musical sub using old speaker components

[mikecon54]

I'm new to DIY and in experimentation mode. I'm curious what kind of results I can get from building a music focused sub using drivers from an old pair of Paradigm Monitor 7 (v1) that I've retired. As an initial step I did some playing around with chatgtp. I'm interested in seeing what folks on this forum think of the results and if the recommended build makes any sense as a way to reuse the 6.5 inch drivers that otherwise aren't worth much given the age.

Repurposing speaker components

A conversational AI system that listens, learns, and challenges

chatgpt.com

Not sure how to tell if what its giving are bad assumptions or hallucinations. If the science makes sense though it seems like it would result in a decent little music focused sub for someone like me that primarily listens to rock, blues, jazz and folk with a little bit of pop for the kids (who don't care about accuracy).

 

[Salt]

Uhm ..
Don't see any driver specifications, but enclosure recommendation?
No Hypex plateamp recommended? ..
And how should the plateamp be isolated?
... AI needs to read some more books ... ehem, sites...

 

[mikecon54]

@Salt , do you know of somewhere I can find detailed driver specs for something like a gen1 paradigm v1? So far as I know, they've never publicized that type of info.

My other option I'm investigating is building a sealed box for an old SVS PC12-NSD driver with a dead amp. I'd replace the amp with a new one but the wife won't let me bring back the baby water heater of a cylinder sub.

 

[mikecon54]

also, as I mentioned I'm a newbie to the idea of building my own speakers but am techie enough to know not to trust what the AI gave me without verification. That's why I came here before I did anything. if either of them things I'm thinking about are dead ends then I can just see what I can get on FB Marketplace. I just hate to get pennies for gear I loved and enjoyed for many years.

 

[DVDdoug]

The drivers seem kinda' small for a subwoofer...

Normally you use the Thiele-Small parameters with speaker design software to simulate/optimize a sealed or ported cabinet. WinISD is free speaker design software or there are other options. I trust the software a LOT more than AI, although AI will have access to the same theory and formulas.

Since you probably can't find the Thiele/Small parameters for a driver salvaged from a speaker, there is a gizmo that can measure them, but it's a bit more than $100 so maybe not worth it. It might make more sense to buy a new woofer with known specs/parameters (maybe a bigger one).

Another option if you still have the cabinets is to copy the box volume and port dimensions from the original speakers.

Or, sealed speakers are more forgiving than ported speakers so you could just build a random sealed box. (To a point, larger is usually better and if you have the parameters the software will help with that.) Sealed boxes tend to roll-off starting at higher frequencies but the slope is less steep than a ported speaker. So with EQ and enough amplifier, the bass can be boosted and often extend lower than the ported box, and you can get-away with a much smaller box. A lot of active/powered subs are made that way with the EQ built-in

 

[mikecon54]

Thanks @DVDdoug, the gizmo is interesting and a possibility. Especially since I have both the drivers from the monitor 7's as well as the old SVS driver.

I would guess if I can get measurements on it I'd have better luck with the SVS 12" driver from the NSD-12 in a properly tuned sealed box than any combination of the drivers from the paradigm monitors. Who knows, maybe I can build both and use one to fill in my main system and the other for my family room where all we have are in-walls with no sub right now. If it doesn't make sense in either case though I'm definitely open to building using an off the shelf driver.

Being new to all of this I'm trying to learn as much as I can right now. Got started down the rabbit hole calibrating my home theater using OCA's scripts and now I'm a bit hooked on trying to maximize every room

 

[mikecon54]

I'm also open to trying out the manual REW method for measurement, but could use some pointers from anyone who may have done it before. I'm definitely someone who wants to gather as much info as possible before starting a project. I realize in either case it may not wind up with ideal full spectrum bass response, but I'm interested in this as a science project just as much as I am as a way to potentially improve one or more of my systems.

Impedance Measurement

 

[Salt]

I'm also open to trying out the manual REW method for measurement, but could use some pointers from anyone who may have done it before. I'm definitely someone who wants to gather as much info as possible before starting a project. I realize in either case it may not wind up with ideal full spectrum bass response, but I'm interested in this as a science project just as much as I am as a way to potentially improve one or more of my systems.

Impedance Measurement

For home theater request experience from @dualazmak or @Keith_W ...
They wait for that

 

[Devnull]

You can use REW with a homemade test jig and get the T/S parameters using the added mass method.

 

[mikecon54]

You can use REW with a homemade test jig and get the T/S parameters using the added mass method.

Sounds like I need to dig into some youtube videos on how to set that up

 

[kemmler3D]

I'd have better luck with the SVS 12" driver from the NSD-12 in a properly tuned sealed box than any combination of the drivers from the paradigm monitors

This seems sensible to me, trying to build a sub from 4x 5.5" drivers, less so.

Bear in mind that regardless of T/S parameters, the maximum SPL of a speaker is limited by Xmax and cone area. One 12" is bigger than 4x 5.5" drivers put together, and almost definitely has a higher Xmax too, so the game is over before it really starts with the paradigm drivers.

You can play with the numbers here: https://www.baudline.com/erik/bass/xmaxer.html

 

[mikecon54]

interesting tool, hadn't seen that before. so for my main system that I play loud, if I want to complement my pb2000 with something more punchy when I'm listening to music I'd likely want to use the 12" to complement the rumbly deep bass of the pb2000.

For the family room system that almost never plays loud (or else we'd strangle my 14 year old playing rocket league), the 4x6.5 may be more realistic? In this case I'd just like to just have a bit of fill on the low end. I'm not looking to shake the house given that more often than not its used for my son's gaming.

 

[kemmler3D]

interesting tool, hadn't seen that before. so for my main system that I play loud, if I want to complement my pb2000 with something more punchy when I'm listening to music I'd likely want to use the 12" to complement the rumbly deep bass of the pb2000.

For the family room system that almost never plays loud (or else we'd strangle my 14 year old playing rocket league), the 4x6.5 may be more realistic? In this case I'd just like to just have a bit of fill on the low end. I'm not looking to shake the house given that more often than not its used for my son's gaming.

Yes, using 4x 6.5" drivers is somewhat realistic for a sub, but that's assuming they can do a good 4mm xmax and handle ~100w each without running into problems... possible but not certain? Also I think the Paradigm site says this speaker had 5.5" drivers which costs you 3dB relative to 6.5" in this situation.

You can hope for about the same amount of bass the Monitor 7s produced originally but probably not much more, and especially not if you put them in a box smaller than the speakers would have added up to in the first place. This might be worthwhile for the family room, your call. If the original speakers needed a sub, putting the drivers into a new box won't change that need fundamentally.

One toyota prius can go about 100mph at most. Bolt two together and you might be able to get a bit faster, but 200mph is not on the menu.

 

[mikecon54]

I like the analogy. My particular monitor 7s are first gen from the late 90s. Back then they were still using 6.5 inch drivers. I think it was around the 6th or 7th gen that the switched to 5.5.

If nothing else I would assume it would get deeper than the circa 2006 Polk in-walls I installed when we bought our house. I could be wrong of course.

Once I find the time to try the builds out I’ll have to report back measurements. Not trying to maximize here just trying to find a creative way to reuse parts from speakers I enjoyed for many years. If it doesn’t work out it won’t have cost me too much.

 

[Heinrich]

... some playing around with chatgtp. I'm interested in seeing what folks on this forum think of the results ...

As always, Large Language Models (LLM) are not designed for solving tech/ problems. They are designed for the only purpose to simulate (!) human language. They regularly succeed, as people think LLMs would think. When designing a chimney a friend tried, and the answers were off by factors of magnitude (10..100).

The mere fact that it answers proves my concern justified. Look, it wants you to put the drivers on opposing sides, and then brace that sides to each other - how would that work? And it predicts best results 'musical, tighty tight' w/o knowing the drivers' parameters.

LLMs do not substitute human ingenuity, they are parrots. If you had one, would you ask it?

add.: gave the above to µSoft's CoPilot. It started an answer but then stopped as to "keep the communication civilized". It said at a first glimpse: do not trust me at all, I'm ... XYZ. That's were the disrespect may originate in, which made it recall its answer. It's pretty much clear, right?

 

[mikecon54]

I'm definitely aware of the limitations of LLMs as a software developer myself. My version of not trusting it was coming here to check with folks who are more expert than me on the audio science whether what it came up with made any sense. I found some youtube videos on how to gather T/S either with a homemade jig or the Dayton. I'm definitely reconsidering using the 6.5's in a sub and will probably do that with the SVS 12" driver instead. Might switch to reusing the drivers from the paradigms to make a pair of bookshelf speakers for a buddy. That way I can reuse the tweeter and internal crossovers. Will have to download WinISD once I find some time to measure T/S on the drivers.

Very grateful to all for the tips and links.

 

[Heinrich]

I'm definitely aware of the limitations of LLMs as a software developer myself. My version of not trusting it was coming here to check with folks who are more expert than me on the audio science whether what it came up with made any sense. I found some youtube videos on how to gather T/S either with a homemade jig or the Dayton. I'm definitely reconsidering using the 6.5's in a sub and will probably do that with the SVS 12" driver instead. Might switch to reusing the drivers from the paradigms to make a pair of bookshelf speakers for a buddy. That way I can reuse the tweeter and internal crossovers. Will have to download WinISD once I find some time to measure T/S on the drivers.

Very grateful to all for the tips and links.

Thing is, you do not need T/S parameters no more. That's not because of LLM, but because of digital equalization together with room integration, while having powerful amps for cheap.
With four 5" you'll have 4x80sqcm of cone area = 320sqcm eq/ to a smaller 10". Roughly 40liters of internal volume would fit. The rest is measurement in-room, and proper adjustment of filters. To make the enclosure stiff I leave to your creativity. Not the least, a good starter into DIY, don't long for the ultimate too soon.