resistor burnt out

[ben craig]

Hi, an amp of mine has stopped working. A carlsbro marin 8400

I have found a ceramic resistor looks blown and the multimeter confirms this

The resistor is a rx27 5w 2k

2 questions,

1) I can't find the correct resistor, would a 2k2 be ok?
2) are other components likely to be blown as well, or is a single fault like this common?

Any help greatly appreciated,

Many thanks, 

Ben

[Loudthud]

Isn't it supposed to be the Marlin? See if you can find a schematic and post it or a link.

[J M Fahey]

It's a 0.27 ohms 5W ceramic resistor.
Most probably an emitter ballast resistor.
The power transistor connected to it is probably shorted.
It's very likely that there are other bad parts, such as its driver.
Carlsbro SS schematics are very hard to find, and, yes, it must be a Marlin.
Probably some powered mixer.
I knew the Marlin 6-150 (6ch 150W) so 8400 hints at 8 ch 200 W (or 2 x 100W)
Post a couple good closeup pictures , maybe we can suggest something.

[Roly]

Agree.

Can anyone tell me what I can use instead of an A1688 transistor?
The PCB is marked A1688 but I cant find anything on the net.
The amp is a Carlsbro Marlin 8400.
Any help would be much appreciated.

Cheers Mitchy
Data Sheet:
sanyo 2SA1688.doc
A1688= 2SA1688 =
matsushita 2SA1532
MITSUBISHI 2SA1602
ROHM 2SA1808

I found some Carlsbro stuff at http://elektrotanya.com/ but nothing even similar.

[J M Fahey]

I don't get it.
Tube Carlsbros, which "nobody" used, a<re all over the place.
SS ones, which nobody admitted using but which I found *everywhere* in the UK, being very low cost and reasonably sounding, have "disappeared".
Oh, the bad karma of never having being pictured with Jimi or Eric on the same stage 

Back to this one, it will be even more difficult to get because it stinks of not even being a *real* Carlsbro but an OEM Korean one , made by the millions, and in this case *labelled* Carlsbro to suit some Distributor.
The fact that SMD 2SA1688 were mentioned in another post, *strongly* hints at that.

[Roly]

{@JM -  :lmao: - yes, and how did the IBM PC succeed?}


When all else fails, technish!   8|

@ben craig - you have a multimeter and seem to know a ceramic resistor when you see one, so that's a good start.

Failing an actual circuit what we need to help you is a set of well-lit, sharply focused closeups of the power amplifier/output stage area (a general establishing shot, and square-on shots of both sides of the PCB, borrow a camera if you must, flash off, macro on {"tulip mode"}, use two or three lights, close); also the type numbers of the output transistors, and of transistors that seem to be the associated drivers &c will tell us a bit.

Did it fail in the middle of a gig, or during setup?  Carefully check all your speaker leads for shorts/damage.

A burnt emitter resistor generally means a failed output transistor, but the trap is that it may have taken some other devices (like the resistor) with it, and they must ALL be found for the repair to stick.

In the meantime you need to make up a limiting lamp, per;

http://www.ssguitar.com/index.php?topic=2093.0

If you are not yet comfortable you should use the net to bone up on testing transistors with a multimeter.

I'm concerned that with this resistor burned out you may also have sustained some speaker damage, so in the meantime arrange to test your speakers individually with a known good amp; if they sound clean at low level you should be okay.

Keep your chin up, you have a couple of hundred years of collective amp fixing experience on your case here, most of it without the luxury of a circuit, and there is a real risk that you will be making a loud noise again RSN.   

[ben craig]

Wow! Thanks so much to all of you!

I have tried to take some pictures but my camera is not the best (only on the phone), If they are not good enough please let me know and i will try and borrow my brothers SLR.

This is the picture of the ceramic resistor that looks charred. -




My multimeter was low on batteries so I have re-checked this and it NOW seems to show a resistance of 2000 ohms which would match the 2k on it, so either this is not the problem or could it still be intermittent?

The rest of the pictures i have taken are here
http://s248.photobucket.com/albums/gg172/BencraigT4/Broken%20PA%20amp/

There is another circuit board behind the front panel, though I guess this isnt the cause of the problem

It stopped working at a wedding, was not on full power, but i was not in the room at the time, so dont know if anyone 'tampered' with it whilst intoxicated!

Thanks you all so much! - Did not expect this much support at all!

[Roly]

Ah, well there's a bit of a problem there.  You see Photobucket "sharing" is only for members, so I can't download these to look in detail, only microscopic thumbnails.

So it seems that the burnt out resistor isn't burnt out after all.  It doesn't look "charred" in the pix.  It is pretty unlikely that the resistor itself would be intermittent.

With the speakers disconnected, locate and carefully measure the supply voltages (large black electrolytic caps at the end).  There should be two equal voltages, maybe around + and -40 volts.  Post these voltages.

Measure the voltage at the output node (it will be connected to that coil of wire) and confirm that it is zero +/-100mV or so.  If it appears to be close to one of the supply rails, this is important.  Post this voltage.

There are four 0.5 ohm 5 watt power resistors.  Measure the voltage at each end of each resistor to chassis/ground, and post ("R1a = ?,R1b = ?, R2a = ?" etc).

It appears to have a speaker protection circuit and relay (white box with red writing on top near the wire coil).  These can give trouble in themselves, disconnecting the output when there is no fault.  You should be able to hear this relay click-in a few seconds after power up.  Can you?

You are our senses; we can only help if you provide answers to our questions.  Please review the posts above and answer the questions.

[ben craig]

 Thanks,

I thought the small brown mark on it was that it was burnt out - shows its worth double checking

I have realised that i missed out that the fuse has gone, and goes within 1 second again if i use a new one so i cant hear the relay click, or measure any voltages- am i right in thinking this was 'presumed' and that when measuring voltages etc, it should be using the light bulb limiter and a fresh fuse?
If so i will measure all of these voltages when i have made the limiter
(there are in fact 6 0.5 resistors, 2 47 and one 2k)


If the transistors are the black boxes attached to the heat sink (all with three legs) then there are
6 large ones:
    - 3x ' Toshiba 25A1943'
   - 3x  ' toshiba 25c5200"
5 smaller ones ( only 4 on heat sink - 2 of which directly soldered to board 2 on wires )
A1306 8F
C3298B 8A
2x   F    215     TIP 41C ( one not on heat sink)
F    151   TIP42C

There are also many normal looking cylindrical flat sided ones such as
S8050    d-143 and   C9013   H-212

Please let me know if the attached pics are any better and if i have missed any other questions...

Many thanks

( I may need to do a few posts, it doesnt seem to like posting what i have written with loads of attachments)

[ben craig]

These are some more close up pics

Many thanks

[J M Fahey]

These are quite good pictures.
Yet, if at all possible, get that SLR which will be sharper.
And, as Roly said, try to add some extra lights at an angle as to avoid reflections; flash is excellent light but always creates a washed out unreadabla spot.
Anyway, now we have a better idea of what you have.
And, thanks God, looks like a reasonably conventional power amp, not some tricky Class H or whatever.
1) *build and use* the lamp bulb limiter, or you'll not be able to measure anything plus continue damaging the amp.
2) measure those big power transistors (so far, on-board is fine) , we are looking for raw shorts here.
Amp off and unplugged, of course.
Also measure the mid sized ones, both on board and on heatsink.
3= the construction looks OEM Korean; but nothing bad with that, unless they decided to get "creative" and add silli fancy twists to the schematic.
So far it does not look so.

[Roly]

I thought the small brown mark on it was that it was burnt out - shows its worth double checking

No, when these go they tend to be pretty major and also burn up stuff around them.

I have realised that i missed out that the fuse has gone, and goes within 1 second again if i use a new one

Tsk tsk.  Okay, fine, but don't do that again - it will only cause more damage.  You will certainly need a limiting lamp, so you can make one up now.

http://www.ssguitar.com/index.php?topic=2093.0

so i cant hear the relay click, or measure any voltages- am i right in thinking this was 'presumed' and that when measuring voltages etc, it should be using the light bulb limiter and a fresh fuse?

Well I was being a bit optimistic and hoping that it might just be the protection relay or its associated driving circuit, but das fuzenblowen means either/or - the rectifier in the power supply is shorted, and/or at least one transistor on the +ve side and one on the -ve side of the output stage have shorted - in other words it's looking like a classically boofed output stage.

If so i will measure all of these voltages when i have made the limiter
(there are in fact 6 0.5 resistors, 2 47 and one 2k)

Ignore for the moment; we will need to know these closer to health.


The four-legged device on the heatsink in your 0129.jpg is the bridge rectifier, four diodes in a single package.  Since the power supply and output stage are on a single board it won't be easy to isolate the two, so with everything dead (actually unplugged!) and your meter on diode test range you need to have a poke about firstly in the power supply, and secondly around the output transistors.  Both the diode bridge and output transistors.  Incidentally these are 2SA1943 and 2SC5200, not "5" (datasheets attached).

Testing devices without removing them is a bit tricky due to the effects of surrounding components, but all of these devices should settle to 500-700mV or thereabouts forward conduction.  If you find a device that has only a few mV or tens of mV between legs, or much more than 700mV either polarity, it is highly suspect.  In this case you have fuse blowing so it's likely that you have semicons that are shorted rather than open.

There are quite a few webpages and videos on testing transistors if you need to brush up.

If you think you have found an output transistor that is shorted you can remove it for a more certain out-of-circuit test.  Be careful of the white heatsink compound; it behaves a bit like white paint, and it's not entirely healthy to ingest.

Hopefully with three transistors on each side, +ve and -ve, there will only be one each side that is boofed, and they will stand out from the good ones as different.

Something to consider - since you don't know the actual situation of the blow-up, you should go over all of your speaker leads very carefully looking for any wrenched and shorting connections inside the plugs, &c.  Amps generally don't chuck the towel in without a reason, and if the cause is a damaged lead you want to find and fix it before it takes out the amp again (been there, been done by that).

(as JH has just posted - thankfully at this point it looks pretty conventional)

HTH

[ben craig]

Hi guys,

Sorry for the delay.

Have now got a working multimeter,

The figures were the same for each type of large transistor
NPN's showed;  Base-Emitter and Base-Collector - 030 in both directions- and C-E as continuous (bleeped)

PNP's showed; B-E 265 both directions, B-C - 050 in both directions - and C-E as 300 both ways

I presumed that meant they had all blown: i took one of each out to test, but they are fine, so i presume this is just due to the rest of the circuitry (not sure if it is worth removing and testing all of them?)

Similar for the bridge rectifier figures made me think it was gone so removed this, but again tests fine off the board.

I have had a look at my leads for the speakers, they look fine except one of the 'male' ends is a little loose where it is riveted. Doesn't look like it could short, but could make an intermittent connection - I will swap the plug before i connect them - If i ever get it working!

You said to have a poke around the power supply, What else do i need to look at?

Thanks a lot,

Ben

[ben craig]

Hey, thanks again for all the help so far,

Can anyone help me further?


Ben

[Roly]

I'm afraid that when it comes to solid-state output stages it isn't safe to presume anything.

All of the power transistors and the drivers that feed them have to be tested out of circuit.  There are no short cuts (or at least none that don't take three times longer, result in a pile of brand new dead silicon, and in the end force you to take the long way).

There is no alternative to going through the power amp section until you are quite satisfied that you have found all the dead devices.

I hope that you now have a limiting lamp connected in series with the mains supply to your amp.

From now on we also must have no speaker load on the amp - its output must be open circuit.

Since there are three devices in parallel in the upper and lower output we can proceed for the moment by fitting only one NPN and one PNP, leaving the other four out until we get some health happening.

Now; with a tested good output pair and driver transistors; output open circuit, and a limiting lamp in series with the mains feed...

*

Power up; observe the lamp, it should dim after a couple of seconds.  If it stays bright you still have a problem in the form of a short or excessive load.

Okay?  Then measure the two main supply rails.  These will be lower than normal but should be roughly equal voltage +ve and -ve of perhaps 20-30 volts each.

Okay?  Then measure the voltage on the half-rail (output point) w.r.t. ground.  You should have a very small voltage quite close to ground, perhaps around 100mV.  If so, that's good.

If you have something else, say close to one of whatever the (reduced) supply rails are, you still have a fault.

If the half rail is about half (i.e. ~ground) then you can try fitting a second pair of known good output devices (taking care to get each in the correct place).

Repeat above (loop to *).

When you have the final pair in place and the half rail is still near ground you can try connecting a speaker and feeding a signal through.  With the limiting lamp in place this may sound a bit distorted, particularly if you try to turn it up loud.

If this goes well you can fit a higher wattage limiting lamp and repeat the test.

If that also goes well you can remove the limiting lamp and power direct, from the mains, and if you still have good signal you're home.   :tu:

The question remains however, what caused the blow-up in the first place?

HTH