Alesis RA-100 Repair

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I recently came across an Alesis RA-100 through a friend. They told me it was malfunctioning, and I am *so* glad I opened it up and poked at a few things with a multimeter before hooking it up and testing it.

After some inspection and testing, the right channel works perfectly fine. The left channel however, has 45~ volts at the speaker terminals. I'm assuming this means a shorted output transistor. The output pair is two Sanken transistors, 2SA1494 and 2SC3858.

So, my question is... What would be best best way to go about fixing this? Should I replace one of those transistors? both? should I check anything else before replacing them to ensure they don't short out again/ensure there isn't a problem elsewhere on the amp board?
 
First, fox up a circuit diagram. They're kind of hard to secure.

Second, I don't remember: is the output push-pull on the RA100? I was under the impression that as a "reference amplifier", it was, but where the pushy-pully sections were biased slightly toward conduction. And I believe the RA100 had no output DC decoupling capacitor(s), but instead relied on a balanced +/- 50V power supply.

So... guessing out loud, you could as easily have a shorted output transistor ... as an OPEN output transistor on the other half of the AB push-pull pair! Its true.

And an output transistor - in my experience - that is open is more like than one which is shorted. They tend to go into thermal run-away (or get zapped by some reverse EMF because someone does something incredibly stupid when hooking up speakers) ... and open. Maybe short first, overheat, and internally explode (invisibly to you, of course). Then they're open.

Best of luck.

I'd secure a pair of output transistors though ... both of them ... and do a double-replacement, myself. Not too expensive, and the most probable way to solve the problem.

If the problem is upwind (i.e. leading to the transistors) then again armed with the schematic, you ought to be able to confirm odd voltage readings at various key places in the circuit.

GoatGuy
 
First, fox up a circuit diagram. They're kind of hard to secure.

Second, I don't remember: is the output push-pull on the RA100? I was under the impression that as a "reference amplifier", it was, but where the pushy-pully sections were biased slightly toward conduction. And I believe the RA100 had no output DC decoupling capacitor(s), but instead relied on a balanced +/- 50V power supply.

So... guessing out loud, you could as easily have a shorted output transistor ... as an OPEN output transistor on the other half of the AB push-pull pair! Its true.

And an output transistor - in my experience - that is open is more like than one which is shorted. They tend to go into thermal run-away (or get zapped by some reverse EMF because someone does something incredibly stupid when hooking up speakers) ... and open. Maybe short first, overheat, and internally explode (invisibly to you, of course). Then they're open.

Best of luck.

I'd secure a pair of output transistors though ... both of them ... and do a double-replacement, myself. Not too expensive, and the most probable way to solve the problem.

If the problem is upwind (i.e. leading to the transistors) then again armed with the schematic, you ought to be able to confirm odd voltage readings at various key places in the circuit.

GoatGuy


Thanks a lot for the advice. :)

If I were to desolder the output transistors, and then hook the amp up and do some testing along the signal path to make sure everything looks good, would that be harmful to the amp? Generally speaking, of course, is that an acceptable way to test to see if the issue is the output transistors without buying new ones first?
 
Usually "desoldering the output transistors" ... is vexing - but entirely useful for figuring out the root of the problem.

FIRST though... you know not just theoretically how to desolder, but have a bit, and do a good, clean, quick job ... right? If not, you stand a good chance of, well, making a mess of things. Better to do some practice first on a few solder joints you make, then desolder, and unmake. Back in practice.

Then once they're desoldered - without turning on the amplifier - you can determine whether the emitter-to-base and collector-to-base junctions seem to be behaving like diodes (they should conduct in one direction, and not the other). Then you can see if emitter-to-collector is essentially an "open" (same for both PNP and NPN, with either polarity applied to emitter and collector). Ohm meters with needles are helpful here.

Now if it turns out that both transistors appear to be working as above ... one still could be ferdutzed ... but that takes more specialized equipment to ferret out. Work however on the assumption then that something leading into the output transistors is amiss. Simple ...

You have to find the ground-plane first. Shoudln't be too hard.

Then, measure the voltage that was supposed to be applied to each transistor's base pin. Again, easy enough. If the design is conventional, then it should be the same voltage (polarity reversed, of course). If the voltages are wildly different ... well that could be the problem.

Parenthetically - the RA100 is an excellent amplifier when it works. Brilliantly transparent, clean. Heats up those little heat-sinks on the side quite nicely when the power output exceeds 10 watts. Has a high enough amount of negative feedback to be super-linear, and to also show a very significant amount of damping factor (pull-down and drag-to a particular operating point).

The one thing that is dangerous about operating the amplifier without the output transistors ... is only that without the diagram, you can't tell whether you're going to be potentially blowing out other parts of the circuit.

So... back to basics.

Get the circuit diagram.

GoatGuy
 
Usually "desoldering the output transistors" ... is vexing - but entirely useful for figuring out the root of the problem.

FIRST though... you know not just theoretically how to desolder, but have a bit, and do a good, clean, quick job ... right? If not, you stand a good chance of, well, making a mess of things. Better to do some practice first on a few solder joints you make, then desolder, and unmake. Back in practice.

Then once they're desoldered - without turning on the amplifier - you can determine whether the emitter-to-base and collector-to-base junctions seem to be behaving like diodes (they should conduct in one direction, and not the other). Then you can see if emitter-to-collector is essentially an "open" (same for both PNP and NPN, with either polarity applied to emitter and collector). Ohm meters with needles are helpful here.

Now if it turns out that both transistors appear to be working as above ... one still could be ferdutzed ... but that takes more specialized equipment to ferret out. Work however on the assumption then that something leading into the output transistors is amiss. Simple ...

You have to find the ground-plane first. Shoudln't be too hard.

Then, measure the voltage that was supposed to be applied to each transistor's base pin. Again, easy enough. If the design is conventional, then it should be the same voltage (polarity reversed, of course). If the voltages are wildly different ... well that could be the problem.

Parenthetically - the RA100 is an excellent amplifier when it works. Brilliantly transparent, clean. Heats up those little heat-sinks on the side quite nicely when the power output exceeds 10 watts. Has a high enough amount of negative feedback to be super-linear, and to also show a very significant amount of damping factor (pull-down and drag-to a particular operating point).

The one thing that is dangerous about operating the amplifier without the output transistors ... is only that without the diagram, you can't tell whether you're going to be potentially blowing out other parts of the circuit.

So... back to basics.

Get the circuit diagram.

GoatGuy

I have some pretty extensive soldering/rework experience, and good equipment to work with so that shouldn't be a problem.

I do understand a fair about electronics and amplifiers in general, I just don't know quite enough to be able to look at the board and determine exactly what is what. I mean, I can tell it's push-pull, and trace the signal path through from input to output and such. I have a good multimeter and a decent analog oscilloscope as well, so testing things shouldn't be an issue either.

The biggest issue here is, while I understand the basics of the pure 'amplifier' part of the circuit, the protection circuitry and biasing stuff I just don't quite get yet, and I don't have a lot of experience working with analog electronics, so I'm doing my best and I *REALLY* appreciate the detailed and helpful responses you're giving me. :)

I'll look around and see if I can find a schematic for the RA-100 somewhere, and let you know if/when I do.
 
I have to respectfully suggest a different approach. SHorted outputs can leave DC on the output, yes, as could opens on the other side, yes. But there are other ways this happens too.

The way amps generally work is the bases of the output stages are more or less connected together (with the bais circuit keeping them a set amount apart) Then moving the whole mess up and down to drive the output. SO a stuck on or an open at the voltage amp transistor stage could leave the output stuck on a rail just as easily. The whole amp is direct coupled, so just about any transistor failkure throughout the amp could put DC on the output.

Do not use a speaker or load on the output until you are sure there is no DC on the output. DC on the output by itself won't draw any excess current, won;t stress the amp. Putting a load on it will.

Instead of pulling parts and just gong down the rows testing parts, get out your volt meter and start to hunt. You have the other channel to use for reference voltages too. I'd first verify we had both power rails into that channel. I don't know the innards of the RA100, but I imagine to have a common power supply rather than separate. But it could have channel railo fuses, look. In any case a quick check to see that both polarities have full rail voltage present. Then look at the bases of the outputs. If you have that 45v more or less on the bases of the outputs, they they are just doing their job. Whatever you put on the base is what will come out the amp output. If the bases are close to zero DC or at the opposite polarity from the output 45v, then the problem is indeed the output stage.

Nothing external to the transistor can make a shorted one appear not shorted. SO you can go down the row of output transistors checking for emitter to collector shorts right where they sit. If none measure shorted then none are shorted. On the other hand other things in the circuit can make good transistors appear shorted. SO if you find some that measure shorted, THEN pull themout and retest them. If they are indeed shorted, well they had to come out anyway to replace. If they are not really shorted, then you have to find out what parallel path in the amp has the shorted condition.

You need to find out what is wrong, not just start changing transistors.
 
In my company we repair an average of 400 amps per year of any kind ... in numbers we replace an average of 6000 electrolytics of any size per year and about 200- 300 output transistors per year ( also of any kind plastics or cans )

Today i am 47 and i am working with amp repairs for the latest 30 years In my life i have never never never and never seen an ""open"" output transistor .... only shorted ones ....

I have seen totally open transistors but only in the first or second stages of an amplifier but that also will be very rare ...

puzzled regards
sakis
 
Well, there you are. Sakis of Greece - gives the on-the-ground clue. Personally, I've seen OPEN output transistors some, but mostly shorted/fused. In the old days (with the transistors usually in those TO-3 cans... especially like the interior of the one I've attached ... OPEN was more common that shorted - because any transistor that shorted immediately conducted the big-can capacitor reservoir to ground, through those little thin wires ...which would then like a fuse, "blow open". Just saying. Depends on transistor packaging.

GoatGuy

TO3 can
 
well the alesis RA100 doesnt feature any cans and that in a way makes your advice /recomentation kinda worthless

Often there is a chance for one to learn only by pointing him to the right direction than explain him bits of history that will not provide solution .

Finally i dont know if this amp is Alesis for real since the built inside is horrible thermals also are bad and RA 500 has a huge book of history with failure ...For RA 100 i dont have that type of statistics and it may as well be better

kind regards
Sakis Of Greece
 
As suggested, I checked the bases of the big output transistors (green - 2SA1494 and 2SC3858). The base of the top one is reading +47.6v. The base of the bottom one is reading -47.6v. So I moved back to the next set (blue - 2SA1668 and 2SC4382). The base of the top one read +46.0v, and the bottom one read +46.0v as well. Onward to the next set (red - MPS8099 and MPS8599). The base of the top one read +45.0v, bottom one read +45.0v.

So, at this point, I have a few questions. Considering all three sets of transistors that I tested are PNP/NPN pairs, and taking into account that only the big output transistor bases read + and - 45~ volts while the bases of the other two pairs all read only positive 45~ volts... does that mean the problem is likely with one of these six transistors, or is this more indicative of an issue somewhere else on the board?

Another question, the transistor in the center of all the others, marked in pink, I'm not sure what it's purpose is. Is it part of the biasing circuit, or something to do with the clipping LED possibly?

(This image is large and not perfect, I know, but I figured it may be helpful since I have not been able to locate a schematic anywhere. I emailed Alesis and apparently they have been bought out since this amplifier was made and no long have the schematic either. :/)

Board Image
 
Hmmm... I have a couple concerns here. First is your output transistors. If I found V+ on the positive transistor base and V- on the negative transistor base, I'd expect both transistors to be turned on hard, blowing fuses. SO if the 2SA1494 checks OK in circuit with your meter diode test, I'd pull it and test it loose. But I'd also want to make sure the 0.22ohm power resistor in front of it was not open.

The pink transistor is your bias transistor, not the bias adjust control next to it. I'd bet at this point you probably have about +45 on all its legs too.


Are you comparing readings on the good channel?


On the outputs, the base is the leg on the left, so on the one with the -45 on the base, I assume it has also -45 on its collector, but what is on its emitter?


The fact you see +45 on all the previous bases tell me the amplifier is driing the putput to +45, and all those transistors are just following along. Hard to find without a schematic on that board of all similar size transistors, but my first suspect if the outputs are OK is the VAS transistor. But if the output wants to go negative and the output stage cant pull it there, the fed back nature of this amp will cause it to do whatever it can to bring it back to center. SO it can slam the whole thing over to the opposing rail just hoping that will center the output.

You said you had 45v on the output, was that symbol for AC or was that meaning approximate? If DC, what polarity?
 
Hmmm... I have a couple concerns here. First is your output transistors. If I found V+ on the positive transistor base and V- on the negative transistor base, I'd expect both transistors to be turned on hard, blowing fuses. SO if the 2SA1494 checks OK in circuit with your meter diode test, I'd pull it and test it loose. But I'd also want to make sure the 0.22ohm power resistor in front of it was not open.

The pink transistor is your bias transistor, not the bias adjust control next to it. I'd bet at this point you probably have about +45 on all its legs too.


Are you comparing readings on the good channel?


On the outputs, the base is the leg on the left, so on the one with the -45 on the base, I assume it has also -45 on its collector, but what is on its emitter?


The fact you see +45 on all the previous bases tell me the amplifier is driing the putput to +45, and all those transistors are just following along. Hard to find without a schematic on that board of all similar size transistors, but my first suspect if the outputs are OK is the VAS transistor. But if the output wants to go negative and the output stage cant pull it there, the fed back nature of this amp will cause it to do whatever it can to bring it back to center. SO it can slam the whole thing over to the opposing rail just hoping that will center the output.

You said you had 45v on the output, was that symbol for AC or was that meaning approximate? If DC, what polarity?

Diode testing the big outputs, for both of them, while still on the board, I put my meter in diode test mode and connected the negative lead to the base, the positive to the emitter, then changed the positive lead to the collector.

Top Transistor:
B-E: 0.060v
B-C: 0.000v

Bottom Transistor:
B-E: 0.060v
B-C: 0.487v

Both of the 0.22 ohm resistors read 0.1 ohms with my meter, while still in the circuit.

I haven't been comparing anything to the good channel, I disconnected it before I started doing these test just to make sure I didn't ruin it somehow by fumbling. :(

negative base positive collector negative emitter

For the output transistor that has -45v on it's base, it has +45v on it's collector and -45v on it's emitter.

The speaker output terminals have +45v on them, DC. All of the measurements I have done so far have been DC.
 
OK, first the diode test. You have to do NPn the reverse of PNP. Your 2SCxxx is an NPN, so you put your red probe on the base, then the black one on the emitter or collector. You should see roughly a half a volt. On the PNP, you then put the black lead on the base and the red on the other poles. And of course on either you want to NOT see a short from emiter to collector.


You may have disconnected the other channel, but you can still compare your diode test and resistance readings. Assume for now that all readings on the working channel are "correct."

Your B-E numbers look very low, but because they are identical, I am lead to believe that in circuit ther is a low resistance parallel. That upper B-C of ZERO appears like a short, flip the meter to resistance and see if you get zero ohms or something very low.
 
OK, first the diode test. You have to do NPn the reverse of PNP. Your 2SCxxx is an NPN, so you put your red probe on the base, then the black one on the emitter or collector. You should see roughly a half a volt. On the PNP, you then put the black lead on the base and the red on the other poles. And of course on either you want to NOT see a short from emiter to collector.


You may have disconnected the other channel, but you can still compare your diode test and resistance readings. Assume for now that all readings on the working channel are "correct."

Your B-E numbers look very low, but because they are identical, I am lead to believe that in circuit ther is a low resistance parallel. That upper B-C of ZERO appears like a short, flip the meter to resistance and see if you get zero ohms or something very low.

Okay so I did all that for both the good channel and the bad one, and this is the result:

Good Channel:

2SC3858:
B-E: 0.064v - 97.6k ohm
B-C: 0.445v - 154.6k ohm

2SA1494:
B-E: 0.064v - 97.6k ohm
B-C: 0.465v - 144.1k ohm

Both 0.22ohm resistors: 0.2ohm



Bad Channel:

2SC3858:
B-E: 0.064v - 96.6k ohm
B-C: 0.448v - 156.8k ohm

2SA1494:
B-E: 0.065v - 99.4kohm
B-C: 0.468v - 142.2k ohm

Both 0.22ohm resistors: 0.2ohm


Since everything is so close to the same, I'm guessing the problem is farther back?
 
.... this is turning bad ..... i wonder how much time consumed to produce these pictures with pcb footage ( which actually doesnt help much because most people are familiar with schematic reading which is also available for free in the internet )

Check outputs with a dvm between C-E with the diode tester for shorts ...no load no power conditions

repeat fro drivers and pre drivers

get a schematic
locate small value resistors ( acting like fuses values 10 to 100 ohm ) check them all ....

the failure lays somewhere there
 
RA-100 hand-drawn schematics

Hard to follow, but scroll to the bottom. There it appears that the output is in a SZIKLAI (darlington-like, but with complimentary transistors) configuration.

The "problem" could quite easily be that the complimentary pair in the affected Sziklai set is the one that's out. Easily. That's where I'd be looking next.

Also ... from your first test and numbers, where you identified the "top transistor" as having 0.000V B-C ... I would have said, "well there's the problem". But in your second set of tests, that canard went away. Why? Was the first measurement in error?

IF YES - then look at the compliments.
IF NO - then, well, the "shorted transistor" is obvious.

GoatGuy
 
Okay so I retested the outputs, drivers, pre drivers, and the bias transistor as well just to make sure... all of them test good and have comparable values to the good channel. The pre-driver set, the MPS8099 and MPS8599 tested 0v C-E after flashing something like 0.9v on the DMM, but the same transistors on the good channel do the same thing so it must be a parallel component messing up that reading. Here are the results (This is all from the non-working channel, though I did compare all the values to the working channel):

NPN:

Output - 2SC3858:
B-E: 0.065v
B-C: 0.435v
C-E: 0.445v

Driver - 2SC4382:
B-E: 0.552v
B-C: 0.549v
C-E: 0.454v

Pre Driver - MPS8099:
B-E: 0.642v
B-C: 0.645v
C-E: 0.000v <-- DMM Flashes 0.934v then goes to 0.000v, Same thing happens with good board.

Bias Transistor - MPS8099:
B-E: 0.580v
B-C: 0.572v
C-E: 0.958v


PNP:

Output - 2SA1494:
B-E: 0.065v
B-C: 0.454v
C-E: 0.469v

Driver - 2SA1668:
B-E: 0.556v
B-C: 0.553v
C-E: 0.434v

Pre Driver - MPS8599:
B-E: 0.644v
B-C: 0.639v
C-E: 0.000v <-- DMM Flashes 0.982v then goes to 0.000v, Same thing happens with good board.
 
Hi!
Did anybody have circuit diagram of amplifier Alesis ra-100?
Is it possible to place the diagram here?
I have this amplifier. It works well, but in my opinion too dim sound due to using the electrolytic capacitors in the audio path. I plan to replace them (like C215) with film capacitors and may be doing some other upgrade.
Can you help me?
 
Alesis RA100 vs. newer replacement model RA150:

Sakis is much more able to comment on workmanship and parts quality, and I am also not expert on repair but I can say few words about how these two sound to me because I was able to hear both in my room with my speakers.

I like the sound of old RA100 more. The schematics are very different. RA150 is fully symetrical from input to output and has complementary Darlington (EF) outputs unlike RA100 which has Sziklai outputs. RA100 is faster, with natural dynamics and more detailed.

Lerpa, if RA100 is dark sounding to you I bet it is source or speakers (or both) that is responsible. It is possible that build quality of RA100 is not great, but the sound is good (I am using Tannoy professional studio monitoring 10" loudspeaker kit K2558 for home listening).
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.