advertisement


Linn Spark Recap.... Is it worth it?

car973

Member
As a complete novice I would welcome some advice and opinion about recapping a Linn Spark. The last time I did anything like this was replacing a tiny cap inside my Numerik which is still going well (and all thanks to PFM stalwarts).

Until this week my LK280s+Sparks had been untouched since 1989 so they were/are well overdue for some attention. Bought a job lot of three 10 years ago for small money and one of them hasn't been used for ages so it was the ideal start point. I did of course check that it was fully discharged!

I have now recapped one of the LK280 amps replacing all electrolytics with like-for-like (apart from size - simply not available). To my surprise and relief, it still works! In fact it's a huge improvement in overall sound quality and encouraged by this I am now looking at the Spark. Has anyone done one of these?

Inside there are 8 main caps (ALP22A103DD063) each with five pins fitted to a single board. Replacement availability? None that I can find. The datasheet shows only one positive and one negative pin so if I'm not mistaken the other three pins are purely for support. I could therefore modify the board for a two pin fixing and use glue or strong double-sided tape to support. Does that work?

First ask is, should I bother? Will replacing the caps make an audible difference? The datasheet suggests these are 26,000 hour caps and this particular Spark hasn't been used for ten years. The others have been on 24/7 although the electric bill suggests it's no longer a good idea!

Recapping the amp's caps has given a much fuller sound particularly the bass and top end. It has gone from 'pipe n slippers' to a breath of fresh air and I now realise I should have done this a long time ago.

Also, are there any safely or risk of damage considerations for Spark replacements to be an imperative? The caps all look to be in good condition.

If anyone has recapped a Spark, is it worth it?
 
If I may offer an opinion divorced from any acquaintance with this particular PSU: it's worth doing sonically, for much the same reason as it was worth doing to the LK280.

It's also worth doing for reasons in line with a particularly praiseworthy facet of this forum's zeitgeist: Repair > Recycle > Refuse. Keeping older hifi going is simply a very worthwhile thing to do, and the performance of well-fettled kit is often nowise inferior to modern equivalents.

It's always possible (and an argument against) that the existing caps are in good health, but by the time you've pulled them and checked them, you may as well replace them with new ones with a much longer life expectancy ahead of them.

Besides voltage and capacitance, make a note of the maximum height and diameter of component that will fit on the PCB and inside the case, and make use of Farnell's well-implemented parametric parts finder to set some minima: working voltage should be the same or higher, capacitance the same or higher (up to at most +50% of existing value), and rated temperature the same or higher. Get the best you can. It's always possible that the ideal cap is out-of-stock, if so it may be available elsewhere - Mouser and Digikey seem to manage stockholding better IME.

Assuming you can't find caps with 5-point connection, rather than gluing the new caps in place - which can be a nightmare when it next needs servicing - I'd place a bit of padding between each group of caps and carefully cable-tie the group together, not too tight, obviously, just enough to support them.

While the PCB is belly-up, I'd also reflash all the solder joints: dry joints are a marvellous source of intermittent faults.
 
The caps on the 280 boards have a hard life as they are subject to heat and dry out.

The Spark caps have a much easier time and id be surprised if there was much mileage in changing them.

I tested some of the ones in mine and they were still in spec.

Try one and report back.
 
Thank you both for great advice. I hadn't thought of testing them so that should be my next step. I don't actually know how to test them but I can find out and if it means removing one or two of the caps for that purpose, it's not much to do. Supposedly they have a +/-20% tolerance so if it's +/- another 5-10% of whatever I'm measuring, then maybe it's not going to make a material difference to my ears. Most of the heat is definitely generated in the amp box so they could still be ok (he hopes!)

If replacement is unavoidable, my search window has now been widened appreciably with the knowledge that there is upward latitude to increase capacitance and voltage, size permitting (all good to know). I haven't been able to track down any matching same spec 5pin types using Mouser, Digikey, Farnell, Rs etc. but I was looking for exact specs. Actually, they are available from Distrelec in Switzerland but min. order is 1008 pieces at EUR13.05+vat!

If I can't find a suitable multipin, I'll avoid gluing 2pin caps to the board. A future owner will be grateful!

Should I re-open the amp box and 'reflash' all of the component joints on each board? There's not a lot to them so it's do-able but they are working rather well; or are you suggesting there could still be less than ideal connections? Are connections affected by heat and become brittle?

Definitely learnt something today, thank you, and will report on progress.
 
With rudimentary knowledge how to discharge and test a capacitor safely and a smart new multi-meter, I was eager to test the caps pulled from the LK280. First up were the small 100uF caps – 12 of them. Most of them were out spec with capacitance less than 50%. Three appear to be completely dead. On the other hand, the four 10,000uF caps were/are still good reading a whisker under 10k. Whoops! That was the bulk of the cost. With hindsight, wiser choices could have been made. I went wrong by not doing any testing. On the plus side, the LK280 now sounds more like an expensive hi-fi upgrade than a £60 parts bill.

At least the new caps will be good for years to come and I may not need to do the large caps in the other LK280s just yet. I’m hoping to sell them anyway so a partial recap may be all that’s needed (if also confirmed by a listening test).

As to the Sparks, I’ll remove a couple of caps from one of them today for testing and fingers crossed all should be good. Watch this space.
 
I went wrong by not doing any testing.

You really didn't. The cost of replacement caps is silly compared to what they once were, but compared to the cost of (say) replacing the unit you've refurbished, or - worst case scenario - a fault in that unit causing catastrophic damage upstream in amps/speakers, the cost is utterly trivial.

Also, those caps may well measure OK on a cheap, basic tester, but what do you really know about their performance in-circuit...? And you simply cannot know for how much longer they might stay good...

The mindset of the professionals commenting on threads like this is always from the POV of people who need to minimise their costs, and who are charging-out their time to other people - and that's how it should be - but for relatively inexperienced, untrained amateurs, the time and ballache involved, and the potential to go wrong and break something, are all considerations that really matter.

So, speaking for myself, if I pull a lifetime-limited component from a board, it doesn't matter how it measures, unless it's Unobtainium it's going in the bin and being replaced with new, preferably better. It's actually daft not to do this IMO.
 
Earl, thank you for your reassurance. It's not a lot of dosh as you say. I don't know if Kemet has a good reputation but they took over BHC, hence like-for-like replacement. True, I am that untrained amateur and you definitely wouldn't want to use my multi-meter in a mission critical setting. That said, it has a specific capacitance setting and for the two caps just pulled from the Spark, I get readings of 9.52K and 9.51K which seems ok. The farad testing range is auto-setting so it gives me a bit of confidence that it is measuring approximately correctly if not precisely. I take the point that age is not on their side but I haven't yet found a replacement that wouldn't involve modifying the board. So that is probably what I will have to do at some stage (to the one I'm keeping!).I have taken the opportunity to reflash the board. In conclusion, I really ought to replace them as a precautionary measure but there's no rush because they are not under much stress. Many thanks
 
All makes sense.

I wouldn't personally concern myself with matching maker and series with the smoothing caps - so long as the replacements have at least the same key values and physically fit on the PCB / under the lid, it's all good. I've no real experience with Kemet caps, the range of BHC-alikes is much smaller than BHC once made, and SFAICS lacks all of the more interesting options now.
 
In closing, I started on the next LK280, removing all of the caps. A previous owner had thoughtfully superglued the large caps into the plastic retaining clamps, not just a dab but liberally right around such that 2/3rds of the surface was properly stuck. Managed to slice through the outer jacket and get them out that way leaving the jacket stuck to the clamp. Not the end of the world but unhelpful. Must have been a sloppy amateur ;)
 
I used four LK100 for a few years and all was fine until a cap sarted to make noises so a friend who does this sort of work in Bulgaria and was here for a few weeks visiting his parents said he would take a look and we decided to replace all the caps in all the amps and ye gods the improvements were huge. After a couple of years one of the amps developed a problem ( not a cap this time) and I had to transfer all the caps to another board. A faff but not difficult to do. At this point I realised that perhaps I had come to the end with old amps and quickly replaced them with Magik 4200 and any logevity issues are now gone and the added bonus of a better performance. Yes it is worth replacing old caps if they do not measure well as long as you do so with like for like measurements.
 
I've got a Spark that I don't use. I bought it from ebay rather cheaply a few years ago. It's incredibly heavy! So I'm reading this thread with much interest... Thanks for posting.

I recapped all my LK280s with Elna and Mundorfs caps. I just did some research on hifi collective and since it was the last time these would be recapped, I wasn't worried about costs. It all worked out fine in the end.

I got a bit stuck with my Dirak and Active Crossover... But that's a different and distressing story, that eventually worked out fine, but it was a close call due my own stupidity ... Its all working nicely now so I'm rather pleased with myself.

One of these days I might open up the Spark and have some fun...I love messing around with electronics... Shame I've got no talent for it, I'm rather cack handed.
 
I love these threads because I’m a big fan of fixing and re-using and if it can improve the sound quality too, even better.
My preparation for retirement (hopefully not more than 2 years off) is to find a basics electronics course (in person not on line as I want to watch a skilled professional use a soldering iron). I would love to do what @car973 has done - without blowing myself up or the unit :D:eek:
 
I finished doing the last LK280 but this time I only replaced the small 100uF caps to compare it with the other two I had recapped earlier. Well, it was difficult to tell any difference and after switching the units a few times I couldn’t draw any obvious differences. To my ears they all sound the same. So I’ll do the big caps on the last one and then it’ll be good for a few more years. What I did notice while doing the small caps was that the circuit is printed on both sides of the board so it was really important to make sure the upper point of contact is good.

Really happy I’ve done this and I’m encouraged now to do some pre-emptive small cap replacement on the other Linn components. I will do one of the Sparks (out of curiosity!) but it will require a little more effort as the original pattern 5 pin caps don’t seem to be made any more. I tracked down some matching F&T caps on eBay but as these must be NOS and of dubious provenance, I passed - better to use good quality new caps even though it will require a bit of fiddling with the board.

BTW, my wife thinks I’ve lost the plot doing this but as all of you already know, it is very enjoyable and satisfying (when it works).
 


advertisement


Back
Top