Ideas to improve a lightweight floorstanding speaker? (Polk T50)

[NotInYourHead]

As the title suggests, I own a pair of Polk T50 floorstanders. Being budget speakers, they don't have the best build quality and is pretty lightweight, around 10kg per Polk's website.

While I haven't seen them jump up or down during high SPL, complex music passages, I found that the cabinet vibrates, especially at the top, bottom and the wall directly behind the 6.5" woofer. At certain volumes I could even feel the turbulent flow of air inside the cabinet - the wood Polk used is that thin! Perhaps as a result of this, virtually all objective measurements show a 4-5db peak centered at 1KHz. It doesn't help that the speaker has 2 passive radiators contributing to the midrange leak.

When I checked the insides, there's only a thin layer of white polyester filling behind the drivers, about 3/4 of the cabinet length. Worse, the last 1/4 is virtually a void, no filling whatsover. Polk Audio was kind enough to brace the cabinet, but they're limited to several corner wedges and window (?) bracing around the speaker cutouts. I then stuffed the bottom with some spare polyester filling I had lying around but the issues remain.

What should I do next?

 

[DanielT]

More bracing off the boxes.
Fill them with more filling/damping material. That can dampen resonances but also reduce bass efficiency (if you stuff too hard/much). You have to test and see what suits you.
Maybe paste the Bitumen on the inside walls:


Is there a gasket between the bass drivers, passive radiators and the speaker box?

More extensive, if you have the time, opportunity, desire and think it's worth it build new speaker boxes.

 

[RickS]

Most importantly, knowing the source of the resonances is key. Fortunately Amir measured the T50 so you have a valuable baseline. Bracing can be helpful but can be difficult to target productively without some mode analysis and/or experimentation. Post some pics of the cabinet internals without the damping material and might be able to spot some more obvious improvement areas.

Since this is a budget tower, there are some likely culprits for resonances. One is the front baffle. Notably if areas between the drivers are not well braced. Another is resonances in the woofer and PR. A measurement of the woofer alone would be helpful. The passive radiators and the woofers likely have stamped chassis and they can resonate. This is more difficult to control but could try some strips of butyl rubber on the chassis.

Finally, would caution against going overboard with bracing or damping efforts. Some more expensive damping can be difficult to apply and remove. Also, too much damping can actually worsen the sound. So, using a Umik-1 or other usb microphone can help diagnose the source of resonances and help determine the efficacy of potential remediations. They are fairly inexpensive and can actually save money by avoiding the purchase of more expensive damping materials.

 

[Sokel]

I would also suggest lower levels, speakers are primarily made with that in mind, SPL.
With the effects you describe it's no so much about cabinets I worry about, it's voice coils and drivers in general.

 

[Fledermaus]

Looking at Amir's measurements, 1 kHz peak might have more to do with combination of midwoofer break-up and passive radiator resonance. More bracing and stuffing might not be bad if cleverly applied but in case it was not done already, I'd try first to EQ out this peak, together with lowering treble high-shelf - which would be a more scalable, reversible and non-invasive means of intervention.

 

[a4eaudio]

In addition to bracing, try butyl rubber damping material. Depending on where you are located, Amazon Basics offers a reasonably priced version compared to the name brands - although they don't make the thickness very obvious so not sure that it is apples to apples.

https://a.co/d/cUE3yCh

 

[NotInYourHead]

Looking at Amir's measurements, 1 kHz peak might have more to do with combination of midwoofer break-up and passive radiator resonance. More bracing and stuffing might not be bad if cleverly applied but in case it was not done already, I'd try first to EQ out this peak, together with lowering treble high-shelf - which would be a more scalable, reversible and non-invasive means of intervention.

EQ is the first thing I did when I got these towers and that improved the tonality, but I'd like to try and address the culprit at the source.
Based on Amir's nearfield, I think it's a woofer issue (3rd and 5th harmonics spike) and the midrange leak/resonance from front PRs unfortunately mixing in with the direct sound. These PRs should've been placed in the back to limit audibility IMO. I did put a mic to one PR, playing pink noise, and saw in REW RTA graph lots of unwanted midrange output, around 500-1Khz.

...Post some pics of the cabinet internals without the damping material and might be able to spot some more obvious improvement areas.

I'll send some pics when I can.

Since this is a budget tower, there are some likely culprits for resonances. One is the front baffle. Notably if areas between the drivers are not well braced.

The front baffle is suspiciously thin, now that you mention it. It generally should be thicker than the sidewalls for driver mounting, right?

Another is resonances in the woofer and PR. A measurement of the woofer alone would be helpful. The passive radiators and the woofers likely have stamped chassis and they can resonate. This is more difficult to control but could try some strips of butyl rubber on the chassis.

And yep, the woofer and PRs are made with thin, stamped steel chassis. I'll think about damping those.

 

[DVDdoug]

Since more stuffing didn't help and EQ doesn't help, I'd say just replace the speakers whenever it's economical for you.

If you thought the cabinet was OK, I'd suggest possibly "rebuilding" the speaker with different drivers but we don't know if anything is worth re-using, plus it's a shot in the dark.

 

[NotInYourHead]

Very unscientific, I know, basically consisting of me pulling the woofer from the right tower and doing a nearfield measurement.
I think the woofer is causing the 1KHz peak, but I'm not sure.
Would a woofer replacement do?

 

[RickS]

Very unscientific, I know, basically consisting of me pulling the woofer from the right tower and doing a nearfield measurement.
I think the woofer is causing the 1KHz peak, but I'm not sure.
Would a woofer replacement do?

Your measurement is interesting. If you had a set of baseline measurements against a known driver would help to establish your measurements are reliable. Can you reproduce any part of Amir’s review?

Polk T50 Review (Floor Standing Speaker)

This is a review and detailed measurements of the Polk T50 tower speakers. I bought this recently for $149 from Amazon. As you can imagine, a floor standing speaker at this price has to be quite light and this one definitely is. I can easily lift it as it weighs less than many bookshelf...

www.audiosciencereview.com

He was able to equalize the resonances and this means crossover work would help. Replacing woofers is an entire redesign and does not seem worthwhile for an inexpensive speaker. Equalization is simpler and relatively inexpensive. If you want to learn, a crossover redesign is the next best option. Beyond that, is questionable investment as opposed to spending your money on a new set of speakers.

 

[NotInYourHead]

Your measurement is interesting. If you had a set of baseline measurements against a known driver would help to establish your measurements are reliable. Can you reproduce any part of Amir’s review?

Polk T50 Review (Floor Standing Speaker)

This is a review and detailed measurements of the Polk T50 tower speakers. I bought this recently for $149 from Amazon. As you can imagine, a floor standing speaker at this price has to be quite light and this one definitely is. I can easily lift it as it weighs less than many bookshelf...

www.audiosciencereview.com

He was able to equalize the resonances and this means crossover work would help. Replacing woofers is an entire redesign and does not seem worthwhile for an inexpensive speaker. Equalization is simpler and relatively inexpensive. If you want to learn, a crossover redesign is the next best option. Beyond that, is questionable investment as opposed to spending your money on a new set of speakers.

So far, due to time constraints, I'm limited to near-field measurements of the woofer. It seems to be the biggest problem with this speaker anyway.
The response with the mic really close to the dust cap, about 2cm away, (NF) aligns with Amirm's for the most part. But 16cm from the dust cap, the peak is very apparent.
Someone in the OG thread once suggested cone-surround resonance. This might be proof of that.

I also took the time to take a few pictures of the insides. Originally, there was not a lot of polyester stuffing. The bunch at the top and bottom of the cabinet is my doing, since I read it could help with internal standing waves. There are two window braces between the woofer and 1st PR, and about 1/4 of the end of the cabinet. In addition, there are a few triangle pieces stuck into corners and between PRs' mounting points on the front baffle. All internal surfaces are bare wood, no damping whatsoever.

The front baffle is 18mm thick at most (3/4 inch). Xover is a cheap and cheerful 6-component job. Nothing is interesting about the PRs and woofer by themselves, though there's no metal weight stuck onto the PRs, just a lump of plastic added to the spider center.

Would love to hear your thoughts.

 

[RickS]

So far, due to time constraints, I'm limited to near-field measurements of the woofer. It seems to be the biggest problem with this speaker anyway.
The response with the mic really close to the dust cap, about 2cm away, (NF) aligns with Amirm's for the most part. But 16cm from the dust cap, the peak is very apparent.
Someone in the OG thread once suggested cone-surround resonance. This might be proof of that.

I also took the time to take a few pictures of the insides. Originally, there was not a lot of polyester stuffing. The bunch at the top and bottom of the cabinet is my doing, since I read it could help with internal standing waves. There are two window braces between the woofer and 1st PR, and about 1/4 of the end of the cabinet. In addition, there are a few triangle pieces stuck into corners and between PRs' mounting points on the front baffle. All internal surfaces are bare wood, no damping whatsoever.

The front baffle is 18mm thick at most (3/4 inch). Xover is a cheap and cheerful 6-component job. Nothing is interesting about the PRs and woofer by themselves, though there's no metal weight stuck onto the PRs, just a lump of plastic added to the spider center.

Would love to hear your thoughts.

Just a few comments for you. Be careful not to add too much damping. Should only be lining the interior panels. If want to experiment add more at the bottom of the cabinet. Next, the lack of metal mass to PRs is not likely an issue as they were custom designed by Polk for this speaker. Finally, it sounds like Polk made a decent attempt at bracing and the front baffle thickness is fine considering the price point.

As for measurements, hobbyist distortion is only useful as a relative measure and so not all that useful. Would take before and after measurements after each mod to determine whether and/or how much change they caused. This is the optimal way to judge which mods result in useful benefit. Being able to measure impedance is very helpful but still unclear on what your end goal is. Money spent on measurement rig might be better spent on better speakers.

 

[Head_Unit]

only a thin layer of white polyester filling

That's probably not doing much. And the void is a bit odd-in a ported design you can't get stuffing near the port end because it interferes with airflow. With passive radiators that should not matter. The big question is where does the 1 kHz resonance come from? It's not really obvious from the measurements https://www.audiosciencereview.com/...polk-t50-review-floor-standing-speaker.24890/ though it seems the woofer is in break-up most likely. And maybe that radiates back out through the passive radiators, perhaps exacerbated by a resonance front-to-back in the cabinet.

I'd want to really brace this thing at a couple points front-to-back and side-to-side, like 1" dowels which then touch in the middle for more stiffness. Then stuff this thing to reduce internal resonance-1 lb of fiberglass per cubic foot is my advice at least until I find my copy of Vance Dickason's seminal Loudspeaker Cookbook which has a chapter about that. And I'd coat the interior with some antivibration goop. I did all this to my best friend's husband's old speakers and it was shockingly transformative. One note: stuffing can make the box act "bigger" than it physically is, which would shift the radiator tuning down hence @RickS admonishment about too much stuffing. Personally I prefer "undertuned" designs.

 

[terryforsythe]

1 lb of fiberglass per cubic foot is my advice

Fiberglass insullation is very effective. I used it in one of my projects after first removing it from the backing. A doctor in this forum commented on it and suggested not using it due to the potential of air-born fibers. I think that mostly is a concern in a ported design, not a design using a passive radiator, but I'm not a doctor. Another good option is to use something like Sonic Barrier Acousta-Blue damping material. That also works very well, and without any potential health concerns of which I am aware.

I'd want to really brace this thing at a couple points front-to-back and side-to-side, like 1" dowels which then touch in the middle for more stiffness.

Polk specifies the dimensions to be 36.25" H x 7.75" W x 8.75" D. I don't know what is the thickness of the cabinet material, but if it is around 5/8" (not uncommon in mass produced speakers), that puts the internal dimensions around 35.0" H x 6.5" W x 7.5" D.

The peak in frequency response appears to be around 920Hz, which has a half-wavelength of 7.36" inches. So, assuming the peak is not caused by the woofer itself, the peak could be due to a standing wave between the front and back walls.

@NotInYourHead I would run a frequency response sweep of the woofer in free air (i.e., pulled out of the cabinet) and with the microphone very close to the woofer to see if the 920Hz peak still is there. (You will need to figure out a way to hold the woofer - you probably don't want it laying flat on a hard surface.) If the peak still is there, I don't think additional cabinet bracing will do much to address the peak.

If the peak is not from the woofer itself when measured in free air, then additional cabinet bracing may help. Since the half-wavelength is close to the cabinet depth, I would add bracing side-to-side to reduce the total area front-to-back that is parallel at 7.25". For that, you may want to use mdf or particle board since wooden dowels won't have a significant impact on changing that total area that is parallel. Additonal bracing front-to-back to reduce baffle and rear wall vibrations also may not hurt.

Additional bracing will reduce the internal air volume, but GOOD damping material (not the polyfill that is in there now) should make the cabinet appear 10-20% larger from an acoustics perspective; it does so by slowing the speed of sound inside the cabinet.

 

[RickS]

As mentioned previously, am strong proponent of finding problems before applying solutions. We have already discussed the woofer measurements, but if the panels are resonating badly, you need to find the source. This can be aided by placing clamps or ratcheting straps in likely spots and doing before and after measurements. Guessing can just get you bracing that feels right but is ineffective or maybe worse. Moving a strong low resonance to a higher frequency may just make it more audible.

Take note of Amir’s eq. He went after the lower frequency of the resonances. This is a good approach notably if you find additional resonances at some multiple of a base one.

Unless you are good with your Polks becoming a science experiment, plan your mods knowing they may need to be rolled back. Expectation bias can be strong with DIYers and, unless you have the patience and skill to make modifications prudently, you may just be better off spending your money on new speakers.

 

[Ktacos]

I would just grab something like sound barriers denim insulation sheets and line the box with it, possibly fill up the the lower portion with quite a bit of material to kill the resonance related to the speakers height. I wouldn't go through the trouble of adding more bracing.

 

[Apesbrain]

You might consider decoupling the speakers from the floor using an outrigger/spike system like this. Also, add mass to make the speaker a more stable platform for launching sound waves into the room. A 10 lb. weight placed on top of the cabinet should help.

 

[Roland68]

As the title suggests, I own a pair of Polk T50 floorstanders. Being budget speakers, they don't have the best build quality and is pretty lightweight, around 10kg per Polk's website.

While I haven't seen them jump up or down during high SPL, complex music passages, I found that the cabinet vibrates, especially at the top, bottom and the wall directly behind the 6.5" woofer. At certain volumes I could even feel the turbulent flow of air inside the cabinet - the wood Polk used is that thin! Perhaps as a result of this, virtually all objective measurements show a 4-5db peak centered at 1KHz. It doesn't help that the speaker has 2 passive radiators contributing to the midrange leak.

When I checked the insides, there's only a thin layer of white polyester filling behind the drivers, about 3/4 of the cabinet length. Worse, the last 1/4 is virtually a void, no filling whatsover. Polk Audio was kind enough to brace the cabinet, but they're limited to several corner wedges and window (?) bracing around the speaker cutouts. I then stuffed the bottom with some spare polyester filling I had lying around but the issues remain.

What should I do next?

Do you really think this effort is worthwhile?
Wouldn't it be better to invest the effort, time, and money in a different speaker or kit that doesn't have so many flaws/limitations in the cabinet?

Otherwise, I'd advise against patching it up and instead recommend doing it properly from the start. In any case, I wouldn't reduce the internal volume.
You could remove the foil and reinforce all six sides with 10, 12, or 16mm MDF.
There are two or three approaches for the front, but that depends on your resources.

 

[Ktacos]

Do you really think this effort is worthwhile?
Wouldn't it be better to invest the effort, time, and money in a different speaker or kit that doesn't have so many flaws/limitations in the cabinet?

Depends on the person, for me cutting some braces and adding new absorption material would take a few hours at most and be worth my time, not an expensive set of mods for me to do either. I mean It's diy audio, if one is even asking then they probably determined that it was worth their time. I would put this pretty low on the scale of time/effort investment in terms of speaker mods.

Though for the 1khz problem they complained about they should just use EQ for that.