Howard_Johnstone wrote:

 

I which to object to the wrongful replacement of capacitors and transistors in old Beomasters as a point of being original without any regard to factual circumstance. Being that output capacitors need to be close to 20 mfd for full bandwidth, 20-20k (D. Self et al)This was well known at the time and led to audio downrating.

This of course relates primarily to AC-coupled receivers with the 3 (single supply) capacitors for PSU and output.

The original values was chosen because of size, cost and availability with the technology available at the time, in full knowledge that the practical values where so low as to seriously affect performance, especially at full load and over time. It is therefore wrong to replace/substitute capacitors with the same values, knowing that they are wrong in the first place.

Today, we can come close to 20 mfd at 50V, 63V and even 80V (BM4000) PSU and output capacitors should have the same voltage rating in the first place, especially running at full load (max voltage swing).

Never ever use NOS caps and trans, these will eventually fail and cause further problems. Don´t replace with higher voltages than needed, most voltages are actually lower than you think, i. e. BM3000 requires only 63VDC. (43-45VAC x 1.414 = 63VDC) Measure it!

Maximum capacitance within the size available will close the gap to 20 mfd and suddenly the classical problems with AC-coupled amps are no longer!

Always replace the rectifiers, broken or not, with even simple available replacements such as GBJA1510 or TS25P05G that can be bolted in place. The original rectifier, PSU and tuner, have degraded and will fail when any modern low-impedance capacitor, whatever value, are used. The old transistors 2N3054 and 2N3055 had a very wide spec. allowance you will never know what performance you will get from them, and the bond wires will have corroded (purple plague). They will ev. fail at load. Again, today modern TO220 (TO66 replacement) and TO218 (TO3) will easily outperform these old units with much greater Ft etc.

Also there where cost simplifications made along the production series not at all addressed in the service manuals, such as the drop of cooling blocks in the later series of BM1400M&K and carried forward into BM1600. The lower output of the initial BM3000 (30W) and 108MHz versions,raised to 40W, not addressed in the service manual, as well as revisions into the later BM3000-2, again no specific information produced.This leads to measured values not corresponding to the schematic diagram when circuits need to be recalculated and checked for coherence.

Please acknowledge this and give the right advice for reconditioning.

Yours sincerely, Howard Johnstone

 

I acknowlegde your right to state your opinion and to present us with your advice.

- And what a post that brought us!

"Don´t replace with higher voltages than needed" (Capacitors).
I agree with that.
I also agree on not fitting NOS capacitors (electrolytics).
But I'm afraid, after more than 40 years of collecting, restoring and servicing B&O audio units and electronics in general, I don't agree with most of the rest - or at least I find it completely unnecesary.

I would prefer to fit NOS output stage (and other "costlier") transistors any day rather than risk buying counterfeits. Been there, - even from reputable suppliers, tried it many times, got the T-shirt.
I gladly snap up unused old stock whenever possible, and I also gladly reuse good original transistors. Never in all the decades did I have any problems with that.

Certain rectifiers were bad from factory.
In Beomaster 3000 (and 3000-2 as the updated version is called - make sure you have the right servicemanual, there are many) this goes for the Siemens rectifier on the tuner board, which I like to replace on sight with a W06 or similar type (it's not very critical).
In the first Beomaster 1900s the rectifier on the main board did not live up to the producers specs and should also be replaced on sight (a B&O tech note actually suggests this).
The rest of the recitifiers I leave well alone (unless faulty of course) and they'll be fine.

- But they may die if you up the capacitance of filter capacitors and output series capacitors!
This is not because the rectifiers are bad, it's because they are put under loads they weren't designed for.

"rectifiers ... will fail when any modern low-impedance capacitor, whatever value, are used".
You would never use low-ESR capacitors in circuits like the Beomaster 3000 power supply (or DC-coupled output stages for that matter).
You would choose something similar to the originals.
Blanket-fitting low-ESR capacitors to improve functionality is a misunderstanding. On the contrary they can cause problems that
weren't there before, like blown rectifiers and self-oscillating amplifiers.

If you fit low-ESR filter capacitors in Beomaster 3000, the rectifier WILL eventually blow.
Replace the rectifier with a more powerful type and the transformer will eventually die.
(Think Beogram 4000 mains transformer, I've seen many dead ones after owners "improved" the filter capacitors).

Not all components are better, just because they are modern.

Larger output series capacitors will give a better damping factor. That's a fact - and that's about it.
Upwards from the original values I doubt anyone will notice any difference, though.
Better bandwidth? - In theory, well yeah... but in a Beomaster 3000 hmm.. not significantly, no.
The original values are adequate for the unit in question and for the frequency range specs given.
Other components and circuits will set a lower limit, anyways.

Please keep in mind, that output series capacitors are charged (and discharged) through the attached speaker at power up/down.
(This is what gives the characteristic thump sound when switching a Beomaster 3000/4000 on/off).
Upping the capacitance of the output series capacitors drastically as you suggest, would put a much higher load on the voicecoil(s) and filters in the speakers (much louder thump sound), and together with larger filter capacitors the load on the rectifier and transformer will also increase, because the rush-in current would then be considerably higher.

That a modern transistor has a higher frequency range (say 30MHZ instead of 10MHz) is of no importance whatsoever (slewrate etc. included) so long as we handle nothing higher than 30KHz at the extreme max.

"Also there where cost simplifications made along the production series not at all addressed in the service manuals"
Why should they be? The servicemanual will tell you how the given unit is made - not what was left out or considered unnecessary in production.
If these details are important to you, make sure you have the right manual for the given unit.
A lot of tech notes, service notes and BoMark pages were released covering production changes.
Not everything is covered, but the major things are.

The Beomaster 1400M/K (and 1600M/K) models have very low-powered output stages, they don't need a lot of cooling fins. They are absolutely fine as produced in this regard.

We are free to disagree here, and I couldn't help noticing, that we've already been there. 😂

Thank you very much for your answer.

It helps a lot to be constructive rather than just negative, as this do not bring the discussion further.

Transistors:

At the time of design and production these units where available and at low cost.

They had insufficient encapsulation that cause substrate corrosion and gold bond

wires that will show up as the ”purple plague”. All this degrades NOS transistors

and limits their use to low power only. At full load they will eventually blow.

This does not apply to many of the small signal transistors used, they where/are o.k

as long as they are not put under load, high temperature or damp/wet environment.

Today when old motorola (semicond.) is run by ON semiconductor, excellent products

can be provided even in the original format, TO66 and TO3 without any problem at all.

However, they are aftermarket products and will cost. Therefore old classics such as

2N3055 can be provided as MJE2N3055T (TO220) at very low cost as replacement.

It was wrong to select the use of 2N3054 in the first place, and they knew it.

I have used D44H11G (TO220) modern and of excellent performance at low cost

and will give you a much better SOA, equals peace of mind.

Today Mouser, Digikey and RS components provide safe and secure certified sourcing.

They will never ever knowingly provide counterfeit as they could be sued.

Buying counterfeit at Ebay is stupidity....

Rectifiers:

The inrush current values goes down, especially the small tuner bridge (BM3000 to 4000)

no matter what capacitors are in use or will be selected. If they fail the transformer will

be in perill, just as you have experienced. If the transformer is without defect (E-core)

it will hold to a load corresponding to about 70W RMS without overheating. (5 mins).

Only in the BM4000 do you come close to this, tested with low-impedance 18mfd/80V.

Original rectifiers, wrong fuse, defect transformer (humming) and the transformer will

be shorted/overheated no matter what load (short) or use it is subjected to.

Original rectifier (PSU) is an accident waiting to happen, why take the risk?

Capacitors:

Today, almost all electrolytic capacitors will be or be close to low ESR (impedance)

capacitors as the electrolyte keeps improving and the market demands lower impedance

as a matter of fact. Unless using NOS it is/will be unavoidable.

I have tested this on BM4000 at full load and if the amp is in order

there is no problem with the transformer or any problems with self-oscillations, which

is a sign of stability problems in the stage itself; again, old transistors, bad SOA.

Using 15 and 12mfd respectivly in BM3000, BC3500 and BM4000 I have not

experienced any thump or noise whatsoever. (18mfd in BM1400) Unless there is

a leakage (which is heavy with the old capacitors) there is no DC component that cause

thumps, noise or damage to the crossover/vcoil in the speaker. (D. Self et al audio design)

There where bad compromises when these where new; bad rectifiers, small capacitors.

But 2N3055 and MJ3001 was a correct choice  at the time, but the drive stage was not

and had to be revised. The power stage (PSU) of the BM3000 series was underused,

but the BM4000 was a simple improvement (darlingtons) which gave much more

output from the same size PSU.

All this refers to AC-coupled amplifiers, which are easy to handle and get in order.

Therefore I would challenge these to a much better performance than when new,

when a recap and other rework/repairs have to be done in any case to the chassis.

DC-coupled amplifiers are an entirely different matter, bode plots show a SOA that

limits everything to a very tight order, usually a large rebuild is then required if something

has to be changed (BM4400). That design was much more up to date and appropriate.

Today, of course, AC-coupled output stages are a thing of the past.

Howard_Johnstone wrote:

Thank you very much for your answer.

It helps a lot to be constructive rather than just negative, as this do not bring the discussion further.

 

Transistors:

At the time of design and production these units where available and at low cost.

They had insufficient encapsulation that cause substrate corrosion and gold bond

wires that will show up as the ”purple plague”. All this degrades NOS transistors

and limits their use to low power only. At full load they will eventually blow.

This does not apply to many of the small signal transistors used, they where/are o.k

as long as they are not put under load, high temperature or damp/wet environment.

 

Today when old motorola (semicond.) is run by ON semiconductor, excellent products

can be provided even in the original format, TO66 and TO3 without any problem at all.

However, they are aftermarket products and will cost. Therefore old classics such as

2N3055 can be provided as MJE2N3055T (TO220) at very low cost as replacement.

It was wrong to select the use of 2N3054 in the first place, and they knew it.

I have used D44H11G (TO220) modern and of excellent performance at low cost

and will give you a much better SOA, equals peace of mind.

 

Today Mouser, Digikey and RS components provide safe and secure certified sourcing.

They will never ever knowingly provide counterfeit as they could be sued.

Buying counterfeit at Ebay is stupidity....

 

Rectifiers:

 

The inrush current values goes down, especially the small tuner bridge (BM3000 to 4000)

no matter what capacitors are in use or will be selected. If they fail the transformer will

be in perill, just as you have experienced. If the transformer is without defect (E-core)

it will hold to a load corresponding to about 70W RMS without overheating. (5 mins).

Only in the BM4000 do you come close to this, tested with low-impedance 18mfd/80V.

Original rectifiers, wrong fuse, defect transformer (humming) and the transformer will

be shorted/overheated no matter what load (short) or use it is subjected to.

Original rectifier (PSU) is an accident waiting to happen, why take the risk?

 

Capacitors:

 

Today, almost all electrolytic capacitors will be or be close to low ESR (impedance)

capacitors as the electrolyte keeps improving and the market demands lower impedance

as a matter of fact. Unless using NOS it is/will be unavoidable.

I have tested this on BM4000 at full load and if the amp is in order

there is no problem with the transformer or any problems with self-oscillations, which

is a sign of stability problems in the stage itself; again, old transistors, bad SOA.

Using 15 and 12mfd respectivly in BM3000, BC3500 and BM4000 I have not

experienced any thump or noise whatsoever. (18mfd in BM1400) Unless there is

a leakage (which is heavy with the old capacitors) there is no DC component that cause

thumps, noise or damage to the crossover/vcoil in the speaker. (D. Self et al audio design)

 

There where bad compromises when these where new; bad rectifiers, small capacitors.

But 2N3055 and MJ3001 was a correct choice  at the time, but the drive stage was not

and had to be revised. The power stage (PSU) of the BM3000 series was underused,

but the BM4000 was a simple improvement (darlingtons) which gave much more

output from the same size PSU.

 

All this refers to AC-coupled amplifiers, which are easy to handle and get in order.

Therefore I would challenge these to a much better performance than when new,

when a recap and other rework/repairs have to be done in any case to the chassis.

DC-coupled amplifiers are an entirely different matter, bode plots show a SOA that

limits everything to a very tight order, usually a large rebuild is then required if something

has to be changed (BM4400). That design was much more up to date and appropriate.

Today, of course, AC-coupled output stages are a thing of the past.

 

Regarding transistors;
Gold-plated pins do not oxidize.
Same goes for tinned pins.
It's the bare (untinned) pins that oxidize (like some BC182/BC183 etc.).

No 2N3055 transistors (be that NOS or used originals from B&O) have untreated pins and none suffer from any plague (most are in fact gold-plated).
B&O used RCA and Motorola primarily. They are all excellent components that will stand the test of time.

The most often seen transistors with bare (untreated) leads are the small signal ones (BC182/BC183 and the occasional MPS5418 etc.).
Never the power transistors.

And no power transistors are pushed to their limits in any B&O.
F.e. the 2N3055 is used in Beolab 5000 (rated at 60 Watts per channel) and Beomaster 3000 (rated at 30-40 Watts per channel) - each of them using two 2N3055 transsitors per channel.
The producer rates one 2N3055 at 115 Watts alone...

Regarding counterfeit components;
"Today Mouser, Digikey and RS components provide safe and secure certified sourcing."
- I have returned counterfeit parts to all three. All handled the returns professionally, understanding and cooperative.

"Buying counterfeit at Ebay is stupidity...."
- I often advice against buying from Asia, but stupidity? Well..  I think I would rather put it down to lack of knowledge and expericence.

I'm sure no well-reknowned suppliers "knowingly" supply counterfeit components.
But the market is a maze and a jungle.
As mentioned above I have received counterfeit components from even VERY well-established and trusted suppliers, - and I'm sure they didn't know, until I told them.

"It was wrong to select the use of 2N3054 in the first place"
Where did you find 2N3054?
As far as I know B&O never used 2N3054.
I am not familiar with it anyways.

B&O used 2N5034, which proved to be an incredibly sturdy and relible output stage transistor (think Beomaster 1001/1200 and the earliest 1400).
I have replaced only very few of them, and it was typically when owners had accidentally shorted an output stage
They are VERY sturdy components.

Rectifiers:

"The inrush current values goes down, especially the small tuner bridge (BM3000 to 4000) no matter what capacitors are in use or will be selected. "

- What?

With a low-ESR (low series resistance) capactitor the current goes up. That's Ohms law.

"the small tuner bridge (BM3000 to 4000)... - ... it will hold to a load corresponding to about 70W RMS without overheating"
- Absolutely not.
It will not carry 20 watts, - and it was never designed to.

The original part is a very small (Siemens) rectifier. Its only job is to power the tuner and preamplifier section, - not a huge task - but still they fail because they break down physically due to their construction.

A better replacement is, as mentioned previously, the very common W06. Not a particular powerful component at all, but by far adequate for the task.

Rush-in current:

You talk about Beomaster 4000.
I am talking about Beogram 4000 (the first tangential record player from B&O).
I have lost count of the owners fitting low-ESR filter capacitors only to ask for a replacement transformer thereafter.
Been there myself, not proud to admit, but my excuse was that I wasn't as experienced then as I am now thirty-something years and hundred-something decks later.

(BM3000/BM4000) "Original rectifier (PSU) is an accident waiting to happen"
- I have replaced only three or four in 40 years. That's considerably less than the number of rectifiers I've replaced in more modern units.

Capacitors:

"Today, almost all electrolytic capacitors will be or be close to low ESR (impedance)"
- Eh.. no.
You can still buy excellent "conventional" capacitors. I have done so as late as last week.

"if the amp is in order there is no problem with the transformer or any problems with self-oscillations"

I can tell you, that is NOT the case.

I have had many amplifiers on the bench, that had run astray for "no apparent reason". They just went "poof", even

when running at what seemed to be low output or silence, with clear signs of overheating.

Beomaster 5000 (that didn't have the extra capacitors fitted to the output stage, that was later fitted to Beomaster 5500 etc.) is a definite example. Typical tell-tales are the 4,7 Ohms resistors R131 / R231 at the outputs.

They will burn if a high-power high-frequency signal is present.

And they can REALLY burn - they should ideally have been flame-proof types, but curiously they aren't.

The original transistors will never be the cause of self-oscillation.

If anything they will lose Hfe with age, and that's rarely a problem.

The problem comes if you fit high-hfe transistors in a DC-coupled output stage. That's asking for trouble.

Particularly if you also fitted low-ESR caps in the output circuit.

Everything gets "too eager".

Not good.

Thump-sound;

Look at the schematics and tell me, how the output series capacitor is charged at power on.

Through the connected speaker - right? There's no other way.

A larger capacitor will take a higher current and/or longer time to charge. = Louder thump sound.

"but the BM4000 was a simple improvement (darlingtons) which gave much more output from the same size PSU"

The same SIZE, yes. - But not the same PSU.

BM3000 and BM4000 use different transformers (BM3000: part# 8013080, BM4000:part# 8013115).

Of course, you can construct better circuits today.

Actually, that was possible already way before the construction of BM3000 and BM4000.

None of the BM3000 or BM4000 were the best construction possible. - And none of them ever claimed to be.

They were excellent, stabile and reliable daily-use livingroom receivers delivering excellent performance and superb looks for decades without much trouble at what was still a reasonable price.

They were very popular and surprisingly many are still in use today - unserviced.

You want BM3000 and BM4000 to be something they aren't and will never be - unless completely reworked, which would be a shame in my opinion.

Is producing units that aren't top-notch, all-systems go, bells and whistles galore, what you call "bad compromises"?
Not all buyers need top range - or want to pay for top range.

I'm not sure, what it is, you are trying to tell us.
That we are all wrong, doing everything the wrong way?
That B&O was all wrong?
What?

Martin

Dillen wrote:

I'm not sure, what it is, you are trying to tell us.

me either.  It's got a "I am very smart" vibe. 

Ok, so B&O don't suit your needs.  Move on.

Regarding NOS transistors:

”Purple” and ”White” plague in gold-alu intermetallic is a matter of fact:

"”Purple” and ”White” plague in gold-alu intermetallic is a matter of fact:"

- Gold is used to cover exposed pins as a means of lowering contact resistance, protecting from oxidation and to ease soldering.
Just like in some contacts/plugs/sockets - the better ones, that is.
Component and plug pins are not aluminium.
The issue you refer to affects bondings in microelectronics (ICs etc.), if gold wires are bonded to aluminium.
It has been known since at least the late 1950s as far as I can tell, so will have been well-known to component producers like RCA and Motorola in the late 60s.
I am not sure if there are any gold-alu bonds inside any transistor - I doubt it since most internal bondings seem to go from the leads
to the substrate, none of which is aluminium, but I know, that I have never seen any issues in this regard anywhere in any B&O.
I'm under the impression, that most semiconductors that failed for no apparent reason did so because of impurities in the substrate from production.
But the majority fails because of being misused/shorted/overloaded etc. rather than production issues.
I'm curious; How can you tell if a failed transistor actually failed because of the gold-alu issue?

"The specification for the 2N3055 was very wide in terms of allowable absolute maximum ratings"

- What?
The 2N3055 used by B&O was from Motorola or RCA. Their specs were certainly not wide.

Counterfeit components:

"You obviously must be doing something wrong, do you have your stock in order?"

- What?
Seriously, this is getting silly now.

Stability and self oscillation:

"This was in regard to the PSU"
- The Power supply will not self-oscilate on its own. If not designed properly, it could start to fluctuate with an "oscillating" load from an output stage and together they can increase oscillation, but what happens in the power supply in these cases is merely reactions to the main problem which is in the output stage.

"Amplifiers that have design faults such as BM5000 should be corrected"

- What design flaws would that be? It's regarded as one of the better and well-sounding amplifiers and a quite solid and stabile build.
Many are still in daily use - and many unserviced, playing along nicely.
Unserviced ones would benefit from a thorough service, after 30+ years, but that's not because of bad design.
If modified it wouldn't be the same anymore.

"But again this whole discussion is about ac-coupled amplifiers."

- Amplifiers in Beomaster 3000 and 4000 can also go into selfoscillation. I have had many on the bench whistling or motorboat'ing.

Thump Sound:

"Have you actually done this and had the experience yourself?"

- Eh... yes.
Sorry, but this is getting silly again.

Transformers:

"Why bring up BG4000?"

- Because this is one model that I have had on the bench many times with blown rectifiers and/or transformers, because owners "upgraded" the filter capacitors to low-ESR types.
The rectifier typically shorts one or two diodes and the transformer usually shorts its primary.
I don't know the exact number of units I have seen with this issue, but I can look it up. It's more than ten out of perhaps one hundred and fifty units of this model I've had on the bench over more than four decades.

"low ESR should be compensated for by a low value series resistor."

- What?
Hopefully, this is not what you mean - or practice yourself?

"Those who like the design (style) but don´t want the hassle of a lengthy repair, just pull the PSU and mainboard and replace with SMPS and ICE modules."

- Now, would that be your way of servicing vintage audio gear?

"If you had visited Wicon in Glostrup and H&K Lubcke in Köbenhavn, you would have seen the issues already when they made these parts."

- I have never come across the issues you mention.

Actually, the Beomaster 3000/4000 transformer was not made by Lübcke but by Skandinavisk Transformer, and according to the blueprints (below, actually Beocenter 3500),
the BM3000 transformer 43V secondary was designed to deliver 4,4A but specifically ordered by B&O to not fail at 10A continous load at apprx 10% mains overvoltage, = hardly underdimensioned...:

Login to view image

Login to view image

Martin

Very interesting discussion... actually both of you have good points but keep it civilized... ;-)