First Build - Rework of Cheap $10 TV Boxes

First build.

I am a stereo 2.0 guy, not a HT 5.1 or 7.1, so most of my interest is in building 2.0 speakers and maybe one or two subs to go with them. I am not a woodworker and my cabinetry skills are below zero, so fit and finish will be the hard part for me. I am far away from any stereo store so building my own is a necessity. Parts are somewhat available via mail order but much of what one can get in the EU and USA is not available here and will not pass customs so I must make do with what I can get via online sellers.

The box:
A friend moved away last year and left me with a pair of faux-wood paneling, vinyl covered speakers with plastic removable grills and an led light show in the base of each box in a small sub enclosure. This is a 2-way bookshelf speaker, 22cm tall X 12cm wide X 11cm deep. Existing wire is 24 AWG running to a hole in the back of each box, no posts. The tweeter is a paper cone 1.5" generic yellow cone and the mid is black, also generic and 3".

First task was to listen. Muddy, muddled, no soundstage, no room reverb, no echo, details were swallowed up, the tweeter had clearly seen many hours with an overdriven SS output stage and everything above 8k was gone. Recessed and lifeless, sounded like a pair of mid-fi 1970's boxes playing an 8-track, but without the midrange life that even an 8-track had.

Ok. That's the start.

I chose cheap replacement drivers, biwired to an outboard xover that I ordered and did not build and design. If I make any mistakes I do not want to do it with a pair of Fostex 120's. Next was to get tools and gear. After a month of receiving deliveries from various online sellers I had a soldering station, tools, wadding, voltmeter, new drivers, 16AWG, sound proofing, a 50W pa mono PA test SS amp, and the rest. Total cost for all the tools was more than 3X the build. Drivers were $10usd for all 4 combined in total and as you will see this choice was wise for this first project.

Xovers were 2nd order $20 Kasun for the pair and chosen for their overkill 150w capacity and the 3.2khz xover point and hefty mass to self damp vibrations when installed to their outboard boxes.

Removed the pop off grill and to my surprise below each mid is a small box with a useless led light show wired off the mids. Clip and toss but there is now an empty space on the front panel. Now to removing the drivers, four screws up front on the mid, two on the tweeter. The xover was a one cap 47uf pass filter across the mid terminals. Clip and into the parts bin and clip the 24AWG and leave 2" of lead to use these drivers as testers. Yank the remaining 24AWG out the back.

Remove the drivers and blow the dust out with compressed air. Now to apply some brush on liquid adhesive to solidify the box and soundproof the inside. I managed to cover most of the inside but the tight space and inability to remove the front panel meant that coverage was not total. I let this dry overnight, next day drill out a larger hole for the 2 X 16 AWG of biwire.

Lesson One: Drill first, then blow out the dust and apply adhesive.

Few days later did the next box, drilling first, then applying.

Both boxes now adhesive on the inside, I apply some auto firewall soundproof material, cutting it into strips and laying it on to the interior surfaces of the box. Let that sit a day and then cut two lengths of wire 15" long.

Lesson Two: Whilst these are intended to connect to the xover next to the speaker, this lead was too short to properly separate the xovers. 24" for hard wired outboard crossover is better for this bookshelf and any floor standing 6+ feet would be optimal as one wants to get the xover as far from the speaker as is reasonable.

Pull the 2 lengths of 16AWG for the two drivers through and apply kids modeling clay inside and outside to seal. Next the wadding and the rubber face. To reduce reflections off the front I got several cheap mousepads, black, no logos, and glued them to the front of each speaker, face up (the part you would place your mouse on facing out) trimmed the round driver holes leaving enough to cover the 1cm thick front plate and act as a vise to hold the drivers snugly as the hole is 4mm larger than drivers.

Cutting the holes and trimming this was not optimal - fit and finish being one of my weak points.

I tried several glues to get the driver holes trimming to adhere the the sides of the hole to make the fit snug, but found that several did not keep the mousepad material in place well.

The small space in the base, where the led light show was, I chose to fill with kids putty and a few rocks to add some mass and damping. Pack it in tight and refill as the putty dries out.

Lesson Three:: Next time, better modeling clay or Blutack that won't dryout.

I check the drivers for polarity and as this is Asia nothing is to be assumed. I pop tested one of the mids and wired it in, copper lead to neg tweeter the same, then hooked the wires to one crossover and connected speaker wires to it. Tested polarity of the system with a 9v at the xover input.

Lesson Four:
Never do this. Pop tests definitely cook off the high end of tweeters but they also destroyed some of the 'life' in this cheap $3 midrange. Never pop test with a battery. Bad idea. I thus need to find a safer way to test for polarity cohesion from driver posts through xover to speaker posts on the amp.

Now you know why I used cheap drivers.

Lesson Five:
My soldering skills are terrible. Leads look awful even worse than when I worked in High School. Practice required. Future xovers will have better lead solder points as the Kasun's had small screw pits, not holes. This made soldering a challenge.

So the R speaker is up and running and works and seems to be polarity coherent from driver terminals to and through the xover and from the speaker terminals to the amp.

Repeat for the L speaker, learning from prior errors.

Trimming the front mousepad finish is not to my liking as I cannot get it 'perfect' and looking as I wish, but they are finished.

When paired up the R is definitely duller and will need driver replacement due to the battery test.

I am uncertain about drive polarity cohesion and want to pull all the drivers to check my connections.

That said they sound much better than the originals, with drive and a toe tapping quality with room ambience readily heard. They do not play loud and when pushed the 50w amp begins to show signs of strain. I can get the mids to max incursion on bass heavy tracks but do not want to push them too far. FR estimated at 90-20k.

I have been listening to them for a few weeks and now will get some better drivers, the tweeter size is limited as any large front plate tweeter won't fit so nothing above $5 is available to me AFAIK and choices are very limited. Mids I am going with a pair of Fountek at $15 each, so re-driver of this box from these first prelims will increase the project price by a factor of 4. Once I get those I will box up the xover and move the four units into the lr from my workbench and use them as the main system whilst I work on the next project.

That is going to be a 5" to 8" full range one driver no xover build. Directly solder no posts. The cabinet will be the crucial part of that one. Cabinet and Q and Ohm matching with the amp, designing them as one dedicated system, will be the crucial elements.

Many lessons learned one this first project regarding basic build techniques, order of procedures, and more ideas for future builds.
I checked the polarity of the speakers and the R hand unit had the mid wired out of phase. I corrected this and then moved the speakers from the workshop into the living room powered by the mono PA amp and the R unit is definitely cooked due to the ill-advised phase/polarity pop test. Hopefully this is only the drivers and not the xover, but if it is replacement is easy as the xovers are outboard of the speakers, sitting 12" behind each.

I am going to order up a set up better drivers on these, budget of $90 for all four, build some speaker stands (that can be adjusted to accommodate future larger boxes) to elevate them, get some improved speaker wire upgraded from the 26AWG that came with the original boxes, get another one of these tester amps and run this as a mono block setup off of a USB source fed to one of my reclockers.

These speakers sound good, well the L hand one does, on acoustic, small ensemble classical, female vocals and such. Nothing below 120hz, no drums or bass or rock or electronic music or so even when placed well into the corners and along the back wall. They are, after all, tiny bookshelf units intended for nearfield listening placed on the shelf in front of you and against the wall. No sound-stage as this is two speaker mono and the R-hand unit is faulty.

Local hardware and electronics stores have thick AWG copper strand and solid core but so far on multiple samples both purchased and scavenged from existing gear, the quality of this wire is low. Brittle, strands easily snap off and break into tiny bits of metallic dust (and I am mindful of what issues that could cause), etc. I am looking into solutions for connecting thick gauge AWG to the tiny terminal posts of drivers and xovers. Most terminals are tiny BUT the wire in the driver itself is much larger AWG and I want to get a fat pipe to the drivers. I could use 2" of 26AWG to lead to the terminals but that would cause resistance and the whole point is to lower that with the fat pipe.

I need to find a USB source player that outputs coax, fibre, spdif etc to put on top of this rig to play music in stereo and accommodate multiple ODAC's in the chain. The USB of the amp only plays the left channel of any USB stick - no blending of channels (tested on Strawberry Fields Forever). This will be either something I find online as a digital media source centre, would have to be powered to run the bus, or maybe a used laptop dedicated to the system playback library, moving files from my main PC as needed.

My soldering skills are still terrible.
Last edited:
Went out this morning and got some 10AWG solid core copper and replaced most of the 24AWG with it, leaving 24 for leads to the xovers. Made a difference, more dynamics, plays louder, mid-range clearer, the R-speaker is still off even when I angle it in and leave the L-speaker facing ahead so new drivers are a must.
S.0 >> S.1 Update
Bought some new drivers for these boxes and will replace the current drivers with some in the $10-$25 range for each driver. Frequency Response is a little mismatched in level so a resistor for each tweeter may be need to pad them down to the midrange SPL level. Then this will be finalised and then time to move on to S.2, the open baffle.

I am going to need to make some speaker stands for the listening/living room.
Need to get some tools to file the openings as the new drivers are a mm too large.

The bigger magnets are welcome but no room to wire the terminals into this existing box.

May need to run wire extensions, violating my simplicity rules for minimal connections.

Boxes for the outboard xo's, and this one will be finished this week.

Still listening to most music with one X $1 X 3 inch OB driver in a piece of cardboard.

Learned a lot on this first builid.
Photos. No measurements on any of these. S.1 drivers in the S.0 boxes. This system has good extension down to ~85hz and with new drivers may get lower. S.0 drivers in the OB with a 10uF cap high pass filter of the 6.1 ohm tweeter (actually a broadband driver in the upper range, goes down to 1400 or so) dc measured not dynamic. Great open sound, and fast on transients - quickness of OB was a surprise.


Paid Member
2003-08-03 11:43 am
I check the drivers for polarity and as this is Asia nothing is to be assumed. I pop tested one of the mids and wired it in, copper lead to neg tweeter the same, then hooked the wires to one crossover and connected speaker wires to it. Tested polarity of the system with a 9v at the xover input.

Lesson Four:
Never do this. Pop tests definitely cook off the high end of tweeters but they also destroyed some of the 'life' in this cheap $3 midrange. Never pop test with a battery. Bad idea. I thus need to find a safer way to test for polarity cohesion from driver posts through xover to speaker posts on the amp.

Now you know why I used cheap drivers.

I've always used a 1.5V battery and not had any issues with woofers or mids. I've not tried it on a tweeter though.

I test tweeter polarity with an actual measurement. Even a cheap webcam mic should be good enough to show up a gross problem such as a tweeter being connected with the incorrect polarity (you will get a big dip in the frequency response if it is the wrong way around.

Software such as holm impulse is excellent for this :)

Note that depending on the crossover order and drivers the tweeter may be better wired in reverse phase. a simple polarity test won't show you this.

Thanks for the tips!
No xo just a cap so no phase issues, I think.

These current builds do not have a sibilance issue, likely due to luck more than anything else.

I used a 10uF cap as the "xo" on the OB above, about 2610 hz with a measured DC impedance of 6.1 ohms to the tweeter. The mid/woofer has extension well up to 7k and the tweeter goes down to 1400 or so. Maybe add a simple xo to the woofer to cut out its shouty upper range.

Maybe will move those to my desktop, put a big baffle woofer under the desk. I have 60cm square to work with - surely large enough, but near-field these sound fine to my ears good down to 125 or so. They play loud enough on a 50w 110v constant output PA amp.
I finally sealed up these boxes with the best drivers that I am willing to put into them, $10 to $30 per driver, and they play so loudly that they vibrate about on top of the speaker stands.

I dampened the drivers and inside the cabinet with the usual materials, put rocks and clay in the base, glue on interior walls, soundproofing material and cotton wadding, sealed all the holes, crossovers are 9 inches outside the boxes, but the limitations of the cheap 1.5cm thick plywood have them vibrating all over the place.

They sound good for a small box, no open OB sound for sure, but the resonance of the cabinet is at several different frequencies when played loudly.

100, 90, 80 the box and xo buzz like bees.

LF extension is much better than it has right to, flat to 55 and then a gentle roll off with audible down to 35hz.

I could measure this, SQ, FR, resonances, etc. But the limitations of the box, the prefab xo, leave little room for improvement.

I will take lessons learned and use them in future builds.
Last edited:
Thinking about what to do for any future boxed builds. Going to get an OB design but no matter a box is going to lack that open sound with a trade off of a bass rolloff. Going to build some very sturdy speaker stands to hold these up.

Someone mentioned about sub-boxes with a layer of sand all around the interior enclosure to dampen resonances. Those granite 2-ways seemed good as well. I could probably get local masons to do faux-stone in a mould if I could fabricate the form.

Trying to hunt down old 12" and larger drivers for an OB H-frame and most things here are quickly scrapped and recycled. They don't stay around for long. If I get this a simple adjustable low pass filter will be in the loop.
FWIW, a small, worn out 1.5 V [my current one measures ~0.98 V] works fine on even the smallest, cheapest [super] tweeters I've tested. If it damages a driver, it's only good for salvaging for scrap to me.

Yes, decent cardboard combined with high Qt drivers can work quite well. If you haven't already seen the foam core board thread, it's another good choice:

Historically, I've had access to a wide range of size/shape Styrofoam boxes, flat sheets, so have built speakers with high Qt drivers up to 12" diameter from this that you would probably consider huge. IIRC the boxes were Honda Gold Wing motorcycle packing 'crates' within the braced cardboard exterior 'crate'. Still got some rectangular 2-3 ft^3 net boxes in the attic awaiting 8" 'full-range' drivers for a computer 5.1 system.

Cut up flip flops, tongs, can be used for isolation footers.

I find suitable footware and cut them into squares and placed them under my 2-way boxes and their outboard crossover as well as under my valve amp. Rubber sheeting can be found and that is used in a wood sandwich for OB projects.

The amp did not have enough clearance to my liking for heat dispersion and the vibration isolation seems to help a small amount with microphonics. Placed it on one of the larger cutting boards on the carpet with the homemade footers under the feet on the chassis.

Placed an order for many pairs of widerange 3 inch and 4 inch drivers to put into my various cutting boards and the foam board I picked up today.

I see what some of you mean about SQ and FR when I listen to the 2-way box, but I think trying to get all things from all speakers is a mistake.