You need to put a 6 -10 ohm 100 W resistor in series with the AC power.
This is typically a 100 W light bulb, although the newer halogen bulbs have some sort of diode in there and are not suitable.
One gets a simple edison base fixture, puts it in a grounded box with a circuit breaker or big (15 A) fuse, put a receptacle on the front. That way if the wire to the bulb comes loose you don't get shocked, the fuse blows.
With that in series of the power supply, every thing collapses in voltage and current and you can look at the voltages with a DVM and find where your miswire or bad solder joint is.
First step of course, copy out the circuit and buzz it with a yellow pencil tracing out the good wires. Count the wires on each node and make sure the circuit doesn't have extras or missing. A wirelist written out from the schematic diagram makes obvious how many connections are on each node. Wirelist writing was something I learned at my second electronics job, not in college. The operators that build circuits at businesses work off a wirelist, not a schematic diagram.
This is typically a 100 W light bulb, although the newer halogen bulbs have some sort of diode in there and are not suitable.
One gets a simple edison base fixture, puts it in a grounded box with a circuit breaker or big (15 A) fuse, put a receptacle on the front. That way if the wire to the bulb comes loose you don't get shocked, the fuse blows.
With that in series of the power supply, every thing collapses in voltage and current and you can look at the voltages with a DVM and find where your miswire or bad solder joint is.
First step of course, copy out the circuit and buzz it with a yellow pencil tracing out the good wires. Count the wires on each node and make sure the circuit doesn't have extras or missing. A wirelist written out from the schematic diagram makes obvious how many connections are on each node. Wirelist writing was something I learned at my second electronics job, not in college. The operators that build circuits at businesses work off a wirelist, not a schematic diagram.
Thw circuit it's fine..transistors are not shorted.
If the value of r5-r6 r7-r8 is not the right one for my output transistors it migh draw a few amps ( happend with other amp, changed that resisotrs and it draw about 100mA quisance current) but why does the power draw rise on this one...on that amp it was steady at a few amps..
If the value of r5-r6 r7-r8 is not the right one for my output transistors it migh draw a few amps ( happend with other amp, changed that resisotrs and it draw about 100mA quisance current) but why does the power draw rise on this one...on that amp it was steady at a few amps..
Buno, all the values of the resistors at the output stage - 0.5 , 13 , 1ohm etc... must be there. The value of these is what set the bias current.
Take things calmly...
As HAL9000 said. "Dave, I feel much better now, I really do. Why don't you take a stress pill and think things over"
Take things calmly...
As HAL9000 said. "Dave, I feel much better now, I really do. Why don't you take a stress pill and think things over"
The resistors from base of bd to rails 2.7 -3.3k change it to 4,5,6k nothing... Changed it to 20k not 0 amps
And i have music at the output now!! I have to " tweak" that value to obtain 20-30mA quisance, it sounds good tho, distorts obly at high volume
And i have music at the output now!! I have to " tweak" that value to obtain 20-30mA quisance, it sounds good tho, distorts obly at high volume
I have no resistor (1ohm) on the 2sa,2sc tho, someone posted here there will work without them, and that they are for current limiting and help for multiple transistors in parallel
Should I pit two pot's and see the " right resistor value " also why the values the author presented don't work for me, is it that I have different output stage transistors? Still 2sc,2sa
Very bad IMHO. You've got to have emitter resistors on Output transistors.
If you have different output transistors than 2sc5200 2sa1943, then you have to measure the current across the emitter resistors to make sure it is a safe value at idle. No emitter resistors no measurement. TO247 packages 20-100 ma (lower better), TO220 packages 10-30 ma .
If this is the spreader voltage stack for the output tranistors (I didn't download this, I don't want one) you set the resistors near the rails to not exceed the power limits on the other two or the current available from the op amp to push it around. You set the middle resistors to get enough idle output current through the output transistors as read across the emitter resistors. As every pair of output transistors has its own design and production history, setting the middle two tends to be experimental. That is generally a voltage spread of 1.2 to 1.4 on the bases of the two output tranistors, but the idle bias current is the important thing.
Use a potentiometer for a middle resistor at first, although the wiper tends to go open over the years and blow up your output transistors if you leave a pot in there.
Generally you want the output transistor idle current to be more than 10-20 ma silent on TO220 packages, and 20-40 ma on TO3 packages with enough heat sink to waste the heat safely in both cases. The lower the resistance of the spreader stack is, the more current you have to put into the bases of your output transistors and make speaker current. The limit on how low in resistance you can go is the current and power limit of your driving op amp.
That is why I am thinking of using a TDA7293 as the driving "op amp", it has a 50W limit. Dip package op amps tend to have a wattage limit of 600 mw for a 2 amp package. So with V*I the wattage, and 18 or 15 your voltage, then you don't have much current from a DIP op amp to push the base spreader stack around.
Next step from this OPA is an LM1875 amp, BTW, about $2.80 each. The stable offset voltage op amp version is LM675.
Use a potentiometer for a middle resistor at first, although the wiper tends to go open over the years and blow up your output transistors if you leave a pot in there.
Generally you want the output transistor idle current to be more than 10-20 ma silent on TO220 packages, and 20-40 ma on TO3 packages with enough heat sink to waste the heat safely in both cases. The lower the resistance of the spreader stack is, the more current you have to put into the bases of your output transistors and make speaker current. The limit on how low in resistance you can go is the current and power limit of your driving op amp.
That is why I am thinking of using a TDA7293 as the driving "op amp", it has a 50W limit. Dip package op amps tend to have a wattage limit of 600 mw for a 2 amp package. So with V*I the wattage, and 18 or 15 your voltage, then you don't have much current from a DIP op amp to push the base spreader stack around.
Next step from this OPA is an LM1875 amp, BTW, about $2.80 each. The stable offset voltage op amp version is LM675.
Last edited:
Yeah, you try a pot for the middle resistor to see what idle bias current you are getting on the emitter resistors. Then you replace the pot with a fixed resistor before closing the cover. I presume the good Dr that originated this Rutgers design calculated your top and bottom resistors to not overheat your OPA op amp.
Package of op amp matters. IF you have an old ceramic package DIP 5532, you can go lower in resistor stack because those were higher rated for power. Plastic dip packages are pretty low in wattage. The 1875/675 is a TO220-5 package with a heat tab for an external sink.
Package of op amp matters. IF you have an old ceramic package DIP 5532, you can go lower in resistor stack because those were higher rated for power. Plastic dip packages are pretty low in wattage. The 1875/675 is a TO220-5 package with a heat tab for an external sink.
30mA per rail, but i still have less than 0.6v on Vbe..
Over 30mA it happens that weird thing again...current consumption rises and rises and doesn't stop.. Why, there are NO errors in the schematic, the op amp is Mitsubishi 5220 (+-25v max) I use +-20 rails. What might be the problem.. This js pretty frustrating, and Im not soo experienced to figure it out.
Over 30mA it happens that weird thing again...current consumption rises and rises and doesn't stop.. Why, there are NO errors in the schematic, the op amp is Mitsubishi 5220 (+-25v max) I use +-20 rails. What might be the problem.. This js pretty frustrating, and Im not soo experienced to figure it out.
Put the 1 ohm emitter resistors in yet?
This is thermal runaway, the Vbe of the output transistors goes down as they get hotter and hotter, and the current runs away until they burn up.
Emitter resistors on the output transistors help. Putting the diodes of the driver stack on the heat sink of the output transistors helps- the driver stack voltage split goes down as the diodes heat up. Using output transistors with built in diodes helps even more, as the heat sink has a time lag versus the transistor die temperature that actually controls Vbe. NJL3281 Datasheet catalog
complementary NJL1302
Then some people like the famed "Vbe multipler" in the driver spreader stack, a transistor mounted on the heat sink or even the leads of the output transistors to sense the temperature and collapse the spreader voltage even more than diodes. See honey badger build document up in the sticky threads at the top of the forum to learn a bit how it works.
The one ohm emitter resistors of the Rutgers design indicate the designer knew he had a problem. A lot of the amateur amps posted on here use .22 ohm emitter resistors, which indicates the designers didn't intend to use them very hard, IMHO. Commercial PA amps that are expected to run 24/7 near maximum power, often use .5 ohm emitter resistors, plus thermally coupled (with heat conductive glue) diodes or Vbe multiplier.
This is thermal runaway, the Vbe of the output transistors goes down as they get hotter and hotter, and the current runs away until they burn up.
Emitter resistors on the output transistors help. Putting the diodes of the driver stack on the heat sink of the output transistors helps- the driver stack voltage split goes down as the diodes heat up. Using output transistors with built in diodes helps even more, as the heat sink has a time lag versus the transistor die temperature that actually controls Vbe. NJL3281 Datasheet catalog
complementary NJL1302
Then some people like the famed "Vbe multipler" in the driver spreader stack, a transistor mounted on the heat sink or even the leads of the output transistors to sense the temperature and collapse the spreader voltage even more than diodes. See honey badger build document up in the sticky threads at the top of the forum to learn a bit how it works.
The one ohm emitter resistors of the Rutgers design indicate the designer knew he had a problem. A lot of the amateur amps posted on here use .22 ohm emitter resistors, which indicates the designers didn't intend to use them very hard, IMHO. Commercial PA amps that are expected to run 24/7 near maximum power, often use .5 ohm emitter resistors, plus thermally coupled (with heat conductive glue) diodes or Vbe multiplier.
Last edited:
You're welcome.
The back diodes between speaker and rails, at the right of the diagram, intercept inductive back voltage from the speakers (which are inductors due to the coils) and keep that back voltage from acting on the output transistors and stressing their junctions. Instead the back voltage is turned into current which harmlessly bleeds off through the diodes.
You get the same thing with a solenoid valve coil, say, when you turn it off. The current tries to keep flowing, and when it can't it builds up a big voltage on the driver transistor. Since diode prices went down from $4 (for seleniium diodes) to $.25 about 1964, back diodes have been installed on almost all valve solenoid coils. Before that, they used 220 ohm 3 watt resistors parallel to 24 vdc coils.
Cool about you having a 50 v rated op amp. that is not one I've seen discussed before. The best bargain I found in 36 v op amps in stock at the distributor, was the TI/nationalsemi LM1875, which is about $2.50 in my market.
One reason the back diodes would be important, the speaker goes through a resistor to the minus input (feedback pin) of the op amp. Op amps have limited voltage rating between the In+ and the In- pins, and speaker back voltage might tend to blow it up.
The back diodes between speaker and rails, at the right of the diagram, intercept inductive back voltage from the speakers (which are inductors due to the coils) and keep that back voltage from acting on the output transistors and stressing their junctions. Instead the back voltage is turned into current which harmlessly bleeds off through the diodes.
You get the same thing with a solenoid valve coil, say, when you turn it off. The current tries to keep flowing, and when it can't it builds up a big voltage on the driver transistor. Since diode prices went down from $4 (for seleniium diodes) to $.25 about 1964, back diodes have been installed on almost all valve solenoid coils. Before that, they used 220 ohm 3 watt resistors parallel to 24 vdc coils.
Cool about you having a 50 v rated op amp. that is not one I've seen discussed before. The best bargain I found in 36 v op amps in stock at the distributor, was the TI/nationalsemi LM1875, which is about $2.50 in my market.
One reason the back diodes would be important, the speaker goes through a resistor to the minus input (feedback pin) of the op amp. Op amps have limited voltage rating between the In+ and the In- pins, and speaker back voltage might tend to blow it up.
Last edited:
- Status
- Not open for further replies.