since the last time I spoke of this repair has gone a bit 'of time, but then I forced myself to go more cautious than ever, given the fragility of the "patient" and, last but not least, the money invested.
Today I can finally say that most of the circuits "suspicious''work.
Everything went this weekend, after Thursday I received the registered letter containing the four TDA1514 and the two power transistors!
Here they are:
I must confess that before you turn on the welder and do everything I prayed for a long time not to see disappear in smoke the result of many hours of work .. But it was the only try.
I started with the replacement of the power transistors "hairpiece" (the BD249C and TIP145) with the "official". The 2SC3519 and 2SA1386 place. I shaped appropriately to fit the pin to the particular form of printed, as well as to rebuild a couple of tracks lifted. I can say today with a good approximation to the path of lightning: the cable is grounded and 'propagated through the mass of cross-gain stages, until reaching the final of the channel left, and that 'once burned. The right channel has suffered the same fate during the first switching amplifier, because of the damage to the power supply. Even the most 'robust chip would jump, with a' power in that state, went to the power transistors (the original 3519 short-circuited, the other open), the regulators do not regulate it, and who knows' such as short circuits, which do not I had not even bothered to verify because 'I was taken by the heat of unsolder everything' that seemed to''go.''
But now, with new parts in hand, and the tests done, I was more 'hopeful.
The replacement of the power transistors did not give me any particular concern, because the situation could only improve, because the transistors were precisely those for which the power supply was drawn.
fact, welds are completed, the tensions reflect faithfully those required: 26 volts on the positive branch and -25 on the negative. All right. Now we must be careful not to burn the transistor just posted, maybe with some careless maneuver!
thousand scruples before welding the precious TDA: schema
Before welding the TDA1514 was proper control, datasheet in hand, all the networks around the chips themselves, from inputs to outputs. The circuit is quite simple and repeated identically for the four chips, it behaves just like the classic "big op" then with the components on the input impedance, the feedback network, the capacitors on nutrition, and safety net SOAR. This last network was more complex than expected, given that the amplifier provides a function of "muting" automatic, as well 'muting a force derived from the jumper mentioned in previous posts.
Looking at the circuit, I noticed that on one hand the network of SOAR follows closely that of the sheets, with a 470K resistor to ground and a capacitor to the negative supply branch (in the datasheet from 3.3 uF, in Linn There are 22 uF!). The time constant of this RC network determines the time of "standby" mode before power amplifier complete.
internal comparators to 1514 mean that a capacitor discharge (0 ~ 0.9 V compared to the negative supply branch), the chip is off, then with a voltage between 2 and 4.5 volts
positions is in standby, and then finally with voltages between 6 and 7.25 volts is in full operation. The network
SOAR enslave the right channel was healthy, with the meter measuring the voltage on pin 3 of the pitch, could be switched from 0V to follow the trend (again with respect to-25V) rose slowly to reach the mass. The internal circuit TDA1514 limit the slope of the tension stop at the maximum value of about 7 volts seen before, so if in the measure could reach the power of mass (Ie 25 times the negative branch) at the end not attached to chip never climbed above 6.5 ~ 7V.
An unexpected failure
The branch that enslave the left channel but it was fixed at zero volts. Look carefully at the tracks that made up the branch, you could see those who were most damaged, a sign that lightning has struck with more force in that area. Something was definitely burned.
E 'needed was a measure with the tester to understand that the short circuit to 25V i-was dry. In fact, the 22 uF capacitor, once unsoldered, was in full short circuit.
Replace the condenser is a piece of cake you say, but where I find it a solid state capacitor 22uF 25V SMD? In my spare parts I did not, and even in the cards that I keep as a rich mine of SMD components.
I have noticed that capacitors like this one there was another trivial level that the voltage of 5 volts, placed in the stadium entrance. I have therefore taken the capacitor "good", I welded in place of the short, and in its place I put a 22uF electrolytic 25V I had at home, for the task that has to play it just fine!
Solved! It all turned
answered correctly, with displays that reacted promptly, and the voltages across the capacitors that came up as expected. MUTE jumper from the tensions down sharply to zero, the fire system went up slowly as it was reasonable to expect. Another problem solved!
Prudence is never enough: a test for all components
feedback resistors of the chips were all healthy, except for a 150 ohm which appeared healthy, but it was fixed with the solder (in fact they were a somewhat 'opaque), and appeared as a short circuit!
E 'enough to redo the welds and everything was fine.
For safety I also refreshed the welds of the other resistances, then made sure of their correct functioning. On balance, if the resistance of 150 ohms offending had acted as an open, on the TDA1514 would acted as a buffer, and the output voltage would be line level, so much more 'lower than that delivered by the''twin''and this would lead to a drastic drop in output power, which would be dissipated on resistors adaptation (and here 'cause I chose not to immediately put those to be 0.47 ohm!).
Offset adjustment
I also discovered the precise function of the two switches in the feedback network of a chip of the right channel and one for the left channel.
As I suspected, they serve to fix the exact value of the feedback resistors (this is a 500 ohm trimmer in series with the feedback resistance of 3.9 K) so as to ensure matching perfect gain between the stages that will end up in parallel.
For a measure with the meter, the series resistance and trimmer fit exactly as the resistance of the mirror stage: 3910 ohms in one stage and another 3915. Precise enough to make me seriously consider leaving them there, without even re-calibrate the system!
Schematics increasingly complex operational for temperature compensation
remained fairly smoky the role of two integrated dual op type TL072C, acting on the feedback of both stages. It seems that they are connected to the heat sink all'NTC fixed, presumably Then fit a network of temperature compensation. Their
operation seemed to adjust, there were no cortoocircuiti, and I'm not going into further investigations. Curiously, these two ICs are supplied to the voltage dual + and - 14V, obtained from 317-337 which generate the dual 15V, which are flanked by two medium power BJT SMD causing the fall of 1V, thus creating the 14 volt dual . Capacitors galore.
Perhaps you wanted to improve the decoupling between the power of the stadium entrance (the only one in fact to be fed to the + and - 15V) and the stage of temperature compensation.
The trick for the end of the parallel
short, everything seemed ready to host the TDA1514, but balancing the resistances of the outputs. Resistance from 0.47 ohm 5W I had already taken off, since they had "open" (always inevitable that headed to TDA as "damaged"), and the other I still want to change them.
decided, as mentioned, that for the first tests, I could use the four power resistors from 18 ohm 11W I had in my drawer.
In case of problems, I immediately felt the heat resistance, rather than the smell of burnt chips!
And so I began to pay off the first TDA1514.
decided to weld them one at a time and switch every time, trusting that in fact the four chips are operated by four separate circuits, and anyway even with a single chip amplifier should operate; at most, in the event of catastrophic failure, I would have sacrificed one of the four precious integrated.
After careful soldering of the pins, I checked the continuity 'of all the tracks and the correct correspondence with the function of each pin that was supposed to play.
seems okay.
TDA tried the first ...
Connect the power adapter and the display, and the heart pounding I put the plug. The processor reacts with a noticeable pulse, to load the big 10,000 uF capacitors. All quiet. After three seconds, switch off, and transistor chip key, all cold. Well, at least I can feel to exclude the possibility 'of a short circuit or faulty connections.
Connect a signal to the corresponding RCA CD input, speaker places a "laboratory", and turn again. All quiet, looks good half a minute with your fingers on the board for possible areas for potential to overheat, and I decided to start the music. It was so
'Radio Ga Ga''that''the Queen baptize' rebirth from the ashes of poor Linn Majik, which at this point could be considered on track for full recovery.
The music coming out of the box, the correct answer of the volume control and gearbox input confirm the full functioning of the entire circuit. Much of the doubts have been dispelled.
The road is plain
Proceed with the same task with only one chip to another channel.
With a thousand heart always light it up, and I can finally enjoy the sound in stereo!
All quiet even with the temperatures, everything gets hot to the touch but evenly and''normal''.
Taken with joy to see that everything is working fine balance of the third chip, and then the fourth.
After checking in obsessive detail do 'tension, and finally everything works as it should. All 4 chips work together, Radio Ga Ga comes out clean from the speakers (though they may be laboratory, they sound just fine! Or maybe it's the quality of the Linn me to hear music so beautiful or maybe my satisfaction?).
reached this point, a nice break and a good coffee 'is a must.
Joking aside, here is a photo caption of the work in the current state of things in more detail about Picasa
Visible: LINK
A curious cooling system
Note the small size of ' heatsink, anodized aluminum. The anodized aluminum makes electrically insulating, and in fact during the disassembly of the amplifier I have not noticed any mica insulation, and even a little 'white paste. To do a more thorough job then I proceeded to fix chips and transistors with good chemical and insulating silicone a fair amount of thermal paste.
During the tests the flap (if you can call it) has reached temperatures high enough to touch. Evidently
heat dissipation will be more efficient with the amplifier in its metal case (aluminum is, and quite often), to which the flap will transmit the heat. However, a choice really ingenious!
0 comments:
Post a Comment