Translate

Wednesday 23 January 2019

Right to repair.

This is a partly political broadcast on behalf of Doz...

If you're in Europe (as I am until the end of March). This is important.

Stand up for your right to repair.

https://openrepair.org/news/towards-the-right-to-repair-in-europe/

If you're in that odd bit of the planet, where earth is ground, and the mains wobbles up and down a bit faster, but not as far as here, valves are tubes, and tubes are CRT's (I'm talking about the USA).

https://repair.org/stand-up/

We can't keep chucking this stuff away! And, repairing brings an amazing feeling of satisfaction (I know, I've been fixing things for 40 years)

Don't let this stall.

Sign the petition, and make the world a better place.


Picture stolen from Ifixit.com (who make an excellent screwdriver kit) , sorry.

***STOP PRESS***

This turned up on IFL science this morning too...

https://www.iflscience.com/chemistry/chemicals-elements-that-make-our-smartphones-are-running-out-new-endangered-list-reveals-/?fbclid=IwAR0-RaNu40VvVLq14TfMor4xJ_CK5iQOY_N1RQCG_Lkzff5I2sjniAPuQTw




Tuesday 22 January 2019

Beocord 2000 De Luxe - no playback.

Pyers put out a distress call on the internet ..

"Does anyone have an old reel to reel which does 3-3/4 ips 1/4 track? I have a load of old tapes to digitise."

A mutual friend came back with this ..


It's a Bang and Olufsen Becord 2000 De Luxe (no less)

"Needs a service. Hamstall Ridware: But it’s bloody tidy. Dust cover and all."










It found it's way here...

Initial testing showed it only worked on one speed (there are three, 1 7/8 IPS, 3 3/4 IPS and 7 1/2 IPS), and even then only when it was tilted on it's side. It was also running slow when it was running...








Two of the world's longest screws, and the bottom comes off.

Doesn't look like it'll help us much, as the fault is obviously mechanical.















Although I can get a glimpse of what's to come!

Off with the top plate...
















It's immediately apparent that the capstan is not being driven properly. Belts look in reasonable nick though!










The spring that holds the idler against the capstan flywheel, and the motor drive hub has broken, there's precious little friction to drive the flywheel.

You'll need to remove both the rewind and capstan drive idlers to gain enough access.







The spring has broken at the hook end, so I shorten it as little as possible and form a new hook.

This is the bias oscillator. We'll need to remove this to get the spring back on. There are two screws each side of the heatsink...






And the module is carefully unplugged, and set to one side.











Then, after some not inconsiderable cussing, the spring is attached at the capstan end, fed through and attached at this little bar underneath the bias oscillator.









The capstan idler is cleaned, and slid back into place, and secured with it's circlip (not forgetting the felt pad).

It's the same procedure for the rewind idler.

The bias oscillator is replaced.








Just need to clean up that grubby tape path, and put it back together!















And it's up and running :)

Saturday 19 January 2019

Linn Majik - humming fault

"Ho, ho, ho , It's magic you know , Never believe it's not so"

Except for this one. Magic it isn't...

It's another of Colin's Job lot of Linn stuff.

A Majik integrated amplifier.


Connecting up and switching on it works, but is humming badly from the right hand channel. 

Probably more caps I'm thinking ... let's get it in bits and have a look.

Same MO as the other Linn products in this range, four screws holding the top cover on, are found underneath. Slide off the top cover and put it to one side...
We'll need to remove the remote / balance input board from the rear panel. Two screws, and a couple of plastic support, and it lifts out. Talking of balanced audio... why oh why oh why an RJ11 socket?? XLR's not good enough? 

OK, having got that out of the way, we're going to need to remove the main PCB. Flip the unit over, and remove the 5 screws holding the heatsink to the base plate. 


Remove the screws holding the phono sockets to the back panel.

There are four screws to remove to remove the PCB from the base plate. Two by the capacitor bank, and two on the other side, marked "tool hole" and "sneaky"!

The front panel needs to come off so we can get the headphone jack to clear.

Two screws at the top, left and right...

and two underneath (the silver ones).

Disconnect the ribbon and the two pin connector to the power LED, and remove the front panel.

One the links for the power / amp amp are removed from the back panel, the main PCB easily lifts out. A cursory visual check shows nothing obvious. The main electrolytics are still in fine shape, as they're still holding charge.
Right, we're going to need to get the thing back together to begin testing... so the transformer is plugged back in, and the front panel. It's all a bit dicey. Be careful not to allow the board to short out on any of the case! There's no need to plug the balanced board back in, it'll run happily without.

Without the heatsink bolted to the base plate, it does get quite warm quite quickly, so keep running times whilst testing to a few minutes, then let it cool down.
Powering on shows all the LT rails (+/- 26V, +/-15V and +5V) are all present. The power amp is now isolated from the pre-amp, and the hum remains. So our right hand power amp is at fault.

Comparing differences, both static resistance measurements, and using the scope between left and right amplifiers under identical conditions, I can't see any difference between the amplifiers except for the output of the right amp.

I short the input of the right hand amp to ground, to see if the hum is getting in through the input somehow. It makes no difference.

It must be a fault within the output IC's (TDA1514A) of the right hand amplifier (there are two IC's per channel, in parallel). The problem is the TDA1514A is obsolete. I won't risk the eBay IC's... they're bound to be fakes.

Damn.

One thing I have noticed is there's a bit of 100Hz ripple on the -26V rail. But both amps share the same +/-26V supply (It's actually +26/-25)... so surely it can't be that, can it? I change C602 and C603 which are the local decoupling capacitors to the power amp IC's. There's slight "change" to the hum, but it's very much still present. Encouraged by this change I look at the supply.

Now the -26V supply is regulated by a 2SC3519 pass transistor which is controlled by an LM337. The adjust pin of the regulator is decoupled with a 100uF capacitor. The input is decoupled by another 100uF capacitor (C308 & C310). These two caps are right up against the tab of the regulator. They may have well been cooked by the heat. Decoupling the input to the pass transistor is yet another 100uF capacitor, C320. I change the lot. 

Shush! Can you hear it? Me neither, the hum is gone! 

So why on earth was the left hand amp not affected?? One of life's little mysteries..

Now the +26V rail has a similar set up, but uses a 2SA1386 pass transistor and an LM317 regulator. The components are mounted in a similar manner. I change that lot too, so we should have a reliable amp.

There , all playing nicely again. Another saved from landfill...




*STOP PRESS*

Steve's been in touch. He writes "The board is a Knekt receiver board and the RJ45's are for balanced audio & 2 way control from main room sources, the other socket is for the RCU (Room Control Unit) a wall pad to control the amp and local or remote sources."

Now the RJ45 makes sense!

Thanks Steve.





Thursday 17 January 2019

Linn Kairn Err 2 "Don't Panic" !

Turns out Colin's got a few items of Linn gear ...

"CD works great, but the pre-amp keeps displaying an error.

Can you have a look?"

Yeah ... why not...



Connecting up power gives this familiar error message... Something's wrong with the memory checksum.









Undo the four screws underneath and slide off the cover....












Oh gawd ... Thankfully this particular flavor of "Brilliant" is quite reliable...











We need to remove the front panel. Two M3 sliver screws near the front on the bottom, and two countersunk screws, one on each side. Remove the ribbon connector.










Remove the screws holding the shield in place.











Disconnect the next ribbon cable and put the shield to one side.











... and there's the issue. The Varta mempac battery. Green and fluffy! A new one is duly fitted. The unit will still show an error for the first time it's switched on again. Once the memory is written to, all will be well :)

If there's any sign of electrolyte leaking on to the board, this must be thoroughly cleaned up with IPA.





The new battery arrives .... I have to say they've upgraded the capacity a bit!!!

110Ah .... I think not !










Ah ... 150mAh ... more like it!













And after powering up twice, all is well...












Plenty of in's and out's!





Linn Klout - one channel down.

More Linn !

This time it's a Klout (where do they think these names up?!)


As you can see from the photo, the left channel is just showing a red LED, and has no output.

First off remove the top...

We're going to need to remove the left hand amplifier board to inspect and repair...









... which takes a lot of disassembly!

Disconnect the transformer, and the remote switching board. Unscrew the single screw securing remote input socket from the baseplate (you can leave it attached to the rear panel). You can now remove the baseplate, remote switching board and transformer.

Unscrew the input socket for the board we want to remove from the rear panel, and remove the two long cap screws that secure the output sockets. The cap screws are located in between the heatsink fins. There are 6 more cap screws securing the output transistors heatsink block, and the board itself to the heatsink. Remove these. You can now wiggle the offending board from the chassis.


Cursory checks on the board show all 4 output transistors to have failed leaky. A sanity check with the good channel confirms this.

There are two pairs in parallel. Two 2SC3519's and two 2SC1386's. Thankfully these are still available at the time of writing, so are duly ordered up. Driver transistors Q20, Q21, Q22, Q30, Q31, Q27, Q26 and Q25 check OK

The ESR of the electrolytics are checked. There are six 100uF 63V caps, two 220uF 16V on each board. They read OK'ish, but not great, so are replaced. The main smoothing caps all read sub 0.01 Ohm, and are not replaced.


The small sub-board has two 68uF 16v caps on the top. They read almost open circuit! Replaced.

Once this channel is repaired, I'll re-cap the other as it won't be far behind failing!









The transistors are held to a heat spreader by these two nuts. Removal is fairly straight forward. Clean all the old, dried up heatsink compound form the spreader with some IPA.









Heatsink compound is cleaned off the heatsink, and a smear of new compound applied.











And the board squeezed back in place!












I reassemble the amp, just to test the freshly repaired channel before re-capping the good one.

It's good news.









All boxed up, tested and good for another 20 years :)