Beolab 150 overload fault

[cloudrock]

Hi everyone,

I feel a bit unsure whether this is the right forum to seek some advice and help ?

One of the Beolab Penta MK1 amplifiers is not working ?!

So far I found a dead output transistor, a 2SC3281 Toshiba, tested all involved transistors the way down, replaced some but still the overload kicks in.
The faulty 2SC3281 was replaced by a MJL3281A.

One question bugs me: how crucial is the matching with its complement
Transistor 2SA1302 ?

The output stage has 3 pairs of these transistors.

All elcos, the big reservoir caps and trimmer pots have been replaced apart from a few transistors down from the output stabe.

Perhaps someone is familiar with that Beolab 150 amplifier and can help....?

Would be very much appreciated.

Regards
Cloudrock

[Max Ripple]

This is a sophisticated amplifier, and usually very robust too. Is is part of a Beolab Penta column, as the model you quote is for the stand alone amplifier only?
The protection system senses d.c. at the speaker output and shuts off the main supply.
The output transistors are indeed critical to the correct function, and matched sets ideally should be used as replacements. You will not be able to get the unit to function correctly by only replacing obviously failed components.
Modern equivalents are available, and indeed too, full kits of parts to restore these amplifiers.
You may need to check the crossover components inside the column too, if dc has been across the 'speaker.
You have an email message too.
Nick

[cloudrock]

Hi Nick,

Many thanks for taking the time to respond, something I sadly missed on the Beoworld forum on that occasion !
A brief history of that Beolab150 amplifier:
its not a stand-alone version but was part of a Beolab Penta MK1 speaker, purchased with exactly that overload issue.
I took the amp apart to investigate, discovered a dead 2SC3281 transistor which I replaced
With a modern equivalent MJL3281A. Followed the circuit path down from the output and basically tested every transistor in the path off-board for damage with a Peak transistor tester....no findings ?
All emitter resistors (470mOhm 2watt) are on the edge of their spec and will get replaced as well.
I also replaced all old trimmer pots and all elcos, including the two big reservoir capacitors.
FYI, i was fortunate to find a pair of Beolab 150 MK2 replacements which were fitted.
All crossovers have been checked and up-graded with higher spec capacitors - a costly exercise.
I am aware those output transistor pairs work best when matched but also read that most modern replacements are pretty much matched already to their compliments. It also seems a tricky procedure for the DIY hobbyist to perform that at home as costly equipment is neccessary.

Anyhow, as it stands I hit a big roadblock:
the green led comes on when I plug it in and turn the ON the amp but as soon I connect P5 to the output stage it goes straight into the orange/red overload.
I hate to think its the main transformer as it delivers the voltage as per SM.....something I am missing here but don't no what ??

Regards
Cloudrock

[Marconi_MPT4]

Looking at the Toshiba 2SC3281 data sheet, two DC current gain hFE rankings were supplied. R = 55 ~ 110, O = 80 ~ 160. However this is not specified in the B&O parts list, just part number 8320645. Back in the day that was all you needed. MJL3281A would need to to matched within either gain band, but there are other differences to consider.

Was the original transistor shorted b-e or c-e, or maybe open circuit?

When replacing balance pot R59, it is important to make sure the initial setting is roughly midway. Transistor TR6 monitors offset which if out of balance might trigger fault indication.
Bias pot R95 should start off at minimum resistance setting.

If P5 is connected and then system is powered on, does the power supply relay enable then trip out or is it just the fault indicator and the amplifier has supply voltage?

Rich

[cloudrock]

Hi Rich,

Many thanks for your input - sadly I don't have that faulty 2SC3281 transistor anymore - there was a short either B-C or C-E from memory.

I also checked the 6 Emitter resistors which are 470mOhm, 2Watt 10%
All of them are going to get replaced as they are all about +20% !!

They should come in early next week. In the meantime I paired the MJl3281 with its modern compliment a MJL1302 instead the orIginal 2SA1302 which should be a better match.

Once I installed all emitter resistors I shall test again. From memory once plugged in and turned on (with P5 connected to the poweramp) the overload indicator comes on while the power replay shuts off at the same time - it is like the moment I turn on ...click, off it goes.

I will confirm after the new parts are in. I took your advice regarding the two trimmer pits R59 and R95.

1TR6 tested fine !

Will report back

Cloudrock

[Marconi_MPT4]

One item I forgot to mention is that fuse (safety) resistor R79 150Ω going open circuit will result in roughly -65V DC appearing at the loudspeaker output and an instant system trip!

Rich

[Max Ripple]

Further to my earlier post,what I meant about matching the output transistors,is that it's best (in my experience)to replace the transistors with the same brand,rather than to take pot luck.
If one device has failed,it's quite likely to have damaged others in the network too.
Are there any signs of overheating on the pcb?Sometimes we see this nowaday's,and the board can actually become conductive too!
It's crucial that the dc balance,and quiescant current setting preset's are replaced,as they known troublemakers.
The dc balance preset has to be set up(for min dc at the output) when the amp is cold,and with no signal applied,then the quiescant current can be set too.
Do you know if the replacement electrolytics are "low impedance" types?
Curiously,these can exacerbate instabillity in high gain amps.
Do you have a service manual btw?
Nick

[cloudrock]

Hi Rich,

Sorry for the late reply, I am still waiting for those emitter resistors

Good thought about that fuse resistor - I checked it before, it tested 148Ohm.

So that side is covered.

I am really hoping fir a delivery today as it has been almost 10days wait !!

Will report back asap

Cloudrock

[cloudrock]

Yes Nick,

Thanks for asking I have the service manual.

At first I only replaced the Toshiba with a MJL 3842 but have now also replaced
Its Toshiba compliment with a MJL brand to stay with the same brand.

Regards
Cloudrock

[cloudrock]

Time to post a quick report:

The new emitter resistors are finally in place.
Ok, if P5 is connected to the main board I can measure a DC voltage on the
Speaker connecting wires of around 60 VDC momentarily !!!

That of course trips the relay in an instant. But the question is how is it possible to get that amount of DC voltage on to the speakers connecting wires ??

Capacitor C48 is specified 10nF / 400V - I can measure 8.9nF but of course I can not verify the 400V stability !

This is a far as I got right now......not that great 😢

Regards
Cloudrock

[Max Ripple]

It seems as though one of the output transistors is breaking down under voltage.
It's significant that one of them was shorted originally,and I suspect that other damage was done then too.
It's a pain I know,but can you isolate all the output transistors by desoldering their base and collector connections?
The amp should still run without them connected,at least to be able to see if the dc conditions are normal.
You are looking for near zero dc at the mid point connection to the 'speakers.
Set R95 to minimum res by the way,and R59 to about midway.
If the amp does indeed run and dc reading's are normal(ish)the safety system should not trip,and it will prove that one or more of the output devices is faulty.
In that case,replacement of them all together with same brand devices would be the way to go,but not cheap of course.
Nick

[cloudrock]

Ok, here it goes:

As suggested I disconnected the B and C of the output transistor pair one after another to check if it was perhaps only one pair but ended up disconnecting all 3
Pairs.

When turning on the amp is doesn't trigger the safety system but still shows the red LED because there is no incomming signal.

I can turn it on momentarily to green by switching the sensitivity or bass level switch.
However, No matter how I turn trimmer 59 I can not get the speaker output DC down to zero....best would be around 90 mV, fluctuating !! 90mV isn't too bad and should not be enough to trigger the overload circuit ?!

So, one of the big questions seem to be:
does replacing all output transistor pairs fix the problem ? Its not about money but I had done this before when I suspected having purchased fakes !

They were all the same brand - MJL..... - and really didn't break the bank.

The alternative is having a non-working Beolab Penta amp sitting in the corner.

When I purchased the complete set of output transistors I tested all of them with a Peak tester...all tested ok.

So, where to from here ?? 😡😡😡😡

Cloudrock

[Max Ripple]

Yes indeed,replace all six output devices.
You can get the equivalents MJL3281 and MJL1302 from RS components.They have a depot in NSW(www.au.rs-online.com)
There are indeed some dodgy devices around,but you are much less likely to find them when sourced from a reputeable supplier such as RS.
The 90mV dc level is not too frightening,but still too high,and this may be due to noise coming through from previous stages.
The two voltage regulators supplying +15 and -15v have decoupling caps (I think 330uf?which dry out.Replacing all the caps in the standby circuit though is recommended too.
Incidentally,you refer to the fusible resistor R79 as "1480ohm" when it should be 150 ohm,and the output device emitter resistors as "470 ohm" instead of 0.47ohm.Hopefully,those were typo's,if not,revisit?
Try and get the zero dc voltage reading lower and check all the other voltages are correct before you install the new output devices.
Nick

[cloudrock]

Yes, I agree the DC voltage at the speaker output should indeed be much closer
to 0 VDC.

Yes again, those damn typos !! Of course the replaced emitter resistors are all
470 mOhm and the fuse resistor measured 148 Ohm !!

I buy quite a lot from RS components - all other local shops here are pretty useless for that kind of work.
Already have my hands on a set of ON MJL-type output transistors - cost is about
$15 for three pairs.

If that is all that's required I would be very happy - BTW, all elcos have been replaced ! I am quietly confident that part is covered.....as for the new set of output transistors.....?? Hope dies last as they say.

I shall be back with another update soon.

Cloudrock

[cloudrock]

Just managed to buy another 3 new pairs, ordered directly from
ON-semiconductors.....so there should be no dramas to expect ?!

Wish us luck 🤞🙏

I'll be back - Cloudrock

[cloudrock]

Ok,
The new set of output transistors is installed.....sadly the (suspected) outcome
Was negative ☹️
Here is what happened:
Switching on the amp (auto or link) only shows the red light, meaning the relay does not click.
If however I move the sensitivity switch the relay very briefly clicks, the green light come on for a split second until its turning to orange again !

So, at this stage I am no wiser as I had the same outcome before. Bear in mind the previous set of output transistors were supplied by RS-Components.

Something else I can not explain is the DC reading on the speaker wires ?
First showing about 0.7V and once the orange light is on this voltage briefly goes up to about 58V.

I noticed two capacitors C35 and C48 are specified for 500V and 400V - i wonder how critical their high voltage spec is ??
Yes, I understand this sounds like fishing, but.....could it be the overload indicator circuit playing up ??

Right now all seems very dark and I might as well abort the whole project.
The hours are mounting and frankly, I still do not have a decisive lead.

Something somewhere isn't playing ball and sonetimes one has to admit defeat.

Thank you all for your interest, much appreciated.

Regards
Cloudrock

[Max Ripple]

We all get these trying faults,and it sometimes pay's to have a rest from it and come back with a fresh approach.
It's easy to get into a situation where you are technically "chasing your own tail"and getting nowhere!
It's often a good idea to "go back to go"
Just as an overview,the amplifier is designed to automatically switch on in response to a signal,when set to "auto" mode."link" is a B&O specific function.
Your description of switch on is consistant with that,as touching the sensitivity switch will create noise and trigger the 'amp "on".
The relay connects +68 and -68 voltages to the output stage,and the stage should (d.c)balance itself,and settle at about 10mV at the speaker output.If it does'nt,and a greater dc voltage appears there,the safety circuit will kick in and disconnect the supply to protect the output stage and 'speakers.
Establishing the dc balance is crucial,and I wonder if you tried the amp without the new output transistors connected as I suggested earlier?
I think that you need to go over all the earlier points again,making sure that all the transistors are reading ok,particularly TR23/28 and IC1,which is a Darlington transistor mounted on the heat sink.As a Darlington type,it's base/emitter junction will read higher than a normal transistor.
Incidentally,you mention the output transistor emitter resistors are "470mOhm".You do mean 0.47 ohm?
Don't worry about C35 and C48.
Soldier on sir,but have a rest from it and restore your confidence repairing something else in the meantime?
We used to say with tricky faults,"I'll think about that when I get back from the pub",but our's are all still shut here!
I dare say that you could pop out for a Tooheys(or two?)
Nick

[cloudrock]

Just a quick one Nick - many thanks for your encouraging words 🤙😁

Yes, I tried the amp with disconnecting all B & C from the "old" output transistors one pair after another ...... 🙄🤗

You are quite right, as there is no incoming signal that would turn on the amp in auto mode I need to trigger it with for example moving the sensitivity switch as
explained.

By doing this it should still stay on (green light!) for a couple of seconds before switching iteself off again...but it doesn't 🤬

So, as suggested I shall go into recovery mode and

Regards, cloudrock

[Max Ripple]

The amplifier should stay "on" for about 15 minutes after being triggered by a signal.If the signal eventually drops out,the system switches off.
When connected to a compatible B&O system,the amp is triggered by a pulse from the other unit,so when the system is turned off,the BL150 also turns off.
Your BL150 still seems to be shutting down straight away due to the protection system detecting dc at the speaker connection?
Hope this further insight helps?
Nick

[cloudrock]

Hi Nick,

All sounds very logical, but: where & how is that high DC voltages across the speaker wires coming from

That is certainly more than enough reason to activate the protection circuit !

Ideally it should be 0 VDC or very near that figure.

Again, I have gone back to disconnect all B&C connections on all output transistor pairs, tested again transistors TR23 & 24 and TR28 & 29...all fine .

All axes have been under lock & key to prevent fatal damage. pm returned !

Regards
Cloudrock