Bad solderjoint at the large 1000uF next to the fuse. 

Or the cap itself is faulty. I replace them always. 

I agree.
Or someone with very little electronic knowledge made the mistake of fitting Schottky diodes as rectifiers.
I have seen this twice now (Eh???)

Martin

Hi Martin

Just curious - why this comment? 

"Or someone with very little electronic knowledge made the mistake of fitting Schottky diodes as rectifiers."

Nothing personal - just the technical reasons.

Today, Schottky diodes have become the defacto first-choice for rectifiers in power supplies.

In fact, their first widescale applications in the 1990's, were for the secondary Low-voltage stages of all VCR power supplies.

Today you will hardly see a power supply without Schottky diodes.

Of course, they are not as forgiving as a regular diode, and the tech should be over-generous in selecting a suitable current rating.

Menahem

Maybe for the exact reasons you've just mentioned?

Schottky diodes are overkill for an old fasion powersupply as used in the BG6500.  Using the Schottky's in combination with the original used Roederstein capacitor, will result inevitable cease of it. The rush-in current will be to high for it and will result in a shortend cap.

The voltage-drop is much lower and results in a higher voltage at the entrance of the rest of the powersupply.

So, juts fiitting some ordinairy older 1N400x or 1N540x would be fine.

If you would use these more modern Schottky diodes into older devices like the BM900/1000 etc, then you will end up with fried output transistors.

So, I think you (generally speaking) need to think twice about where you are going to use these parts and if they are suited for the job, instead of just thinking that newer more modern parts are always better.

Thanks for your reply, but personal experience proves otherwise.

Firstly, if a tech is going to change the diodes, then smartness suggests that he also upgrades the filter cap as well. Modern filter caps have a much better capability to handle the inrush than the older ones, so THAT is not a problem. If one is going to do the job, then do it completely, not just a half-job.

Secondly, the extra voltage is NOT a problem. ALL BeoMachines have a voltage regulator after the bridge, and an extra volt on the regulator input is NOT going to make any difference at all. Not only that, but the smoother waveform of a Schottky vs a 1N400x, reduces the input stress on the regulator.

Fried output Transistors? Perhaps on some isolated ancient machines, but not on anything made after 1975.

My own BG3400 was upgraded with Schottky's well over 10 years ago. I haven't had to open it since.

Ditto my BeoGram 4000.

Martin mentioned the BG6500. My own BG4500 (the same as the BG6500), was upgraded with Schottky's on EXACTLY that circuit over 10 years ago, during a complete overhaul and upgrade. I have NEVER had a minute's problem with it.

I have done the same overhaul on many other BeoGrams of the same series, and am in regular contact with the owners - none has ever had a problem.

I intend to upgrade the 2 round rectifiers in my recently-acquired and previously-dead BM4400 with 4 Schottky's wired together, with absolute confidence that there will be no unfortunate side-effects, ever. I'll keep you posted honestly, if I'm subsequently proved a fool.

The only caveat which I can issue is stay away from cheap Chinese Schottky's - you're asking for trouble. Anyone who is investing the effort to do such an upgrade should not be looking anywhere else than Vishay's diodes. Yes, they cost more, but they last n times longer than the Chinese look-alikes.

Of course, I do not willy-nilly change components - I make a very  comprehensive examination of what the circuits require, and match the new components to those requirements.

There may be some ancient BeoMachines which will experience cardiac arrest with Schottky's, however, to make a blanket condemnation both of the installer's intelligence level and the equipment, simply does not suit the integrity of the people on this forum.

Martin mentioned in another thread how important the right belt is for the BG1000 and other round-belt turntables.

He said perfectly intelligent things about the cheap Chinese look-alikes.

So it is with Schottky diodes. There is Vishay, and there is junk. Do not expect junk to last - in fact, expect it to damage your BeoMachine.

But do not write off all Schottky's, just because the Chinese junk damaged the machine.

Menahem

Conventional diodes are more forgiving, yes, - and far more reliable.
I've seen two cases of the above Beogram types with shorted and/or physically exploded Schottky diodes and in both
cases I could tell, that they weren't factory fitted.
In both cases the diodes were the only culprits, and fitting conventional (I like to use 1N4002-1N4006) diodes
brought back normal function.
You can design circuits to use Schottky diodes but replacing conventional diodes with Schottkys in an existing circuit is pointless.
Waste of time and money in the best of cases, and in rectifier positions it would be asking for trouble and unstability.
Been there, done it, saw it twice, got the T-shirt.
Paid for it too.

I never saw a blown output stage transistor because of a Schottky-based rectifier, but it sure is possible.
And it has nothing to do with the circuit being produced pre- or post-1975 (where did that year come from anyways?),
10 ohms year 1960 is still 10 ohms year 2016. 10 Volts are also still 10 Volts.

But how come that now, when Schottky diodes are available, we are still using conventional diodes?

Schottky barrier diodes have a lower turn-on voltage and faster recovery time.
They are generally "faster".
There are some places, where a Schottky diode is preferred but there are many more applications, where it's
not recommended to use Schottky diodes.
- And for a number of reasons;

They have a higher reverse leak current.
They don't handle power particularly good - and if put to work with higher currents they lose their Schottky properties anyways.
They are considerably more temperature dependent.
Their internal capacitance is even more temperature dependent, often to the extent of being quite unpredictable.
They are typically larger.
And they cost a bit more, up to twice the amount of a similar silicon diode, though this is probably mainly of interest in large production.

The temperature dependency is a problem, which together with the higher reverse leak current and the almost ever
changing internal capacitance can cause a thermal runaway, which typically ends with the diode shorting - or
even exploding, depending on the current flow in the given application.
(The diode doesn't even have to work in a high voltage circuit to do this).
In other words; In areas where temperature changes are to be expected (f.e. power supplies and amplifiers), Schottky
diodes should only be used with the utmost care.
Any place a conventional silicone diode can be used instead of a Schottky, it should - to make a much more reliable circuit.

In existing circuits, like power supplies, where the first component following the rectifier and
filter capacitor is a voltage regulator, the lower forward voltage drop of the Schottky diodes
will make the input voltage to the regulator higher, which again increases the voltage drop that the
regulator has to make to output its nominated voltage. This puts a higher load
on the regulator and it will have to dissipate more energy (it will run warmer). The diodes will then of
course dissipate less but the total heat dissipation will be the same and the heat is often better distributed at the
four diodes (that rarely needs additional cooling anyways).
If you design a conventional power supply circuit from scratch, you can choose a transformer that outputs a lower
voltage but there's really no point in doing so if only to spend money on more expensive rectifier diodes
(- that won't do the job any better - and give unstability with changing temperature).
If you at the same time put a low-ESR filter capacitor after the rectifier, keeping the lower forward
voltage drop of the Schottky diodes in mind, then the transformer (and the fuse!) will see a higher
rush-in current.
And so will of course the diodes...
In most cases, the rush-in current is not a problem for the capacitor but potentially for all the other
components upstream, that has to carry it.
- And there's really no need for either a low.-ESR filter capacitor nor "fast" rectifier diodes in the Beogram power supply (100Hz).

High frequency circuits are where the Schottky diodes usually do their best and where it makes sense
to call in "faster" diodes.
Radio frequency circuits, IF-stages, mixers/frequency changers and the likes.
Transistors used in high frequency switching circuits can also, in some cases, benefit
from a "Schottky clamp" between base and collector. Think of the 74LS TTL logic IC series, where the LS stands
for Low-power Schottky - allowing them to operate at higher frequencies than the standard not-LS types.
Do remember to fit a 100nF capacitor across the supply very near to the IC and preferably directly on top of it - or
the schottky-based circuit will give unpredictable results (!!!!!).

If you design your own circuit, you can design it to use Schottky diodes if you are absolutely sure you
know, what you are doing.
Replacing conventional silicon diodes with Schottky diodes in an existing circuit will
not improve the circuit in any way. Quite contrary if anything.

For 50Hz rectifiers, replacing conventional diodes with Schottky diodes because they are "faster" would
be downright ridiculous.
The good old 1N400x series (and something like 1N5401 in higher current applications) - would be an
excellent choice for rectifier purposes in most stereo/audio applications.

Substituting conventional diodes with Schottky diodes to make circuits run cooler simply doesn't make sense.
Doing it in the name of increased stability is electronics misunderstood. Regardless of the brand.

And this is by no means meant personal. It's just facts. 😕

- And I don't buy chinese.
Been there too. 😂

Martin

All quite correct - no arguments.

But there is always room for well-thought-out exceptions ;-)

Test a Vishay 31DQ10 for reverse leakage - let me know what you find.

Initially this was my skepticism "barrier" to investigating Schottky's - I was pleasantly surprised.

As I stated, the tech considering a Schottky upgrade must be over-generous in his choice of current rating - meaning double or triple.

The main benefit of using Schottky, is to my ear, better sound with a wider and more dynamic soundstage.

I've A-B'd it, and so I find it.

I mentioned 1975, because that's the approximate year that B&O stopped using the TO-3 2N3055-type low-gain output transistors, and changed the amp circuit design to the TIP141/6 high-gain Darlingtons.

That's where a change to Schottky's will be beneficial and audible - in my experience.

Take note that I did NOT (and still do NOT) advocate replacing the large rectangular B80C main 50Hz bridges with Schottky's.

In fact, I've never had reason to replace any B80C bridge in any European machine on which I've worked.

As I mentioned - the round 1A bridges are the components I look at for upgrades - and they are generally feeding the IF/RF circuits in BeoMasters.

Wholesale "upgrades" with no forethought is not an upgrade, and may be destructive.

Test any Schottky diode for reverse leakage, put to work with varying temperature.
And test their Schottky properties when handling current. (Save your time - there won't be any 😂).
Schottky diodes are Schottky diodes.
With their benefits and drawbacks.
It's in their nature.

TIP141/TIP146 etc. darlingtons already have an internal diode between E-C so I would hesitate adding a
Schottky clamp here (or put a Schottky diode anywhere else in AF-circuits for that matter).
Seeing as the benefit of Schottky diodes really first becomes evident above several MHz, I really
don't think that there will be any audible difference below 20KHz - let alone if used in the (50Hz) power supply.

No Schottky diodes in power supplies, power amplifiers and AF-circuits.
Keep them away from places where power is handled and where circuits have feedback.
Use them in tuners and particularly in high-speed digital/switching circuits (computers and the likes), that's where they belong.

You are free to do and fit whatever you feel for, it's your machines (and your clients) but I
would never use a Schottky diode where I could use a conventional diode. No exceptions.

We may not agree but electronic facts are electronic facts and
facts ask noone to agree. 😂

Martin

Most circuits have a regulated power management.

But most, if not all, the output stages have an unregulated powersuply.

I referred to the older amplifers with germanium transistors/output stages. These are fed with selenium rectifiers. The used "diodes" inside regulate the rush-in current. Replacing these with more modern version (e.g.Schottky) diode will result in silent output stages.

Schottky diodes are only neccessary where low drop voltages are required at higher switching supplys (PWM). Frequency's here used are above 30khz.

And As Martin already stated, you're shifting a part of the power dissipation from the diodes to the voltage regulator. And the older regulators are far from efficient.

Ever thought of what happens in a circuit when you switch the power on? That's the moment where the most parts cease and at older used age ,( like e.g. light bulbs). The highest current will be at that moment. All, or most current equipment are suited with special rush-in resistors. To protect the circuits (with used schottky's) diodes behind it.

The main thing is to think what the appropriate needed components are. It's like a chain with it's weakest link.