Creative Labs SB0300 24bit external USB sound device

Faulty Creative Sound Labs SB0300 24bit external USB sound device.

The fault symptoms are: The internal RED SPDIF optical digital audio output LED is lit but the power, CMSS and audio mute LEDS are all OFF and the unit is not working.

The fault turned out to be a shorted turn on the internal switched mode power supply series 220uH inductor that would go away once it was heated gently. Once heated up it would continue to work correctly until powered OFF and left to cool down whereupon it would fail to start up again.

The internal power supply uses a series and shunt P channel and N Channel MOSFET transistors respectively, to obtain a stablised +4.8VDC and -4.3VDC power rails. The specialty switched mode power supply IC is a LTC1771 device which is essentially designed for "current mode" of operation with a second voltage feedback loop.

The circuit is robust in that if an overload occurs the LTC1771 will limit the peak current and the ON time duration and this is what was happening with my unit.

I could not open up the special surface mount ferrite core used for the inductor as it was cemented together and not being able to obtain a suitable replacement core I made a rather drastic modification as a "work around" for the problem. Other more astute readers may be able to get the 220uH inductor off the board, open it up and re-wind it with new wire.

My solution was to make a little daughter circuit board with a -5VDC generator circuit so that the shunt N channel MOSFET could be removed. The inductor was left in situ as it no longer matters if it has a shorted turn, the P Channel MOSFET will essentiall be ON all the time and given the 240VAC mains power pack has a nominal ouput of +5VDC the switching regulator is almost superfluous.

Many others have had internal power supply problems with this particualr model but another common problem is faulty electrolytic filter/storage capacitors on the internal +4.8VDC rail.

An exatract of the circuit involved is attached as is the -5VDC generator and some photos of the unit.

 

Hello Anton, I do not have the circuit diagram for the SB0300 unit, I did send Emails to Creative seeking help and copies of the circuit diagram but they did not reply so I then telephoned them in Singapore and the reply was that they do not provide this information which I thought was quite unhelpful given this model has not been produced by Creative for nearly eight years now.

So I'm sorry but I cannot help you with a circuit of the complete unit, the circuit of the power supply section that I have provided I produced myself by tracing out the circuit board, it would be very difficult to trace out the entire circuit and creative use several specialty IC's and there is no published information for these.

 

Hi Nexusian

The DC current required for the SB0300 model is about 600mA which exceeds the USB 1/2 current limit of 500mA hence the need for an external power pack/supply. The 220uH inductor therefore needs to be able to handle this current without saturating and the surface mount ferrite core used has an air gap to prevent core saturation.

It is very difficult to remove the 220uH inductor without first removing the electrolytic capacitors surrounding it because of the amount of heat required due to the large mounting pads of the inductor and there is a major risk of accidentally de-soldering all the components mounted underneath the inductor on the other side of the printed circuit board.

First of all you need to establish that the inductor is your problem though as it is rare to have a shorted turn these days, in other forums several other people have had problems with the SB0300 power supply due to bad electrolytic capacitors made in Taiwan between 1998 - 2005 (http://www.badcaps.net/). The bad electrolytic capacitors are usually identified by bulging tops or even rupture due to the build up of hydrogen gas but this is not always the case. The capacitors can go either short circuit or open circuit and in my experience they usully go open circuit which is even more catastophic if they are used in switched mode power supplies to filter out the peaks to produce an average voltage.

You can measure the 220uH inductor in circuit but it's best to do so with a DC current of 600mA, alternatively if you have a surface mount re-work hot air gun, carefully remove Q3 the shunt FDC655 N Channel MOSFET transistor used to produce the negative 4.5 volt rail and see if the +4.5 volt rail comes up. You will need to use the aluminium shielding tape to remove Q3 so that you do not accidentally dislodge the adjacent components which is easily done.

The 220uH inductor is not needed for the +4.5 volt rail as the SB0300 is fed from an external +5VDC power pack but the inductor is critical for the -4.5 volt supply as the shunt N channel MOSFET is turned ON at the same time as the series P Channel MOSFET and the 220uH inductor is used to store energy and provide a high impedance during this time.

If you are really sure the 220uH inductor is at fault then if you cannot easily buy a sutaible replacement that can handle 600mA then the alternatives are:

1. Remove the old one somehow and get it apart and re-wind it, boiling in hot water is worth trying to get it apart and prevent damage to the plastic bobbin.

2. Wind a new inductor on a suitable core

3. Install the small daughter circuit board with the simple -5V DC generator as I did.


I put together a quick spreadsheet with a simple VBA program to allow the calcualtion of inductance with an air gap and the subsequent value of the DC H magnetic field amplitude for a DC current of 600mA.

A toroidal core without an air gap would need be about 20mm in diameter to prevent saturation but a 10mm diameter toroidal core would be suitable if a 0.1mm air gap was present. Alternatively a 10mm small potcore will do the trick with the characteristics as detailed in the attached spreadsheet (ie Area = 15 square mm, Mean Magnetic Path Length = 15mm, relative permeability = 2200, Hmax at the BH curve knee = 160A/m, total of all air gaps = 0.1mm, the number of turns = 35).

Of course correcting the 220uH inductor means that you still have to remove it with all of the pitfalls in doing so.

Best regards Frank.

 

Creative Sound Labs SB0300 Inductance Calculations

The missing attachment from my last reply, the attached zip file contains an MS Word document and Excel spreadsheet detailing inductance calculations for cores with and without air gaps.

 

Steve beat me on this response and I can't see why powering it from a USB3 port should not work as long as you get the polarity right with the USB2 Type B receptacle and the power socket. Pins 1 & 4 are +5VDC and ground respectively for a type B USB socket/receptacle. If you are happy to always power the SB0300 from a USB3 port then to stop the accidental connection of an unregulated power pack which could blow up your computer I would either:

1. remove the power socket from the SB0300

2. block it off

3. Use a schottky diode instead of a wire from pin 1 of the USB connector to the nominal +5V pin of the existing power connector and that way you can keep the power socket and at worst you will trip the polymeric fuse in your computer if an unregualted power pack is connected that is lower in voltage than the computer +5VDC supply.

Remember, the SB0300 has an in-built switched mode regulator so can operate from unregulated power packs.

Whilst reputable computer manufacturers protect the USB2/3 power lines from continuous short circuits by way of self healing polymeric fuses they seldon have over-voltage protection and rely upon the low impedance +5V supply to limit transients.

 

It looks okay to me Julian, centre pin of the power connector is positive and pin1 of the USB connector was not connected to anything and pin 4 was connected to common (ground) as made by Creative Sound Labs so it should work fine.

Of note the power pack that comes with the Sound Labs SB0300 is rated at 5VDC 1.5Amps but I have not measured more than 600mA consumption in my unit so powering from any USB3 port should work fine, as Steve said the USB3 standard specifies a maximum current of 900mA for USB3 ports. If you retain the power connector so you can still use the external power pack instead of the USB port then again it will be okay as long as you don't have the USB cable connected as a precaution or use a schottky diode to connect from pin1 to the power connector centre pin. The diode will at least block the power pack from reverse feeding into your computer and using a schottky type diode will ensure the minimum voltage drop, the voltage drop from a normal silicon junction diode may be too large at 600mA for the SB0300. I salvage my surface mount schottky diodes from old computers or computer equipment (CD or DVD drives etc).

I think that the reason Creative used an external power pack for the original SB0300 was simply that the standing power consumption exceeded the then USB2 standard of 500mA, the SB0300 was re-designed to work off a USB2 limit of 500mA but at the loss of the SPDIF electrical and optical inputs, the re-design only had SPDIF outputs but did not need or indeed support an external power pack.

 

Interesting post andreidi, in the USA (where I'm not from) they would say that "you got lucky"

The 500mA rating on USB2 ports is really a nominal limit and not hard and fast, all the Laptops and Desktop motherboards that I have worked on in the past ten years have used polymeric self resetting "fuses" to protect the internal 5VDC supply from external short circuits and these devices are not exact current limiting devices but rely upon internal self heating to rapidly increase the DC resistance. As such it's entirely possible that different computers will allow perhaps 600-700mA to be drawn with still only a 0.5 to 0.75 volt drop from 5V which may well be enough for a lot of equipment to keep working but that limit will also be dependent upon the internal temperature of your computer, you may be on the brink of failure but not know it.

I have deliberately modified the externally powered 6 port USB hubs that I have got by cutting the internal +5VDC circuit board track(s) so that power from the external power supply cannot be fed back into the host computer but rather the external power supply only goes to four expansion ports and power from the host goes to only two expansion ports. A lot of the +5VDC plug packs are of such poor quality that they feed high amplitude common mode voltage spikes onto the low voltage secondary outputs. Most computers do not have internal blocking diodes on their +5VDC rail to the USB ports.

As an aside, some idiot has created a USB computer killer module that steps the +5VDC up to 200VDC then dumps the charge back into the computer USB power line specifically to kill the computer.
(http://www.itwire.com/security/74750-usb-stick-will-kill-pcs-–-on-purpose.html)