0

I'm troubleshooting my Roland TR-909 drum machine, which doesn't power on anymore. All of the voltages come up fine, but the reset line stays low. My suspicion is one or more of the components of the reset circuit have failed. Edit: the trace connecting D702 and D701 is highZ and C704 in the same area looks like it's leaked some electrolyte, which probably caused an open on the trace.

reset circuit

My understanding after staring at it for a little while is as follows:

At first I thought D701's purpose in life is to bring the drive voltage of Q701 down below 5V so as not to forward bias it's Vbc, but the excess voltage would still be dropped across R702.

Question 1: What is D701's purpose? (Note the schem was revised from a 6.8V to a 5.6V zener at some point)

Question 2: Why not just drive Q701 from the 7805 directly by removing D701, and connecting the output of the 7805 to where the anode of D701 used to be?

Q701/Q702 and surrounding resistors function as an inverter use output is high when power is applied. When power is not applied, the transistors are off, the 7805 is off, and it's output is low, pulling reset low via R705.

R705 and C701 set an RC time constant such that RESET goes high roughly 30ms (6 time constants) after 5V is up.

Question 2: Why is an inverter needed here? Wouldn't the circuit work easily as well with just R705 and C701?

mhz
  • 1,010
  • 2
  • 9
  • 16

2 Answers2

1

The first thing that may help you understand what is going on is to put a scope on C704 and see how long it takes for the voltage to reach equilibrium. I'm going to guess that it takes several cycles.

The cause of this voltage ramp is the winding resistance of the power transformer.

Now think about what the function of the reset circuit is. You want to hold the controller in reset while the 5V rail is stabilizing. The 5V rail ramps up as the voltage on C704 is ramping up. Starts off at zero and ramps up to 5V.

By holding the reset RC network discharged until the 5V rail is approximately stable, you ensure a clean power-up for the controller.

That's what the reset circuit does. Q701 is OFF until the voltage on C704 reaches about 6.2V. This allows Q702 to hold the reset RC timing network discharged.

When the voltage on C704 reaches about 6.2V, Q701 turns ON. In actual fact, it probably turns ON and OFF once or twice but the reset RC timing network doesn't let the controller start.

The 5V rail is fairly close to 5V when Q701 begins to turn ON. The drop-out voltage for a 7805 is typically about 1.5V. When C704 reaches 6.2V, the 5V rail is sitting somewhere near 4.7V. This is safely above the 4.5V minimum that most 5V devices will operate at.

Dwayne Reid
  • 23,792
  • 2
  • 38
  • 70
0

The intent of the zener diode and the two transistors is to keep reset low until the voltage at the input of regulator IC 703 is adequate for the circuitry to operate. Since D701 is a 6.8V zener and it will take a volt or so to turn on Q701 this means at least 7.8V at IC701 which is enough to guarantee that the +5V rail is up to voltage. R705/C701 add another few milliseconds to that to ensure the circuitry is stable before the reset is released.

A simple RC time constant would not guarantee this and although may work would not be as good.

kevin

Kevin White
  • 33,153
  • 1
  • 48
  • 78
  • "A simple RC time constant would not guarantee this and although may work would not be as good."

    Wouldn't you be able to guarantee it with a large enough time constant to cover any other variabilities in the PS circuit?

     – mhz Apr 05 '15 at 22:06
  • If the AC input voltage was low (for example you had the transformer setting incorrect) the RC reset would release the reset to the circuitry even though its 5 volt supply might only be 3 or 4 volts. Enough for the circuity to do something but not operate correctly. – Kevin White Apr 06 '15 at 00:52