One of my sons friends asked me if I could take a look at his speakers which had stopped working. he and my son had been mixing some songs using a pair of M-AUDIO BX5a monitors. These are bi-amped active speakers, with Pro audio style inputs (XLR and 1/4" mono jack). They each have two integral amplifiers so only need a line or balanced input. Each speaker also has a mains power input and a volume control, and a blue LED on the front to indicate power. The speakers have two drivers, a small domed tweeter and a small 5" Kevlar woofer. One had developed a loud buzz, coupled with a very loud 'thump' sound when powered off. The other had completely lost all bass output.The speakers were only 13 months old, which seemed a very short life for a pair of speakers - I have speakers made in the 60's that are still going strong. However with additional electronics in these BX5a's I'm not really comparing like with like.These speakers are self contained audio systems. All you'd need is a source like an iPod or CD player and you'd have a pretty decent stereo system. They are primarily aimed at pro audio monitoring use and retail for around £200. They are also very well reviewed on the web. I was also intrigued by the name. My own main speakers are a pair of Goodmans BBC LS3/5a speakers - something of a legend among small monitors. I thought the naming of the BX5a was perhaps a nod to these classics.
My first advice was to see if the shop he bought them from would be prepared to fix them as they were not that old. They would only fix for free when the speakers were within warranty, which these were not. Oh well ....I googled the speakers and some of the symptoms and pretty soon found dozens of hits on the buzzing problem, with the loud thump on power off . Again and again people 's BX5a had developed loud buzzes through both tweeter and woofer. It seemed that the speakers had fallen victim to a more widespread problem of poor quality capacitors. This problem has been the scourge of the IT Industry too, I have had a Dell Optiplex fail the same way. I understand that Apple Time machines have had problems with caps leaking/exploding too. Possibly it's due to the caps being a faulty batch or the temperature upper limit being exceeded and prematurely ageing them.
The power supply in the speakers uses a couple of 6800uF 25v caps to smooth the output from the 4 diodes making up the bridge. NOTE : the switched mode supply, ubiquitous in every other aspect would be too electrically noisy for these applications. Here we have good old linear supplies with diodes, and smoothing caps.These two caps appear to have a common problem where they leak. The top of the cap appearing to bulge, and in some cases electrolyte oozing out. These symptoms certainly seemed like one of these speakers , so I offered to replace the caps . I figured it was within my capability.
I should state at this point, that this is quite a hard repair to carry out. Some forums imply this is easy, but it took me around 90 minutes to do the first speaker, and 60 to do the second. You have to substantially dismantle the speaker and remove two caps which are soldered and glued in position. There are a lot of steps along the way.
I looked on-line at caps and in the end RS Components did suitable caps 6800uF 25v . In addition other important values are temperature (105 degrees) diameter (18mm) and lead pitch (gap between the leads - 7.5mm). The Cap length is not so important but you need to get all the other parameters the same. RS and Farnell have thousands of caps, but by the time you specify all these constraints, you are down to one or two , I went for these Nichicon caps. RS states that the caps come in packs of 2 , so I ordered a total quantity of 2, assuming this would equate to 2 packs of 2, meaning 4 caps. I had decided that while the second speakers problem was different, it was worth replacing the smoothing caps in both. It might fix the other problem, and I reasoned that it looked like this might become a problem in the second speaker at some point anyway. The package arrived 2 days later but only included 2 caps. I think I had been confused by the quantity & the fact that 2 came in a single packet. Anyway I ordered a second batch and set to work on the speaker with the buzz. When dismantling these speakers you will need :
WARNING : DISCONNECT THE SPEAKERS FROM THE MAINS BEFORE ANY OF THESE STEPS. THESE SPEAKERS CARRY MAINS VOLTAGES INSIDE SO IF IN DOUBT, GET A QUALIFIED ENGINEER TO CARRY OUT THIS WORK.
I put a towel down on the table and laid the speakers face down. The front does not have a protective grille, but the drivers are recessed enough into the bezel that they don't protrude and laying them on their front is fine . 10 Philips screws hold the back panel in place and I carefully removed these, cupping my spare hand around the Philips screw driver point, so as not to slip and scratch the cabinets. This is more important later when removing the bass driver. It's very easy to stick a screw driver through a speaker cone if you are not careful - I have done this before and it's easy to avoid with care. Once all the screws were removed, the back is loose, but I soon realised that the cables inside connecting with the drivers on the front , were too short for me to gain access to the amp Cables ties had been used to gather up any slack, so I couldn't get the back panels with attached amps clear via the back.
I turned the speaker over, laying it carefully on it's back. the towel is important as the volume knob on the back protrudes, so laying the speaker on something a bit soft will save your tabletop and the speaker volume pot. The front bezel is held on with No. 3 Allen key headed screws. At first this seems frustrating but actually is a smart idea. It means you can apply pressure in a plane across the speakers, not toward them (as with conventional screws). I still cupped one hand over each driver, while removing the Allen screws with the key in the other hand.
Once they are removed you can lift the bezel. The bass drivers are not attached to it, they are still screwed to the front of the cabinet. However the tweeter is attached to the bezel. You will need to disconnect the speaker wires from the tweeter to free it. The wires are coloured black and white and have identical sized pushed connectors. I photographed the tweeter at this point so I'd remember the tweeter polarity.
You can pull the speaker wires from the tweeter but take note of the polarity or better still take a picture. Once the tweeter is disconnected you can take the bezel and tweeter and place safely to one side. You then need to remove the woofer. This is held to the speaker cabinet by 4 Philips screws. Again, cup one hand over the driver while you carefully remove the 4 screws. It' very easy to slip and pierce the driver with a screw driver - I know, I have done it before, and kicked myself for my stupidity.
Once you remove the bass driver you do not need to take note of the polarity as the push connectors are a different size for positive and negative. Note the silver wires behind the tags - their significance is important - read on.
Once you have disconnected the woofer, place it safely to one side.
You will find you still cannot remove the backplate as the LED which indicates power is fastened to the front of the speaker. The cable slack has been wound with the excess speaker cable using a cable tie. I carefully cut the cable tie, releasing enough slack cable to allow the amplifier module to be removed from the speaker while leaving the BLUE LED cable in place.
Once the amplifier module was out I could see that the smoothing caps were failing as predicted.
You can clearly see the electrolyte oozing out of one of the smoothing caps and that the top of the cap is curved outward. Something inside is trying to get out !
To replace the caps you will need a de-solder pump or de-solder wick, a soldering iron and a craft knife.
However before you can gain access to the back of the board there is a metal plate which has to be removed. You can see the plate in the picture below. It's held on by a couple of screws on the right hand side, which also hold two earthing wires in place. These screws have been secured with green enamel paint or possibly nail varnish, which I had to chip away to access the screw heads. On the left a cable tie through a hole secures the plate.
I removed the screws on the right hand side and cut the cable tie on the left. The plate is further secured to the back of the circuit board with an adhesive strip. With gentle pressure the plate slowly came away. This gave me access to the back of the capacitors. I took a picture to remind me of the polarity of the cap leads (negative -ve and positive +ve). You can see below how bad one of the caps appears. The sticky electrolyte has got some of the speaker acrylic wool stuck to it, making it appear hairy !I heated the solder pads on the reverse side of the board where the capacitor leads passed through. I used the solder sucker to remove as much solder as possible. I then carefully cut with the sharp craft knife around the base of each cap to cut through the glue which has been used to further secure the cap to the board. This is a slow and delicate operation, hence my comment earlier about this being a difficult fix. You don't want to damage the surrounding components nor flex the board. It's possible to raise the copper tracks from the board if you apply excessive heat or try removing the solder too many times.
I did more de-soldering until I could bend the remaining length of lead up, allowing the cap to be slid out from beneath . Eventually each capacitor slowly loosened and I could gently pull them away. I further de-soldered the remaining holes and trimmed as much yellow glue from where the caps had been. It was then a case of inserting the new caps, checking the leads were correct. (Negative lead has a white strip down that side of the can ), re-soldering them and cutting the spare from the protruding leads.
If you chose the right value of capacitor ( 6800uF, 25v, 105 degrees, 7.5mm lead pitch, 18mm diameter) you should find that they fit perfectly, though might be a little taller, but there is plenty of space.
Once fitted I carefully reversed the process, fitting the plate back and applying a new cable tie through the corner hole. Fitted the rear panel in the box and then the bass speaker, finally the bezel. Always remember to use one hand to protect the drivers - that treble unit has a strong magnet in it and a wandering screw driver can easily be attracted to the soft dome if you are not careful.
I repeated this entire sequence on the second speaker, thinking that it a) might solve the other problem (loss of bass) and b) is probably a good thing to do. While the capacitors were not in as bad condition, there was a distinct 'doming' of the ends. I also checked the continuity of the two fuses on the amplifier board. Both were fine.
Powering up each speaker revealed no buzzing, the LED still came on and no big 'thump' noise on power off. Placing my ear next to each driver revealed a small amount of background hiss from each treble unit and a tiny amount of background hum from the bass, but this seemed normal to me. Unfortunately I have no interconnect cables to test the speakers with to hand, so returned them to there owner with the advice that he try them ASAP.
Well the good news was that the speaker that had buzzed, now worked perfectly. However the speaker that had no bass .............still had no bass.
I realised at this point that I had omitted to do the single simple test which would have saved me the following rework . I should have tested to bass speakers continuity while I had the driver out i.e tested that there was a circuit between the positive and negative terminals on the back of the disconnected speaker. An open circuit /no circuit would have indicated the speaker had failed internally, as it now appeared it had. Of course it could be other problems with the power amplifier ?
The offending speaker was returned to me. I had also found a cable I could use to test with an RCA plug on one end, and a 1/4" stereo jack at the other. The BX5a needs a mono 1/4" jack , but I knew it should work if I slightly remove the jack so that the internal contact touched either the left of right band on the jack. This was a mono cable, so both were wired to the signal in a single RCA.
I repeated the steps above as far as removing the bass driver, and tested it for continuity. No circuit ! This was good in so far that had the driver blown completely it was a relatively easy procedure to order and drop in a new driver.
I wanted to double check the speaker was not functional , so wired it up to one channel of a small desktop Sonic Impact T-amp I had. I turned the volume to quite low and played some music. Without a cross-over the sound would be bass heavy, but I figured it should produce some sound.
Nothing, as the continuity test had predicted.
I sat down and started web surfing to see get a cost for a replacement bass driver. The M-AUDIO website was not very helpful here. I could find no links for spare parts.
I stared at the driver and , without much thought double checked the continuity again. My trusty TANDY/Radio Shack digital meter has a simple beep continuity mode, where it beeps if it sees a circuit. I tried the positive and negative terminals again. No change.
Then I tried the wires that lead to the terminals from the centre of the speaker voice coils - short lengths of silver wire.
BINGO ! We had continuity.
Further tests showed that the negative terminal of the bass speaker was open circuit. It looked OK, but it appeared that the solder joint was not good. I de-soldered it and resoldered it, and the speaker was fine. I put the second speaker back, after testing it quickly outside the cabinet with my CD player (A Sony Playstation 1), my RCA to stereo 1/4" jack plug and a much loved copy of "The Kinks are the Village Green Preservation Society".
Worked fine so put the speaker back together again. Job done.
A final puzzle: "phased and confused"
It's impossible to connect the bass drivers incorrectly as the spade connectors are different sizes for positive and negative, red and black. However the connectors on the tweeter are the same size.
I noticed that the tweeter connections are mirrored. Rather than both tweeters being connected the same way (black left , white right connector), the arrangement is inverted in the second speaker. I'm not sure about this. I don't know too much about this tweeter but would assume, that like the bass driver, both speakers should be wired identically to preserve absolute phase ?
Here is one speaker, with the bezel disconnected, photographed from the top:
Here is the other speaker, with the bezel disconnected, photographed again from the top:
I guess if the treble sounds a bit odd, you might want to check if your treble units are wired in phase or out of phase with each other ? I think they should both be in phase, both with each other and with the bass drivers, though how you tell is unclear as neither tweeter connection tabs were labelled.
Postscript: revisited 19/4/2012
My sons friend contacted me to say that one of the speakers had lost its bass output, and he thought it was the one that originally had the hum problem i.e not the one that previously needed the bass tag resoldering.
He dropped the speaker off, we followed the dismantle sequence above, I removed the bass driver and put a multimeter in continuity mode across the two tag terminals on the speaker. Sure enough - nothing. I then repeated this but gently putting the probes on the silver wires behind the tabs. Once again i had continuity i.e measurable low resistance, not open circuit. By trying the four combinations of two probes on two terminals and two wires , I figured out it was the right had tab looking at the speaker with the woofer facing upward.
Pleased it was a problem we had encountered before i used a solder sucker to remove the solder around the connector, then resoldered it. I put the speaker back together and played some music, all seemed well as "Picture book" by The Kinks played out. A great track and as i had the mono recording it gave me a familiar sound through only one speaker. Pleased with our results , my son suggested something louder and so "Achilles Last Stand" by Led Zeppelin, which is in stereo so we'd only get one channel, but still good. I mention this because half way through we lost all bass. The problem had returned !
Repeated the process, but this time spent longer using the solder sucker . When the connector tag was as clean as i could get it and i could see the hole through it, this time I poked the silver wire from the voice coil through the hole from back to front and gently bent it over. Don't pull it too far through, otherwise you could damage the driver. Just a mm or so, then held it with tweezers and made a good solid solder joint.
Repeated the Zeptest and it survived all of Bonhams drum fills, and is working and back with its owner.
....So it looks like some bass speakers, certainly the ones in this pair, suffer from a poor solder joint , in both cases on the black wires terminal, where the voice coil wires connect to the push on connector terminals. A good re-solder seems to fix it.
I have since been advised that this is a stress fracture. The braided wire which liks the coil to the tabs has to flex. If soldered badly the solder creeps along part of the braid making it stiff. At this juncture between solder and non-solder braid, you get stress ractures. A tip which I will try next time is to de-solder the braid completely and apply some super glue close to the end where the-joint will be . This will prevent solder from creeping along its length by capillary action when molten.
Another BX5a and a BX5 (update 16/9/2012)
My Son wants to get a pair of BX5a and I was on the lookout on ebay for any examples for spares or repairs, reasoning that I could get them working , and that even a new pair might need the above fix anyway over time.
Not many are listed for spares or repair but eventually a single BX5a came up and the seller also listed a single BX5, its earlier precursor. The seller had sold the working second speaker of each type to a friend. We hatched a deal where if I could get both broken examples working , we'd swap and I'd get the BX5a's and the friend would have the working BX5. What could possibly go wrong ?
I should add that the seller was completely honest about the speakers, describing the BX5a as having the hum and thump problem as above. However the BX5 had a "static" sound. I wasn't sure but thought it worth a go.
The BX5a responded to exactly the process described above. The bezel was slightly different . As before the main problem is getting the caps off the board, where careful , firm pressure is needed with a craft knife.
The BX5 is a completely different beast. Different drivers, bezel, and different inards. Actually marginally easier to work on as each drivers cables attach to the motherboard with push connectors, as does the LED. The problem was, as described. When I turned on the speaker with no connected source, there was a loud rustling through the bass driver, rather like someone rubbing a Microphone. This was independent of volume settings. The BX5 also has a number of EQ settings and these too made no difference. The Caps looked OK, no leakage or signs of stress.
To be honest at this point I was out of my depth with the BX5. I asked a local repair guy who I have used before to have a look at it. He concluded that the chip amp was the source of the problem, but having de-soldered it and re-soldered it once, he was very dubious he'd be able to complete the maneuver a second time with a new amp. The boards are not made to be fixed.
So, I have a working BX5a and a non-working BX5. A gentleman at the other end of the country has a working BX5a and a working BX5.
If anyone (ideally UK based) has either :
1) a BX5a with the hum & thump problem they want to sell
2) has a single working BX5 they want to sell
please contact me.
Alternatively if anyone wants a BX5 for parts ( drivers, cabinet, bezel), also please contact me through the comments section
Postscript February 2013
Well my son needed a pair of active monitors for his college work on music production, and I'd only managed to source one fixable BX5a from ebay see above). hence I faced the need to buy a pair of BX5a or similar. In the end I opted for a pair of Fostex 0.4 monitors, similar to the BX5a in terms of features. They are a little smaller. he seems pleased with them. Why didn't I opt for the BX5a ? Well the story above shows that while they have a good sound, I think their build quality could be better.
September 2013 - More pictures
I had a comment on the Blog from Joey McGowan asking about the wiring from the transformer. I hadn't taken pictures of this area, so opened up my one remaining fixed BX5a and took a few more. Its quite densley packed in the area of the transformer, AC selector, rocker switch and AC socket, there are also shink wrap "boots" over the backs of the AC socket, switch and AC voltage selector, with cables ties. I left these on as I don't really want to have to replace.
I took these notes which I *think* are right (please double check) and I hope are useful combined with the pictures below . Please double check and remember that these are correct for my unit, but colours may change. Anyway here goes:
White -> Rocker switch
Red -> AC voltage selector
Yellow -> AC selector
Green -> AC selector
Dark Blue -> Mains IEC socket
White -> TX
Red (Fat, different wire than from TX) -> Mains socket
Yellow + Green combined Earth -> Screw on Sheild and Heat sink
Dark Blue (Fat , different from TX) -> AC selector
Dark Blue -> TX
Red (FAT) -> Rocker switch