APPLICATION NOTE 74
18 of 39
PGMEN is the active low input to activate the pullup switch. If not connected, PGMEN will be held high
through R3 to avoid unwanted activation of the pullup switch. The voltage converter can be controlled in
three ways: a) hard-wired for continuous operation, b) activated by an external signal, or c) permanently
shut down. Case a) is intended for applications which never require a strong 5V pullup. If there is no
control signal available from the master and strong 5V pullup as well as EPROM programming is
required, then a mechanical switch can be used to switch between case a) and c). Case b) offers the most
flexibility. For EPROM programming SHDN needs to be high, for strong 5V pullup it should be low. If
the voltage converter is shut down, L1 and D6 together with a conducting Q5 provide the required low-
impedance path to 5V.
If the signal PGMEN becomes active (i.e., is low) then the voltage at the gate of Q4 rises from 0V to
approximately 5V, causing Q4 to conduct. This is equivalent to feeding a low level from PGMEN to the
gate of Q5. The capacitor C1 between gate and drain of Q4 slows down the rise and fall of Q4’s gate-
source voltage and therefore determines the ramp rate of the programming pulse. As soon as the gate
voltage of Q5 changes from the quiescent state of 5 or 12V (depending on mode of operation) to 0V, the
P-channel transistor becomes conducting and pulls the 1-Wire bus either to 12V (if the voltage converter
is on) or to 5V, bypassing R1.
From the strictly logical point of view, the double inversion (Q3, Q4) is unnecessary. The reasons for this
circuit are to convert from a TTL-level system to a 12V system (5 volts on the gate of Q5 is not sufficient
to turn the transistor off if the voltage converter is running), to avoid high voltage feedback from the
voltage converter through R5 to the TTL-level control input and to extend the rise and fall time of the
12V programming pulse to the required minimum of 5 µs. High voltage feedback from the 1-Wire bus to
the receive input of the microprocessor is avoided by the diode D5, which becomes conducting only when
the voltage on the 1-Wire bus is lower than 5V.
The monolithic voltage converter IC1 requires L1, C2 and D6 for operation. It is activated by a low level
at its TTL-compatible input SHDN . The right choice of L1, D6 and C2 is essential for reliable operation.
D6 is a Schottky diode, recommended part number 1N5818, C2 is a low ESR tantalum capacitor of 10
µF. L1 must be a low ESR device between 20 to 100 µH, capable of withstanding current peaks of
approximately 0.5 A without magnetic saturation. To avoid EMI problems, L1 should be a pot-core or
toroid type; a rod core type is not recommended. For further details on the voltage converter and its
external components please refer to the appropriate data sheet and application notes. The LT1109 is just
one example of available parts. Other manufacturer’s components or modules can be used as well.
The duration of the programming pulse (pulse width of PGMEN ) or the strong pullup is determined by
software. Program examples are given later in this document.
D. 5V RS232 R/W All (Voltage Converter)
The universal upgrade of the interface of Figure 10 is shown in Figure 12. The components Q1, R1, D1,
D2, Q2, R2, D3 and their functions are the same as before. Additional requirements include the control
signal RTS with the diode D4 for ESD protection, the diodes D5, D7, the cascaded inverters R3, Q3, R4
and Q4 & R5, C2, the pullup switch Q5 and the controlled voltage converter IC1 with its external
components L1, D6 and C2.
RTS is used to activate the pullup switch Q5. DTR may act as power supply, if it is able to source 5V at
25 mA. The strong pullup is made possible by L1 and D6 if the voltage converter is shut down. If idle
(i.e., the bit controlling RTS in the Modem Control Register of the UART is set to 1), RTS will be at 5V.
To activate the pullup switch, the RTS bit in the UART (see Figure 8) must be cleared to 0. This will