same (German) OPTs, one used as a parafeed choke. Parafeed swaps a cap for a non-gapped OPT. I don’t feel it is necessary.
There is only 1 filter choke, 100 mH Hammond, we bought a box of them they were something like $15 CAD each. The flter choke is very important. The caps are pricier. In a SET power supply is more important than the rest.
Map:
Looks like there might be a couple elcos.
dave
Keep in mind the music is dynamic and that specified driver excursion is a pretty wobbly spec with no real definition. And there is xmax and xpeak.
So this number is pretty squishy with a high set of error bars.
That does not hurt.
Keep the circuit simple, and add a switch to go from triode to Schade feedbaxck pentode. No guessing. And you might find each mode is better (for you) on some source material, your mood.
dave
Yup -- better yet I've just guestimated the excursion to be ~1.5mm based on similar drivers. So there's plenty space either way.
On the EF86? EL84 seems to be fairly locked into pentode operation from the schematic. I would also probably substitute the EZ81 for a 5U4GB, simply because I have two of those since a few hours ago.
You mean the 10H/100mA? Those go for about three times as much 'round these parts, but I could probably get my hands on something similar for a reasonable amount... somehow. I'll have to see what I can scrounge up!
Tends to be the case, yes. However -- I'm treating any EQ dips as "lost SPL". The speakers are going to be plenty loud for my use case, so I can afford to lose a bit of maximum loudness in favor of a flatter response. Even at a peak input power of ~1.5W (Down at 40Hz, the remaining frequencies north of that use ~0.8W) the speakers will be putting out a theoretical ~105dB. Granted this is all in theory and should be taken with a healthy heaping of salt, but I trust it's not ridiculously off the mark, at least.
And yes, very good is always the aim, but yet I routinely find myself in "good enough"-territory. 🤔 Don't mind a bit of compromise as long as I get to pretend the decision was mine
Last edited:
We did 7 iterations, with an 8th i’d try a beefier triode (6N6P), and lose the parafeed.
The switch would connect the screen differently and switch off the Scade feedback (plate-to-plate)
EF86 in tridode was a suggestion by Eddie Vaughn, it is kinda whimpy and i think it could get better yet.
dave
Ok, gotcha on the switch situation.
What's the reasoning behind using a pentode on the input? I'm only versed enough to know that triodes are typically preferred there due to generally lower noise figures and snappier response.
Hey folks!
Since last posting I've gone and graduated and secured a position as a hardware test engineer. I've thusly also been able to rack up the necessary funds to buy the wood for this speaker build.
All things going well, the 1" birch ply should arrive some time next week.
I've tweaked the dimensions of the speaker to retain the exact same internal volume as the original with 0.75" ply, and once I've triple checked it all I'll be posting the CAD files for posterity.
Naturally, progress pics will also be shared. Stay tuned!
Since last posting I've gone and graduated and secured a position as a hardware test engineer. I've thusly also been able to rack up the necessary funds to buy the wood for this speaker build.
All things going well, the 1" birch ply should arrive some time next week.
I've tweaked the dimensions of the speaker to retain the exact same internal volume as the original with 0.75" ply, and once I've triple checked it all I'll be posting the CAD files for posterity.
Naturally, progress pics will also be shared. Stay tuned!
Took delivery of the ply last Thursday, things were HEAVY.
Also entered the dimensions into Opticutter -- stock usage is sitting pretty at 95%😬. No mistakes allowed.
Next step is to properly square the gantries on the laser cutter at work and do some test cuts to confirm dimensional accuracy. Once that's done, I'm going to cut some jigs out of 6mm MDF to ensure that critical angles are accurate.
Not sure how I'm going to go about joining the pieces, though. Wood glue is a given, but screws, dowels, box joints... dunno. I'm open to suggestions!
Also entered the dimensions into Opticutter -- stock usage is sitting pretty at 95%😬. No mistakes allowed.
Next step is to properly square the gantries on the laser cutter at work and do some test cuts to confirm dimensional accuracy. Once that's done, I'm going to cut some jigs out of 6mm MDF to ensure that critical angles are accurate.
Not sure how I'm going to go about joining the pieces, though. Wood glue is a given, but screws, dowels, box joints... dunno. I'm open to suggestions!
Attachments
Excellent resource. I can see those types of corner pieces working in some parts -- Primarily the longer runs along the rear and bottom of the speaker. But when it comes to the more intricate horn sections I'm thinking it won't be feasible.
Will definitely add some chamfered pieces with screws through them along the interior of the horn mouth.
I'm just now realising, though, that I won't be able to do box- or dovetail joints on the sides since I'd run out of material having to make the pieces wider. So butt- or some manner of 45 degree joints it is.
I'd rather not have to run screws from the outside of the speakers but since birch isn't the most visually appealing anyways I figure I should do away with that imperative -- Veneer or some sort of paint is probably the better option.
Yup, that's what I'm thinking. Not sure how I'll go about the internal bracing, but I know for sure they're going to be made from some at least marginally cheaper stock.
Going to brace the horn mouth with a section of threaded rod to tension it inwards, figure that'll stiffen it significantly. Probably going to "sleeve" the rod in something more aesthetically pleasing when all is said and done.
Great progress has been made on the Dallas Horns! Managed to snag one of my dream motorcycles a few months back so this project was put on the back burner, but these past two weeks I've been chipping away on the build after work.
I'd say a solid 80% of the panels are cut out now, and I intend to have the rest cut out before the end of the year. I've also modeled and 3d printed corner braces to ensure consistency when lining up the inner panels.
Can't wait!
I'd say a solid 80% of the panels are cut out now, and I intend to have the rest cut out before the end of the year. I've also modeled and 3d printed corner braces to ensure consistency when lining up the inner panels.
Can't wait!
Attachments
GM, that'd be a Honda CX500 Turbo! Rare gem, 'specially here in Sweden. If all goes well I'll have it on the road next summer, was a bit of a fixer-upper.
Ah! Remember them; got 2 like new/low mileage for $6&10k in Fla, though nothing to indicate why the wide $ gap.
Prices can vary wildly. I only gave ~1.2k for mine, though the engine was shot and it had been dropped more than once. Only way I could ever afford one.
Through sheer circumstance I somehow managed to secure a display engine with 0(!)miles on it, which I'll be picking up come Jan/Feb if all goes well. Hope to have the speakers done before then!
Through sheer circumstance I somehow managed to secure a display engine with 0(!)miles on it, which I'll be picking up come Jan/Feb if all goes well. Hope to have the speakers done before then!