Kevin McMurtrie wrote:

My Roku's power supply is a mess of burned parts under burned glue in a smokey smelling woofer box. I did some searching and it looks like a Mean Well PD-2512 is a perfect match. Cut the tips of two corners off and it fits. Output is +/- 12V at 1A. Since the output is isolated, that can just as easily be +12V and +24V. It powers up the Roku fine and I'll do some more testing to see how stable the unbalanced load is. I'm not re-using that plastic box of metallic duct tape so more work is on hold until I build proper enclosure of perforated aluminum.

Are you going to run both rails (blue and red wires) on 12v or use both the 12v (for 9.5v)and 24v (for 19v)? If the latter, be carefull, I think the whole amp section can't take voltage that high (24v).
Just want to make sure, Matt

I've got a classic R1000 boat anchor and am looking into the options. I noticed the repair discussion Burkhardi has had lately with grandefuturo and was wondering Burkhardi if you have changed your repair procedure at all?

Do I understand the pros and cons of each fix properly:

Repair factory PSU:
+better components in place of cheap ones
+easier to do
-underdesigned (is this overcome with the better components?)

Replace PSU
+better components with better overload protection
+more modern PSU design
-less volume at higher volume settings if use 12V
-risk of killing the amps if go with -12 V + 12 V = 24 V (how big a risk?, I thought that grandefuturo felt it would be fine to run them at 24V)

Get factory coverage
+????
-????

Replace with external wall wart
+Easily replaced if fails down the road
+Thermal issues are gone
+relatively easy
-ugly Wall wart
-lower max volume

Anything I missed?

What about the extended warrany +'s and -'s?
Have they made any changes to improve the PSU life?
Does anyone know if they repair them or replacement them?

I know there are a few questions here. Hopefully someone can spend a moment for some answers/feedback. Thanks

Regards
S.

May be it's possible to drill holes to areate the PSU?

@ Hiphouser:
If you mean to drill holes in the black plastic housing that the woofer uses and the PSU sits in, then I would not. The enclosure is a tuned box and there is even a port on it. Drilling holes will make the bass sloppy and you may even hear it whistle too (if the holes are small). So in closing, I would not drill holes in the housing to vent the PSU if it were me.

@ TimeVacuum
I think it was McMurtie that brought up the 24v deal, not grandefuturo; but I have to go back and read.

Repair factory PSU:
+better components in place of cheap ones
+easier to do
-under designed (is this overcome with the better components?)

Yes, relative, yes this is overcome with the better components.

Replace PSU
+better components with better overload protection
+more modern PSU design
-less volume at higher volume settings if use 12V
-risk of killing the amps if go with -12 V + 12 V = 24 V (how big a risk?, I thought that grandefuturo felt it would be fine to run them at 24V)

relative, relative, yes, yes. The amplifier IC is rated at a MAX voltage of 21vDC. So with the OEM voltage of 19vDC there is no issues, with 24vDC your running it past the maximum rating and I think you will let the smoke out of the IC...

Replace with external wall wart
+Easily replaced if fails down the road
+Thermal issues are gone
+relatively easy
-ugly Wall wart
-lower max volume

Yes, yes, yes, some like wall-warts, yes lower volume.

This is the order I would choose for mine.
#1 Rebuild the OEM supply with Japanese quality parts.
#2 Wire up a 12vDC wall-wart and whilst your in there, leave the dead PSU there to keep the same interior air volume and then fill the outside plug there the AC cord went with silicone (or glue from a hot glue gun) so no one can plug an AC cord into it. Add a fuse and if you are really worried you can diode isolate the two rails.
#3 replace PSU with another type.

Reagrds, Matt

CAVEAT: I am just a tech and not an Engineer; I don't play one on TV but have stayed at a Holiday Express once. This is just my opnion and 2-cents worth so take it with a grain of salt. I am bad with a math and can't spell very well; but I can sure fix things and have plenty of Engineer friends to bounce ideas off of. Seacrest out...

Thanks guys for the vital info, I would not have been able to repair it without it. I bought my R1000 about 18 months ago and it went dead, figured it was the PSU seeing as it is overloaded and several other users have also experienced failures after about a year in use.

I bought it in Germany and the PSU has the same label mentioned by grandefuturo. I decided the easiest solution was to go for the external power supply (PS), as I had a spare 18V DC PS available from an old printer. I bought a barrel connector for 2€, made a hole next to the old AC inlet to fit the connector and superglued it to the housing. BUT I had to buy a soldering iron, a multimeter and a bit set for those damned TORQ screws attaching the 2 halves together, which cost me 15€ but I may use them again. Of course the manhours did not make the repair cost-effective, as it took me a couple days´ work (mostly researching the best approach and ordering equipment) but I get some satisfaction out of repairing stuff.

An external PS is IMO the easiest option unless you enjoy repairing PCBs and are able to do it (I would struggle). Finding an alternative internal PS that would fit where the old one lies would be easier, but I haven´t heard anyone come up with a suitable one. And I wonder if the internal PS is not only overloaded but may also have thermal issues, it´s obviously a flawed design (even the redesigned PS fails), maybe size and location constraints are the main problem? My radio may still be under warranty (I read Germany has 2 years) but I suspect they would install another PSU with issues which would fail after a year, so better to play it safe and go for the external option, and it has no aesthetic drawback for me.

If anyone follow my approach some tips:
-barrel connectors are by no means standardized, make sure your PS and connector are compatible and that the positive is connected to the inner contact, I couldn´t find a compatible connector for my PS so I had to change its connector.
-PS should be (if I´ve followed the discussion correctly) 7-19V DC, and rated for 550mA
-the housing is made of bloody hard plastic which would not drill easily, so I mostly melted it away with the soldering iron, by stages so that the connector would fit in snugly with the housing. Ensure the position of the connector and cable leads does not interfere with the back panel or any other parts on assembly.
-I doubt that for low voltages arcing is a problem, but just in case I used some of the AC inlet potting compound (melted some chunks off and remelted it onto the connector).

And to finish: shame on Roku for putting such a flawed design on the market. I bought if for 370€!, a lot more than its competitors and an outrageous amount for a radio, and I rarely used it, but evidently it´s unreliable even on standy mode. I expected a lot more.

Last week both of mine died - one I'd upgraded the caps and resistors and the other was a warranty replacement. Both within 24h so perhaps it was a mains spike (no other appliances etc dead though).

So out came the original power supplies and in went a tag strip with two 4700uF 35v caps and 9 diodes wired in series.

A 2A 15VDC regulated plug pack now provides the juice and the diodes drop the logic supply down to 9.5V

http://jaycar.co.nz/productView.asp?ID= ... rm=KEYWORD

I don't expect to be working on the power supply again... YAY!

Having previously boasted about repairing the PSU, it eventually failed again. This time I was so fed up I decided to remove it and run the Radio from an external unit.

To this end I thought I would share my experience.

The first problem was that I only had a 12v linear regulator in my component box, so I decided to run the 9.5V line from this IC. Measurements on a bench supply indicate that the 9.5V line drives a DC to DC inverter (or similar), as the overall power consumption did not change at the higher voltage. So although I am fairly confident all should be OK, I must admit I would probably have used a 9V regulator if I had one! (9.5V = 520ma, 12V = 420mA).

One benefit is the lower current usage at 12V means that the linear regulator and heat sink dissipates less power! This leaves more power available for the 18V feed.

Having constructed the simple regulator circuit I powered the unit up with the single 18V supply using a bench PSU. This consumed about 420mA at idle and even at full blast I was struggling to see more than 800mA!.

Hence I purchased an 18V 1A regulator from CPC PW00526 (UK) @ £15.73. Everything is now working a treat with no buzz or distortion.

I took a few pics along the way so feel free to have a look. Sorry some of the images are a bit blurry but they are mobile phone pics and I couldn't tell until I uploaded them.

http://picasaweb.google.co.uk/steveMroe/Roku?authkey=Gv1sRgCLqX3qP1koaXMA&feat=directlink

Cheers....

@ blubi

-PS should be (if I´ve followed the discussion correctly) 7-19V DC, and rated for 550mA

I think it's more of a 9-12vDC (caveat: without an extra onboard regulator for the 9.5 rail) with 12vDC being the popular one.
The amp's IC is only rated down to 8vDC (if I recall the spec sheet correctly) so not only would it probably not work (audio) at 7vDC, but you would have next to no output (IMHO) and 19vDC is too close to the 25vDC rating of the capacitor that is going to the regulator on the main PCB that steps down the 9.5vDC; I would worry about it failing. I think that is why 12vDC is the popular one. Also, I think 550mA is too low, that should be enough for the main PCB, but then you need the amperage for the amplifier section and that is what will take the bulk of the power. I would saya wallwart that is rated at least 12vDC @1A and I would go larger if you could just to let it run cooler, like a 12vDC @ 1.5 or 2A.

@sausagePaws

I must admit I would probably have used a 9V regulator if I had one! (9.5V = 520ma, 12V = 420mA). One benefit is the lower current usage at 12V means that the linear regulator and heat sink dissipates less power! This leaves more power available for the 18V feed.

Even though it uses less current, it's more or less the same power/wattage, right?

9.5v*520ma=4.92watts
12v*420ma=5.04watts
right?

If so, then the 12v one is a tad higher in power consumption, or perhaps I read your post wrong (plus I do stink at math).

I do like your approch though.
If it were me, and I was not going to fix and update the OEM power supply, I would either do a single 12vDC wall-wart and leave the OEM PSU in the chassis for air volume and just cut the mains inside and then use hotglue or silicone on the outside of the power plug (so someone does not try to plug an AC cord into it) and then add a barrel connector to the chassis with a fuse (and perhaps an isolation diode and some caps in the inside. Or do like you did and use a 18 or 19v wallwart with a regulator for the 9.5v rail or do like mark2c and use some diodes in series to knock it down (rates at the correct current/power). Anyway, just thinking outloud...

Thanks for the info!

@Burkhardi

Even though it uses less current, it's more or less the same power/wattage, right?

9.5v*520ma=4.92watts
12v*420ma=5.04watts
right?

If so, then the 12v one is a tad higher in power consumption, or perhaps I read your post wrong (plus I do stink at math).

Agreed - I think rounding errors are the main difference but I think they are close enough to justify that the power usage and hence dissipation is effectively unaltered.

The main benefit is the reduced current drawn from the 18V (1A max) supply - thus giving more surplus current / power to the 18v supply. (18 * (1000ma - 420mA) = 10.44W) as opposed to (18 * (1000ma - 520mA) = 8.64W). This is more important since the received supply is actually labelled as 18V / 0.85 / 15W so CPC details are slightly in error. As I said still works fine though.

Thanks for commenting....

Several months ago, my Soundbridge Radio R1000 died out of warranty. Since Matt Burkhard had previously repaired my Soundbridge M1000 and returned it in perfect condition, I sent him the radio for repair. Now, my R1000 looks and sounds like it was new. He re-assembled radio perfectly so that there is no evidence that it was ever opened. The only difference is that the new power supply should last much longer than the inferior factory part. I never met Matt, but I greatly appreciate his talent and the service he provide to the Roku user community.

hi there

my radio died last weekend after 13 months...

I was very happy finding this forum and the loads of infos about the broken PSU. so I opened it tonight, ending up confused about the following:

the capacitor looks blown up, the area at the bottom and over the diode looks burned. also the area around one resistor on the top looks burned (like on the posted foto).

I measured both output voltages expecting nothing or flapping voltage on the 9.5V as read in this forum. but I have the 9.5V and the amp's supply has about 25.5V.

so if the 9.5V is there, why is the box not booting up? I also tried to disconnect 230V and supplied 10.5V from an external PS. I did not disconnect the wires from the board, but that shouldn't matter... and still no boot...

any ideas?

Hi.

Presumably you measured those voltages on load?
Ie with the main unit connected.

If so 25.5V is far too high and may even result in main speaker damage.

Quite why the unit did not boot when you applied the 10.5v I cannot answer at this time, other than again suggesting that you check the voltages on load, and that you wired things up correctly. I have never actually had any unit that failed to boot, once the PSU had been correctly rebuilt that is. The logic board in these is very reliable.

The damage you describe is exactly what I see on every unit that I refurbish. For lasting results ALL the capacitors must be replaced and ALL traces of the burnt glue must be removed too, as this can and does become conductive.

I normally use a kit of about 10 components, and lots of patience when repairing these, you need it



Pete

g0pkh wrote:

Presumably you measured those voltages on load?

shame on me, I didn't....
on load, I measure the 19.5V stable and an instable 5.5 - 5.9V output.

so I'll go for it, and will order the capacitors for my upcoming soldering session. is it an open secret which 10 components you're changing?



alex

spiffy_monkeys description of guts of the Soundbridge ("dollops of glue"...) makes me not want to buy a Soundbridge again.

The assembly operation does not sound very professional. No wonder these things have a lifetime of 1 year or thereabouts.

My Soundbridge just turned 1 yr old in July...I'm waiting for its end to come any day now !

I've noticed that since changing to external plug-pack power supplies that the FM reception on both of my Soundbridge Radios (R1000) has improved.

They appear to be more sensitive, produce less sibilance and have less crackle/noise.

I use external whip antennas and these have not changed.

Has anyone else noticed a similar improvement in FM reception?