Anyone here own a Rel Sorm III sub?

[Buriedcode]

Hi,

First post so be gentle..

I've been meaning to join this forum for a while as I almost always end up coming across great information from here. I started my interest/vocation with electronics with audio and moved onto digital/power, well, its about time I got back into the swing of analogue audio design.

My first question is about repairing a Rel Storm III sub for a neighbour. It's an older 2002 model, and by the looks of things the bipolar driver transistors for the MOSFET have blown, taking a resistor with it. After contacting Sumiko about it, I managed to convince customer support to provide the value of he burnt resistor, and I have the datasheets for the driver transistors. (BFN16 NPN, and BFN19 for PNP). As these transistors are SMT SOT89 they can't dissipate much heat, which is probably why they have failed (even with a good copper heatsink on the PCB). They added through-hole footprints on the PCB in parallel, making it easier to use TO-220's (or TO-126) for replacements, so I was thinking, MJE340's and 350's?

I am not after 'improving' sound for my neighbour as he is happy with it, so the closer I get to BFN16/19 equivilents the better, and being a sub, high end frequency response is not an issue. I picked these transistors for availability, and their higher power dissipation, along with the possibilty of adding a heatsink.

There's also the MJE15034/35, which are most expensive, higher gain, greater power dissiaption, highg frequency response etc.. but I fear the high end won't be needed for a sub, plus, perhaps the higher gain would result in a change in sound, or require a change in the values of passives.

As cheeky as this is, if anyone owns a storm III, any cance you could post a pic of the power amp PCB? Specifically the driver transistors not the MOSFET's). I read that Rel upgraded the design in 2005 due to failures (most likely of this kind) so a picture of the later model would be ideal, as I could see any changes they made. As you can guess, Rel is understandably a bit hush hush on this, as they want money for repairs, plus, its a specialists field, for qualified 'rel engineers' only.

Ultimately, despite my (somewhat limited) understanding of audio electronics, I am not trying to reinvent the wheel. Purely to get the sub working again, as close to the original sound quality as posible, but improving reliability so parts don't overheat again. Hopefully he can enjoy another 10 years of it, without having to upgrade

Cheers,

Buriedcode

[Buriedcode]

Ok, so it seems I made a typo, its the Storm III if you hadn't guessed Also, its a 2002 model! so its coming up for its 10th birthday. Something tells me that PSU caps are likely to fail within ten years..

Seems it has now packed up completely, I replaced the burnt resistor - no dice. Just a constant buzz regardless of input. So, until I trace out a schem, my initial guesses are (remember, I've never worked on a power amp before..):

1. Failed power supply cap.
The positive rail reads a good 60V on DC, and with an AC check, 0.000v. Using 'AC' voltage measurement across a power supply doesn't exactly tell you much, because of he limited bandwidth of multimeters, but it can be a cheap way of seeing if there's big transients. The negative rail reads -60V, but on the AC measurement, it jumps around.. a LOT. I have to scope this to see exactly whats going on.

2. Failed output transistors.
I have checked the MOSFET's in-circuit for shorts, but will probably desolder these and check them again. There is high resistance between source and drain of all four, with the body-diodes Vf of between 0.4, and 0.3V. I'm guessing the low value is because it is in circuit, along with resistors, but a short would be <0.1V in my experience.

3. Failed driver transistors:
Although I have tested the Vbe voltage drop of all four, as well as the collector-emitter (for shorts) I will ahve to desolder these and test outside of circuit. I suspected these first, purely because two had discoloured solder joints (rust colour, a sign of reheating and burnt flux) as well as one being directly under the burnt resistor - which btw, was completely burnt, enough heat to lift the pads off the board, and melt the solder in the joints.

Aside from that, I don't believe its likely anything has failed in the preamp/EQ section, theres not really any power components in there. As you can probably tell, I don't want to just order replacement MOSFETs and power supply caps if there is no need. Some MJE340s/350's are on order - but I will replace the drivers anyway, whether they are the source of the fault or not, just because TO126/TO220's are much nicer to heatsink.

I know without a sound sample or a scope capture, chances are someone cannot diagnose the fault and I'm not expecting as much, but some guidance as to what to check first would be helpful

Buriedcode

[Dabble]

The REL storm III sub-woofer has the same amplifier as the Strata III, albeit with additional power transistors, and suffers the same crackling / buzzing when R8 overheats and deteriorates. A simple fix is to replace R8 (12k, 2 Watt) with a 12K, 3 Watt.

Comparing a repaired unit with an original, the commercial repair for Strata appears to be as follows:

Change Q9 & Q10 (BF469, NPN) to MJE340
Change Q4 & Q5 (BF470, PNP) to MJE350
Change R8 (12k, 2 Watt) to 15K, 3 Watt, 5%
Change R6 (10k, 1/8 Watt?) to 22K, 1%

The TO-126 transistors are direct replacements.
It's worth mounting R8 well clear of the PCB.
It's probably only Q10, and possibly Q5 that would benefit from (electrically isolated) heat sinks.
The PCB is double sided so soldering will need to be done thoughtfully.

REL say in their sales literature that with the quality build and components of these subs they should outlast their owners. Hopefully, now they might.

[Buriedcode]

Wow, those were the exact repairs I did.... the SOT89 BF469's/70's to MJE340/50 and R8 to 15K - used a 3Watt too.

I used a machined piece of aluminium for a heatsink for all four transistors, with isolating mounting kits, and fixed it to the board with M3 bolts.

The reason I believe the resistor, R8 (12k 2 watt) failed is because by my simuations /measurements, it was disappating around a watt of heat. So whilst 2 watts would be the bare minimum rating, it was mounted almost making contact with the board - under it was an area of copper used as a small heatsink for one of the transistors. Thats a big nono. Not only were the copper floods too small for adequate heatsinking of the transistors, but to add another 'heat disappating' component right on this was asking for a slow failure. Thankfully though, since they probably use the same PCB for many subs, there were footprints for TO126's/220's which I used.

The '15k' value for R8 was the only thing I didn't have to work out for myself. The resistor was so burnt its value was unreadable, so whilst I measured the resistor half way along (reading ~6k) and doubled it to get 12k, I also emailed the US sales rep for REL (cant' remember the company name right now..) who was quite helpful considering - he just said 'R8 should be 15k'.

Sorry if I've just repeated what you said, but its pretty weird that we both had almost exactly the same 'fix'. I can confirm that this works, the sub is now super quiet, no hissing, or bangs/buzzes and is still going strong.

I would have posted photo's, or a schematic, but I didn't want to **** Rel off - and although I'm a digital/embeded/RF electronics guy (never repaired, or designed an amp before..) it was surprisingly simple, and seems to be pretty much the same sort of design as most MOSFET amps I researched brfore starting the repair - no magic, just a straightforward quality design.

Thank Dabble for posting this. I wasn' unsure whether to post my repairs for teh above reasons, but also - I'm a noob here and didn't want to break any unspoken rules...