Sign in   |  Join   |  Help

Beosound 9000 MK1 power problem

rated by 0 users
Not Answered This post has 0 verified answers | 20 Replies | 2 Followers

Peter
Not Ranked
Germany
8 Posts
OFFLINE
Silver Member
Peter posted on Fri, Jun 4 2021 8:43 PM

Hi,

my Beosound 9000 Type 2526 SW 1.4 had problems wird CD but Tuner worked more ore less until yesterday. The display took approx 10-30 sec to show up and after approx. 5 min the unit shuts down with a plopp on the attached BeoLab 8000 speakers and went to standby. Restart with Beo4 worked so far, but now it ist completely dead! The Standby Led stays dark.

I joined "silver membership" ,downloaded the service manual, opened the device and status is as follows:

- Standby Voltage after rectifier D21 is 13V (seems ok)

- 5V Voltage ist missing (0V at Emitter TR32 PCB33, should be 5V also in standby?)

I suppose, 5V supply is defective. Unfortunately, the required schemata of the powersupply (2-25, Diagram M) and safety Unit (2-24, Diag. L) is not included in the service manual (file BS9000 Mark 1 T 2521-2530).

I would higly appreciate any help, especially the missing pages or a comlete service manual for the MK1 unit.

Thanks in advance!

All Replies

manfy
Top 500 Contributor
135 Posts
OFFLINE
Bronze Member
manfy replied on Mon, Jun 7 2021 7:42 AM

Hi Peter,

yes, it's frustrating to have pages from the service manual missing and no, I don't have those pages either, I'm afraid. But you can download the MK3 manual, which is complete. The boards and component designations have changed a bit, but I don't think that B&O changed the circuit design substantially.

But first look at the power supply block diagram in your mk1 manual (page 2-9). There you find a truth table showing which voltage will be present in which mode.
In standby you should have +5V (one of the many) and some 12V on the +25-30V rail. That 5V (let's call it 5Vstby to avoid confusion) comes from IC5 and TR37 (and yes, you should also be able to measure it at the emitter of TR32).

That's where you should start your troubleshooting. Use the MK3 manual to get an idea of how the 5Vstby regulation circuit works.

rgds,

manfy

hemenex
Top 150 Contributor
568 Posts
OFFLINE
Founder

Hi Peter,

I do have the BS9000 MK1 type 2521...2530 service manual # 3538847 09-96 but...

The diagrams M and N are here on other page numbers than you mentioned.

I have attached my 2 pages related to the power supply

and the power switching.

the diagram named L in my manual is called "motor control and detection"...

 

HTH.

By the way - where in Germany are you located? North or South?

Greetz,

  hemenex

 

Peter
Not Ranked
Germany
8 Posts
OFFLINE
Silver Member
Peter replied on Mon, Jun 7 2021 7:39 PM

Hi manfy,

 

thanks for your help. I attached 5V from an external power supply and found that there is a short circuit on the standby voltage. After disconnecting the servo and other units the shorting stays - must be somewhere on PCB3.

Thanks so far

 

Peter

Peter
Not Ranked
Germany
8 Posts
OFFLINE
Silver Member
Peter replied on Mon, Jun 7 2021 7:48 PM

Hi hemenex,

 

thanks a lot for the schmatics - now i was able to locate the defect. I suppose it is diode D22.

Unfortunately it is placed in between the impedances L11 and L9 an difficult to unsolder.

I will post more information after removal success.

Thanks

Peter

 

p.s.: I am located in south of Germany, nearby Mannheim.

manfy
Top 500 Contributor
135 Posts
OFFLINE
Bronze Member
manfy replied on Tue, Jun 8 2021 12:47 AM

Hmm, what makes you think that D22 is shorted? It's a rather solid 3A, 40V freewheeling diode.
Granted, it may be under quite some stress in such a buck converter, but so is C97, the 10V, 8200uF electrolytic capacitor -- and e-caps are more likely to fail in SMPS circuits!

Thanks, hemenex, for the missing circuit diagram. That must be an updated/corrected version because we do have Diagram N on page 2-26. But ours shows more typos and different transistor models.

--------------------------

PS: I just found a picture of that board on the net. I see that the inductors and the large electrolytic caps are leaded components. That makes it easy: desolder them first and you have easy access to D22.
In any case, what I was trying to point out above was: Be sure what component is damaged before you wildly start desoldering and re-soldering SMT components. There are several other components on that 5Vstby rail that are equally likely to fail.

Another word of caution: This board also holds the microcontroller. Pay attention to ESD safety when handling that board and when doing the surgery. If the uP gets shot, you're dead in the water! Huh?

(But ebay to the rescue: I found one guy who offers a used (but probably untested) replacement board for a mere USD650.- Surprise  )

Peter
Not Ranked
Germany
8 Posts
OFFLINE
Silver Member
Peter replied on Tue, Jun 8 2021 10:11 AM

Hi manfy,

you are completely right - it also could be a defective capacitor around. I measured the resistance between these components and ground and found lowest value between D22 and groud - this was my reasoning...?

But desoldering of L9 is a good idea. Nevertheless, in MK2 schematic there is a capacitor parallel to the diode which is not there in MK1 - so i first have to check this.

 

Thanks

 

Peter

 

manfy
Top 500 Contributor
135 Posts
OFFLINE
Bronze Member
manfy replied on Tue, Jun 8 2021 11:58 AM

Peter:

But desoldering of L9 is a good idea. Nevertheless, in MK2 schematic there is a capacitor parallel to the diode which is not there in MK1 - so i first have to check this.

Yes, I've noticed that too in the mk3 manual. There's nothing to check. This 470pF/100V cap is just filtering out the switching noise.
Much more interesting, though, is the fact that the Beo guys changed from an SS34 to an SS36, ie.from a 40V schottky to a 60V version !!!
If you find that it's really D22 that has died, we know they had a problem in all MK1 power supplies with voltage spikes caused by the L9/C97 circuit. I seriously doubt that this Schottky died of over-current since it can handle peaks up to 100A.
Additionally, if that diode died of excess voltage peaks, I can almost guarantee that C97 has taken a beating too! ....about 20 thousand beatings every second it was powered on to be exact. (considering that the DC/DC converter runs at approx 45kHz).

You catch my drift, right? Desolder C97 too and test it out of circuit. Capacitance and ESR - if you have the tools.

To round it off, I would guess that the Zener D30 is also a good candidate for failure. Even though it's outside the LC circuit, I'd guess it will participate in the freewheeling job that's done by D22.

hemenex
Top 150 Contributor
568 Posts
OFFLINE
Founder

D30 is what I wanted to point to as well. It is a Zener so at regular 5V it shouldn't do anything.

I think it's an overvoltage protection if for any reason the 5V line should go up higher. Seen this kind of protection elsewhere.

Switching current for the regulator is only D22's task...

I'm not sure it was neccessary to go up to the 60V Schottky; The higher the inverse voltage of schottkys the higher is their Vf.

I've been there as we had an assembler thinking he'd do us a favour by using a 100V instead of our 40V Schottky ending up getting hot. Vf was almost as high as the one of a standard silicon diode (more than 1V). Was in an switching regulator, too.

manfy
Top 500 Contributor
135 Posts
OFFLINE
Bronze Member
manfy replied on Tue, Jun 8 2021 3:05 PM

hemenex:

D30 is what I wanted to point to as well. It is a Zener so at regular 5V it shouldn't do anything.

I think it's an overvoltage protection if for any reason the 5V line should go up higher. Seen this kind of protection elsewhere.

Right! But they removed this Zener completely in the mk3 design and replaced it with multiple filtering caps.

I guess this mk1 was an early DC/DC converter design for them and they anticipated high inductive spikes because of high load on the 5V rail and a fairly high inductance of L9.
After collecting some actual experience they probably realized that the Schottky was fast and reliable enough and that the Zener brought no benefit. So, they dropped it in their first redesign, the mk2, and added the noise filter C120 and upped the reverse voltage on the Schottky.
The change from SS34 to SS36 was surely not just an "ad-hoc change by the manufacturing guys" because they may have run short on 40V Schottkies. Such things do happen all the time but in that case the change would never find its way into written documentation!

Peter
Not Ranked
Germany
8 Posts
OFFLINE
Silver Member
Peter replied on Tue, Jun 8 2021 7:49 PM

Hi guys,

first i unsoldered C97 which seems to be ok (8830uf measured with Fluke 115, ESR i cannot measure), then D22 which died definitely. Looking at D30 - i would test with external power supply current limited (e.g. 0.5A).

For D22 i already ordered SS36 to replace the original SS34 type - i keep you informed as soon as i got the spare parts.

thanks for your help

 

Peter

manfy
Top 500 Contributor
135 Posts
OFFLINE
Bronze Member
manfy replied on Wed, Jun 9 2021 10:46 AM

Thumbs Up
With a bit of luck your unit will be fine after replacing D22 only.
The Beo guys really did go a bit too low with the 40V Schottky:

Nominal trafo out (24V + 10%) x SQR(2) = 37.3V  Crying

37.3V max nominal level at D22 is scary. That leaves barely any headroom for surges from the grid or transients from inductor switching.

Peter
Not Ranked
Germany
8 Posts
OFFLINE
Silver Member
Peter replied on Sat, Jun 12 2021 5:19 PM

Hi,

D30 is ok. I replaced D22 and powersupply works now.

The original fault described in my first post ist still there, thus the powersupply might have died due to this error (after some time 3-5 minutes, the device shuts down with a "plopp").

Eventually a problem in the motor control unit? I disconnected the energy to the motor control unit (Cable P9 on PCB 3) and radio is now working, but i wonder why the display is not working (when radio is switched on, standby led turns out, but the display stays dark...).

I already checked +5V Display, that is ok. Should i use my scope to look at the display signals?

Any tipps how to proceed further?

 

Thanks

Peter

 

manfy
Top 500 Contributor
135 Posts
OFFLINE
Bronze Member
manfy replied on Sun, Jun 13 2021 7:33 AM

Peter:

The original fault described in my first post ist still there, thus the powersupply might have died due to this error (after some time 3-5 minutes, the device shuts down with a "plopp").

The first step is taken. Congratz!
No, this diode did not die from an overload coming from a connected board. First, the switching controller limits the output current to 1.8A and the back EMF coming from such low current is easily handled by that Schottky. Besides, when a diode fails of over-current it usually fails open. If it fails due to over-voltage, it tends to fail shorted.
It is more likely that a random power surge caused the input voltage to rise beyond 40V and that's what killed D22.

I've seen there are other threads on troubleshooting the laser and the CD circuits of BS9000. You should study those and see if that relates to your issue.

Since you have a scope, of course you use it!! First I'd check all output voltages from the power supply. See if there is a high ripple on any of those rails. If there is, it's a sign of overload in that path OR (and that's important) a sign that the output cap on that supply rail is deteriorating or failing. Fix this! Ripple can cause all sorts of problems in digital electronics, e.g. your display issue, random erratic behaviour, etc etc

Beobuddy
Top 25 Contributor
Utrecht, The Netherlands
3,928 Posts
OFFLINE
Founder
Beobuddy replied on Sun, Jun 13 2021 12:32 PM

If the startup procedure of the PSU works, then you might be able to use the testmode by using the testpoint. 
If you’re not getting the BS9000 in testmode with a Beo4, then this can give you more inside by shorting the testpoint to ground before switching on the  power. Some faults can be caused on a different location than you might expected at first. 

Page 1 of 2 (21 items) 1 2 Next > | RSS
Beoworld Security Certificate

SSL