The Workbench

Re: Workshop Beolab 2500 Audiophil

stampfki — Sun, 19 Dec 2010 14:09:46 GMT

Hi,

Today I have started the overhaul :-). Do I have to consider any PIN numbers on the OPA2134UA? I cannot find any sign on the chip showing a certain direction.

THX

Re: Workshop Beolab 2500 Audiophil

Die_Bogener — Thu, 16 Dec 2010 13:34:00 GMT

BM7000 and BC9500 are very similar from the design. C204 can be replaced by a wire bridge, the others not.

Check and measure the dc offset across these caps. If it is less than 10mV, try to replace them by wire.

Also check, these chips are working internaly with some volts dc offset. So it is important, which cap you replace and which not, these chips are hard to get.

The diodes are not necessary, because i removed the in/out plug. So there is no possibility to get overvoltage into the amp.

Re: Workshop Beolab 2500 Audiophil

MartinM — Thu, 16 Dec 2010 03:02:06 GMT

The BC9500 preamp board circuit appears quite similar to that of the BM7000. The workshop article you had on modifying the preamp got me thinking about doing something similar to the BC9500 board. Looking at the diagram above, with the OPAmp replaced with OPA2134, could C204, 205 and 211 be replaced with a wire link? I see also that in your BM7000 mod you removed Diodes D200, 201, 202 and 203 but not quite sure reason for this.

Re: Workshop Beolab 2500 Audiophil

Die_Bogener — Tue, 14 Dec 2010 12:37:32 GMT

C201 can be replaced by a 10uF Wima type, no problem.

Just look where it comes from... the signal comes from C200 which is a 10uF cap.

It makes no sense, when the first cap is a 10uf to use a bigger second 22uf, the signal does not become stronger or different.

So the same size is good enough :)

Re: Workshop Beolab 2500 Audiophil

MartinM — Tue, 14 Dec 2010 02:39:11 GMT

@ Die_Bogener: Excellent workshop article! Thanks for sharing the information.

I’m hoping you can help me. I have recently acquired a Beocenter 9500 which I’m in the process of upgrading the opamps and caps. However, looking at the circuit diagram I see that one of the caps in the audio path (C201) between the output of ic1 on the preamp board and the STK chip amp is 22uF. As you mention this value is hard to find in WIMA. What would your thoughts be on this cap being replaced with a 10uF WIMA? Here are the relevant parts of the circuit I’m referring to:

Re: Workshop Beolab 2500 Audiophil

Die_Bogener — Wed, 01 Dec 2010 09:34:41 GMT

The picture is repaired

Re: Workshop Beolab 2500 Audiophil

stampfki — Wed, 01 Dec 2010 04:09:23 GMT

Hi,

Please, could you update or reload the missing picture?

Thanx

Re: Workshop Beolab 2500 Audiophil

Mohawk — Tue, 06 Oct 2009 11:41:24 GMT

Hi Stampfki

I changed the 2 caps in the crossover to new polyester types instead of the old bipolar electrolytic caps.

In the poweramp and power supply all old electrolytes were replaced by new ones (105 degrees).The input cap of the poweramp was replaced by a polyester type instead of the electrolyte.

The 2 OP amps4558C (IC201, 202) were replaced with OPA2134UA. Those OP amps are surface mounted so be careful when desoldering and soldering them.

/Fredrik

Re: Workshop Beolab 2500 Audiophil

stampfki — Fri, 02 Oct 2009 05:50:40 GMT

Hi Mohawk,

I'm would like to overhaul my Beolabs 5000 as well. Could you provide some more details what you did replace exactly? My Labs still running as they were built in 1988.

Thanks in advance!

Re: Workshop Beolab 2500 Audiophil

Mohawk — Sun, 10 May 2009 13:43:12 GMT

Hi again

Did a complete capacitor overhaul in my beolab 3000 speakers and decided to try your modifications in them. Changed the OP amps in the signal path to OPA 2134 but I didnt remove any "unnecessary" caps, just changed the input cap to the poweramp to a polyester type. This actually resulted in a much clearer sound than before. I didnt think it would do that difference but it did. It's like if one removed a very thick cloth in front of the speaker. I compared them with my unmodified BL5000s and there was definately a difference in clarity. Im really happy of the result and Im going to do the same modification to the BL5000 also.

/Fredrik

Re: Workshop Beolab 2500 Audiophil

Mohawk — Wed, 01 Apr 2009 14:10:14 GMT

Great story!

Im doing an overhaul on a pair of beolab 3000/5000 amplifiers right now and I will try some of your modifications where it's possible. Would be fun to hear the difference when Im done.

/Fredrik

Re: Workshop Beolab 2500 Audiophil

bayerische — Mon, 29 Dec 2008 17:30:17 GMT

A fun and interesting story indeed!

Thanks.

Re: Workshop Beolab 2500 Audiophil

geearr — Fri, 26 Dec 2008 20:11:57 GMT

A very interesting story - seems to show just how much "black art" there is in electronic circuit design. I am looking forward to hearing the rest of your narrative - it is very informative.

Thanks

Geoff

Workshop Beolab 2500 Audiophil

Die_Bogener — Fri, 26 Dec 2008 11:17:06 GMT

Hello B&O friends,

merry chrismas and a happy new year!

Maybe some of us dont really know what to do in our holidays... so let's modify some of the Beolab 2500. ;)

I just got a pair of BL2500 for my office and they will be part of an Beosystem 2500. The BS2500 has already got all audiophil modifications and is playing perfect with a pair of modified BL4000. The sound is superiour... but when i connected the standard BL2500, hmmm, something was missing. Bass and treble were so "normal", small, sounding just like as if the loudspeakers were still wrapped in bubble foil. I had them better in memory... well, this was the normal sound, quite normal series.

The Beolab 2500 is very similar to the BL4000, the same bass loudspeaker is used, the tweeter has similar data, the STK amplifier chip is almost the same... so why do they sound so different? OK, there is almost nothing that couldn't be made better. :)

Here we are. Just open 7 bolts. This is a look inside the BL2500. Left side the transformer, power supply and the STK chip. And some ELNA electrolyt capacitors. Just a little work to do...

Step 1: Replacing Caps:

http://img360.imageshack.us/my.php?image=pc250049yu9.jpg

A look into the manual shows that only 4 caps need to be replaced by a better cap... WIMA MKS-2-5

What you need for the first step:

C5, C29 = 2,2 uF 50V (edit: not necessary if you continue with all modifications, repace by a wire bridge)

C7 = 10uF 16V

C27 = 22uF 16V

C5 and C29 were replaced by a 2,2uF Wima MKS-2-5, also C7 by a 10uF type. C27 is more difficult, there are no 22uF type available in this size... so i replaced it by a 10uF type, it works the same. Just if somebody has a problem with a smaller cap... the driver is IC4 and it uses C29=2.2uF as a driver. So it makes no sense to use a 22uf type. Later some more about these caps... some of them are not necessary ;)

So, after the cap mod, the board lookks like this:

http://img396.imageshack.us/my.php?image=pc250054xw2.jpg

OK, first sound test: Yes, there is something new to the BL2500 now... they sound better, little bit more clear, the bubble wrap foil is thinner. But still no comparison to other modified Beolabs... ok, it will never be a Penta or BL8000 or BL1, but they should be similar to a BL4000 at least. But you will hear some improvement.

If you are not experienced with hardware and smd parts, you can stop here and reassemble the BL2500. You can be proud of some improvements to the BL2500. Or continue to step 3... also an improvement :)

STEP 2: Replacing the Opamps by audiophil types:

B&O uses for the opamps the standard NE4558 and the LM833. OK, not so bad, but there can be done more...

Replace the 3 opamps by an OPA 2134 UA. This chip works already in some dozen other Beolabs and amplifiers, they sound very good and are not too expensive... around 4 Euro each.

So you need 3 OPA chips for replacement: Replace IC1, IC3 and IC4.

OK, here we are, the opamps are exchanged:

http://img360.imageshack.us/my.php?image=pc260058aep6.jpg

Sound test again: Hmmm, somewhat cleaner, a little bit more bass, some more treble... better. But still something is missing, there are some details missing in the background of the music... i always compare it to my modified Pentas, i like The Eagles Hotel California ... :D

Very nice already... much better than the standard BL2500, but still something missing.

Step 3: Parts you really dont need:

OK, here some parts you really dont need: C60, C8, C28. These small 2.2nF caps are installed to protect the amplifier against noise, oszillations, to fast response... let's say: to become the amp more stable. But these small caps kill some details of the music, they filter small details out of the music. I dont like them, they are the reason for the bubble foil effect, at least a little bit of it. I have removed these caps in all my equipment, every beomaster and every beolab in my house and all my friends and relatives equipment ... these small caps were removed and i had no problem at all.

So, this is something to believe in ;)

Remove the three caps marked red.

http://img526.imageshack.us/img526/8307/pc260058b.jpg

Just remove them and make the sound test:

Yes, there are more details now in the music. The bubble foil effect is gone, it's clear, a punchy bass, crystal clear. You can hear things that were not there before.

Step 4: Some more parts you really dont need:

As mentioned before, there are some caps not really necessary: C7 and C27, even C5 and maybe C29 are not necessary at all. B&O used the caps to remove an offset voltage of the opamps to the power amp. The standard opamp LM4558 and the LM833 are really not good in the data book when you look at the offset voltages. The LM4558 has typical almost 2 - 6 mV, the Burr Brown typical less than 0.5mV, that's really a difference. With these Burr Brown 2134 UA chips it is possible, to build a DC coupled power amp... almost. I have done this mod already on the BM7000 and the BS2500 several times, it works with opamps like these. This mod is already shown in the Workshop BM7000 and it works.

The advantage is: less but better parts. Less parts means also: there is no information lost in the signal.

OK, some measurements: The DC offset measured over C5 and C29 is less than 2 mV. That means that C5 and C29 are not necessary at all. Nice :)

The DC offset over C7 and C27 is around 160mV. That is too much, C7 and C27 have to stay.

Hmmm, that means, remove C5 and C29 and use a small wire bridge instead... that's the cheapest tuning ever :)

The board looks now like that:

http://img443.imageshack.us/my.php?image=pc270003rl9.jpg

Finished!

Sound test 4: Crystal clear, a little bit stronger than before... and pretty close to the sound of the bigger BL4000. Wow, sounds really nice! :)

If somebody modifies his BL2500 now and has similar experiences, please post it here! I'm very interested wether this sound tuning can be validated by other Beoworld members too.

Have fun!

Martin Schmitt

Bogen