FAQs

According to MIL-STD-1553, the stubs are technically not part of the RTs. Therefore, the two RTs, tested individually, should each be able to pass the MIL-STD-1553B input impedance test (1000 ohms minimum, transformer coupled).

In terms of input impedance testing, if the two RTs connected together were to be considered (and therefore tested) as a single “RT,” it is highly doubtful that the two RTs taken in parallel would pass the MIL-STD-1553B input impedance spec.

Depending upon the overall electrical configuration of the main bus on a platform (e.g., an aircraft), i.e., the number of stubs, lengths, spacings, impedances, etc., it would be highly likely to be able to connect two RTs using the same stub cable and have a reliably functional system. The most likely outcome is that such a configuration would work reliably, especially for a relatively short bus, low number of stub taps, and short stub lengths.

Note however, that in terms of bus system topology, that connecting two RTs to the same stub is not compliant to MIL-STD-1553.

For a transformer-coupled (stub-coupled) RT, MIL-STD-1553B recommends (but does not require) a maximum stub length of 20 feet connecting between a bus coupling box and a single terminal (BC, RT, or monitor). The main intent of this stipulation is to minimize the stub loading on the main bus. Excessively long stubs and/or stubs terminated in low impedances can load down the main bus and result in transmission line reflections, and therefore waveform distortions. This can have the effect of increasing the bit error rate for terminals receiving data on the bus or, in extreme circumstances, cause terminals to stop receiving completely.

However, it is not all that uncommon for implementations to exceed the 25 foot recommendation. If you don’t have too many of these types of loads on a particular 1553 bus (and the bus is not too long), then the terminals on it should operate reliably.

A transformer-coupled BC or RT is required to transmit a stub voltage of 18 to 27 volts peak-to-peak on its stub, which translates to a voltage on the main bus in the range of 6.36 to 9.54 volts peak-to-peak. A direct-coupled BC or RT is required to transmit 6.0 to 9.0 volts on the bus. Using the lower (direct-coupled) number and assuming a very short data bus (i.e., no attenuation on the main bus) this results in a minimum received voltage of 4.24 volts peak-to-peak on the stub
for a transformer-coupled BC or RT.

In considering cable attenuation, MIL-STD-1553B requires that a bus network must be designed to provide a voltage between 1.0 and 14.0 volts peak-to-peak at the input to every stub-coupled terminal. The minimum stub signal of 1.0 volt corresponds to a voltage of 1.41 volts peak-to-peak on the main bus. Assuming a minimum voltage transmitter, this allows for an attenuation ratio for the main bus of 4.24 to 1 (6/1.41), or 12.6 dB. This provides for a wide margin, in terms of cable lengths as well as terminal transmitter output and receiver threshold characteristics.

MIL-STD-1553B also requires that a transformer-coupled receiver must accept (in the case of an RT, must respond to) any voltage in the range of 0.86 to 14.0 volts peak-to-peak. This provides 0.14 volts margin between the lowest terminal input voltage and the maximum receiver threshold voltage. In addition, a BC or RT is required to not accept a voltage below 0.2
volts peak-to-peak (i.e., an RT receiving a voltage below this level must not respond).

Of course, the other consideration is transmission line reflections. Reflection problems can interfere with the operation of terminals on the bus. However, assuming there is a limited number of long stubs (longer than 20 feet), the bus should  operate reliably.

For the effect of the 25 foot stub cable, refer to Figure I-1.7 on page I-16 in our MIL-STD-1553 Designer’s Guide, Sixth edition. For a transformer-coupled terminal, assuming a “1553B transformer”, increasing the stub length from 20 to 25 feet will have the effect of reducing the (worst-case) impedance looking from the bus from about 500 ohms to 400 ohms,  representing a 25% increase in the stub loading.

Note that a “1553B transformer” implies a bus coupling transformer with a minimum open circuit impedance of 3000 ohms, as specified in paragraph 4.5.1.5.1.1.1 of MIL-STD-1553B.

MIL-STD-1553B has no maximum bus length, and I have heard of instances where 1000 foot buses have operated successfully. You need to consider the following:

1. You may want to perform a bus loading analysis/simulation. That is, consider the cable attenuation (for a long bus, you can buy heavier twinax cable with lower resistance per foot, and therefore lower attenuation), the number of bus taps (stubs), and the length and spacing of the stubs. I suggest that you either simulate this or build a mock-up. However, for a 300 foot bus with a reasonable number of terminals, it is unlikely that you will see a problem.
2. You need to consider the BC response timeout time value. A 1000 foot bus results in a roundtrip time of 3 to 4 μs. With our ACE, Mini-ACE, and Enhanced Mini-ACE series bus controllers, the minimum nominal BC timeout value is 18.5 µs. In addition, you may program this parameter for higher nominal timeout values of 22.5, 50.5, or 128.0 µs.

For the impedance measurement, we (DDC) use an HP 4192 impedance analyzer. The correct voltage for measuring input impedance, per the 1553 test plan, is in the range from 1.0 to 2.0 Vrms, applied on the stub. Assuming a 15 volt transceiver and therefore a stub-coupled turns ratio of 2:1, this results in a voltage of 2.0 to 4.0 Vrms applied at receiver inputs to the 1553 terminal hybrid. In order to get a stable impedance measurement, the guard conductor from the analyzer needs to be connected to the transformer center tap on the stub side.

At one time, the noise gate was required for our older generation series terminals, the BUS-61553 (AIM-HY) and the BUS-61559 (AIM-HY’er).

 However, the need for the noise gate has been eliminated by the decoder design used in the ACE, Mini-ACE (Plus), the Enhanced Mini-ACE, and the SP’ACE. The new decoder provides improved filtering at the end of a received message, eliminating the need for the noise gate.

There are several additional requirements:

1. The stub-coupled transmitter voltage must be a minimum of 20 volts peak-to-peak (the requirement for MIL-STD-1553B is 18 volts peak-to-peak).
2. There are requirements regarding the RT address. Part of these may be satisfied by latching the RT address (from the 1760 connector) soon after power turn-on. For more details, refer to the answer to the next question.
3. A 1760 compliant RT must be able to respond on the bus within 150 ms following power turn-on. At this time, it is permitted to respond with the “Busy” status word set. This indicates that while the RT is “alive,” it is not yet able to transfer data. Alternatively, it may respond with valid data.
4. Within 500 ms following power turn-on, the RT must be responding with data as defined by the 1760 standard, with the “Busy” status bit not set. This means that the RT’s host processor must be fully up and running at this time.

5. For a bus controller, most of the MIL-STD-1760 requirements are fulfilled by means of software, rather than by hardware. One “gray area” is with regards to the 1760 requirement to transmit and verify (at the receiving end) a checksum with every message.The checksum requirement may be implemented by either hardware or software. DDC’s opinion is that it is better to fulfill this requirement in software rather than hardware. The reason for this is that a software verification provides for a complete “end-to-end” integrity verification.

   What we mean by “end-to-end” is that the checksum, if implemented in software, encompasses the operation not only of the 1553 (1760) communications interface, but also of the BC’s and RT’s host microprocessors and memory, along with the BC’s and RT’s embedded processor and firmware.

   Keep in mind that the integrity of the 1553 (1760) communication link is still checked by the BC and RT hardware, by means of the parity bit transmitted and checked with each word of every message transmitted across the bus.

Yes, There are several considerations:

1. For MIL-STD-1760, the RT address must be provided from the 1760 connector. That is, it must not be programmed by the RT’s host processor software.
2. When any of the RT address signals are connected to logic “0,” there must be a minimum of 5 mA of current in the wire to the 1760 connector. Assuming a 5V power supply, this implies the need for pull-up resistors of less than 1KΩ between +5V and each of the RT address signals.
3.  Note that for routing the RTAD signals to the 1760 connector, it is suggested that some form of ESD protection – either capacitors or clamping diodes – be used. Note that this is just a suggestion, it is not a MIL-STD-1760 requirement.
4. The RT address must be latched into the RT within 10 ms following power turn-on. In addition, it must not change after it has been latched. There are a few different methods for doing this, depending on one’s interpretation of the 1760 standard.