Linn Ikemi CD player

[Radio Wrangler]

Now why would they have a high voltage linear regulator in an SMPS?

The only thing that comes to mind is for a start-up supply.

Quite often in SMPS a dropper resistor/zener is used and the high value droppers have a habit of going high, leading to a supply that fails to start. Apple supplies were well known for this and it makes for an easy repair.

Have you identified which control chip it uses, Michael? Data on that and its application-note circuits is a good starting point for unravelling what's going on. You may have just seen that regulator and thought as the largest device on the board, it had to be the main switch. In a good, efficient, design there will bee very little heat in the switch device(s) so they could be fairly small.

With SMPS, in the past, I've identified their main power rail and used an external bench supply to operate the control chip and checked for oscillation and gate drive to the switch device(s) all this with no mains applied. A second bench supply can then be used to put a lower voltage , say 12-30v onto the reservoir capacitors. This is enough to see if the switch device is alive and to get rectifiers turning on a bit on the secondaries for outputs over 10v

In this way you can try power without risk of destruction or personal danger. Bench supply current limits protect hardware.

I've been involved in SMPS design, including some very weird ones, and in trying out something new, you don't even have the comfort of knowing it once worked. The low voltage supply trick pays its way. I had a 350v bench supply for when I wanted to get things going properly and see it going into proper regulation. Only when something worked on current limited supplies would I go on to applying mains.

David

[Michael Maurice]

Originally Posted by Chris55000

Hi!

The VB408 wasn't a switching device, it's only an LM317 type circuit made to withstand up to 450V h.t. Voltage differential!

Have you tried using the Supertex/Microchip LR8 with an additional small high–voltage pass transistor between the LR8 and the load?

Something like an MJE340 with C to the H.T. input, B to the LR8 o/p and E to the load?

The original VB408 was only limited to 40mA max – you might get away with an LR8 if you clamp it to a heatsink – the LR8 is rated 30mA!

You can also get discrete H.T. regulator PCB kits that will operate up to about 450V as well – one of these might get your unit working if there's enough room inside!

Chris Williams

There is no room inside for a heatsink and to be honest i'm somewhat unhappy with trying to modify SMPS's when I dont know enough of what I'm doing.

This unit is not mine and I cant take a chance of something going wrong and causing further damage or danger.

I'm far more comfortable with either repairing what I've got or redesigning a conventional PSU when i know what the outputs are.

[Radio Wrangler]

It ought to be repairable. Despite the name on the front and its status as a culturall and religious artefact, it's just electronics.

At 20W rating, I can't really see any reasoning for the use of a switcher. It's hardly going to be much smaller than a 50Hz transformer job, but we're outside the rules of normal science in this area of hifi. Strange things are done simply to be strange. But this PSU will work by rational means.

Some photos might help people assist in unravelling it.

David

[Ted Kendall]

Originally Posted by Radio Wrangler

At 20W rating, I can't really see any reasoning for the use of a switcher.

Worse, the Brilliant was sold initially as an upgrade for the then-current Linn CD unit - Karik, I think. No reson for this, of course, unless the original linear supply was a dog...

[Piccy2valve]

Theres mention of a link to one version of brilliant smps in this thread if it helps

https://pinkfishmedia.net/forum/thre...s-naked.51778/


Poss direct link here
https://www.diyaudio.com/forums/anal...atic-smps.html

Mark

[Radio Wrangler]

I've found some photos etc on the interweb. It looks like an expensive custom cylindrical housing with two boards inside it.

One board has mains input, a RIFA timebomb capacitor, rectifier, reservoirs and an automatic bridge/doubler switch with a triac. I assume that's the triac with heatshrink over it. There's a plastic-jacketed TO220 which might be that ST high-voltage regulator, which I suspect powers the control chip(s) There will be two resistors near it to program it's output voltage.

The switcher proper is on the second board. I see two capacitors which look like one side of a half-bridge. Are there two small power mosfets near the big transformer and the two orange capacitors? The control chip is said to be on the underside of this board.

Discussion of faults seem to cover a low temp 85C 47uF electrolytic on the mains input board that gets cooked, and also short-circuit tant reservoirs on the switcher board.

Swapping these might give you a quick stab at getting it going. It's a shame the fancy housing stops airflow, this may be a the basic cause of unreliability.

https://andydoz.blogspot.com/2019/01...ly-repair.html

However, there is mention of a 'slimline brilliant'.


I've found a photo of the outside... this is a conventional rectangular box. Probably a cost reduction exercise. It's lost the fancy housing. Rectancular PCB is cheaper than round. Now 20W PSUs hardly need voltage range switches, automatic or otherwise. A single range PSU can routinely be made to span 80v (Japan on a bad day) to 260v (UK in cheerful times) in one range. So I guess that's one change to make it smaller.

Someone is trying to sell one and wants the best part of 200 quid. It must have the fairy dust intact.

There seems to be a lot of people looking for these things. Are they dropping like flies?

No real information anywhere.

OK Michael, I don't think anyone can do much without photos. And then I expect we'll find it a lot more mainstream than the round one... and likely a lot simpler.

David

[Michael Maurice]

Its not the cylindrical version, this is the later slimline. I've attached pictures, the switching device has been removed, the two diodes under the red capacitors are short circuit. The output has four capacitors.

I hope this helps, though I may soon give up!

[Ted Kendall]

When the mains transformer on my Quad FM4 failed after more than ten years' service, Huntingdon fixed it for a modest charge in double quick time. Go figure why I stick with Quad.

[Piccy2valve]

Michael ,
The second link in my post #25 has some schematic and discussion on the psu but i am not sure itx the same model as yours

[Radio Wrangler]

I'm not convinced that was the main switch you removed, Michael.

The two orange capacitors right by the main transformer look like the capacitor pair of a half-bridge, and the two larger semiconductors under them... with their centre pins chopped look like the the pair of power MOSFETs of a half-bridge The pins in sight are gate and source. The drain is connected via the heatsink tab soldered to the board. Chopping the drain tab short is a common trick because the SMT pad pattern for the board for that package is a bit low on creepage distance for the voltages involved.

To the left of the transformer is a 1n0 capacitor and a 6-pin semiconductor. I guess the latter is an opto-isolator as the regulation feedback path across the high voltage isolation cordon, and the 1n0 is an RFI capacitor also across the cordon. The output rectifiers and filters are crammed hard into the bottom part of the board, below the cordon.

Its general appearance is that this board has been run hard and hot for a long time.

The next thing to do is to identify the two control chips. I guess the lower right one is a half-bridge gate driver. The two large-ish SMT ceramic capacitors would fit with this idea, though plenty of SMPS control ICs have the drivers buil in, but still need the capacitors. So if the right hand chip includes the controller, then what is the left-hand chip?

The two Schaffner lumps are common-mode anti-RFI chokes. The orange gumdrop at top right may be a metal oxide varistor to protect against surges, the black cylindrical thing below it may be a thermistor to control inrush current at turn-on The two connectors on the top are at a guess L and N inputs on one and maybe a voltage range switch on the other?

THe part you removed, below the reservoirs has its centre leg joggled out of line with the other two. This is another trick to increase creepage distances for high voltage parts. Could this be that ST high voltage regulator chip?

If so, then it may be the startup supply. The other bits around it may be the main supply for the control circuits, derived from a winding on the main transformer.

The close proximity of the half bridge (orange) capacitors and the switch MOSFETs of the other arm of the half-bridge is a good way of minimising radiated field. The copper strap around the transformer will be soldered along its join, it is a shorted turn around the outside of the whole transformer, designed to stom flux leaked from the core.

This looks fairly followable and with the data for those two chips ought to be manageable.

The question comes down to cost and time to repair. On the other hand, these things seem to not exactly be blessed with wonderful reliability and there are people chasing spare units. That sounds like a recipe for a worthwhile market. If you get this one going, you will have learned enough to be able to fix others, and there seem to be plenty of others out there with people who want them fixed.

In my hifi system, there is a Sony XA5400ES CD player. Inside, it's well made with linear supplies and iron mains transformers (With R-type core shapes) Their marketing schtick is having separate mains transformers for analogue and digital sections in the hope that some people will think it makes a difference. But nice, low stress linear supplies might make up for in reliability what they lose in efficiency. Looking inside the Linn stuff, I don't see anything added which fits with the dramatically increased price.

David

[Craig Sawyers]

Linn ought to hang their heads in shame to use an SMPS that is unreliable in this day and age.

Since design of an production-ready SMPS is somewhat specialised, I strongly suspect that they paid someone to design it, and then outsourced manufacture probably to points East.

And I agree that Quad, even though owned by the Chinese, still have a superb team in Huntingdon who have a limitless ability to help in keeping many decades old kit up and running. A breath of fresh air as compared with Linn's risible willingness to help.

Another one is SME. They are now owned by an Indian businessman called Ajay Shirke, who is fortunately also a hifi nut. He has breathed new life into SME (who were on the point of financial collapse), and who are still totally awesome in ancient parts supply and customer care.

Craig

[Radio Wrangler]

Quad, SME, these are companies who treasure their historic products and think that their quality and continued function reflects well on them.

This throws into sharp relief those firms who cut support to force equipment replacement and basically treat their customers as involuntary blood donors. What is surprising is that some of these firms customers exhibit brand loyalty!

To some extent it's the secret sauce mentality. 'Our products contain magic. You can't see it, but it's there and if anyone else services that box, it'll be gone.' If they let circuit diagrams out, those who could read them would see through the flim-flam. Quad, on the other hand, published diagrams and explanations in Wireless World. They were proud of what they'd done because they'd done something real. People who live in imaginary worlds daren't do that.

Still, mustn't grumble. Things are better than in the past. Previous attitudes to witchcraft had a terrible death-toll. Now it only takes victim's money, not their lives.

David

[Ted Kendall]

Originally Posted by Craig Sawyers

Another one is SME...who are still totally awesome in ancient parts supply and customer care.

Generally agree, but, as I have found out, the Series III is not supported at all, nor is Series II prior to the Improved. I know the III was regarded as a failure at the time (a mere 100,000 plus units), but it's actually a good piece of kit, and just about the only arm to allow an ADC 25 to work to its full potential.

Still light years better than Linn's service, though...

[thermionic]

From my experience.... a big thumbs up for Sugden Audio, who were more than happy to provide me with helpful information regarding one of their 1970’s integrated amplifiers.


And a big ‘snotty’ thumbs down for Mission, who refused to supply information or a circuit diagram for one of their early 90’s tuners which I was trying to repair for a friend. Needless to say, the respective praise and distain was passed onto both of these end users!