HOME > PRODUCT COMPARE > STATIONARY DECODER COMPARISON > DIGITRAX DS44
 

Digitrax DS44

Digitrax DS44 (buy)

General

The DS44 is the Digitrax equivalent of the North Coast Engineering Switch-It*. It is designed specifically to operate stall motor switch machines and will not work with most other types of accessories due to a limited current drive capability. The DS44 is easily the smallest stationary decoder of the group. It consists of a 0.6875" X 1.625" surface mount circuit board with approximately 7" wire leads. The wires intended for connection to the switch motors are together in a harness that plugs into the circuit card. The plug/socket arrangement is the same one that Digitrax uses on some of its engine decoders. The two wires intended to connect the unit to the track power (Red and Black) are hard wired to the circuit card.

The DS44 allows the control of up to four stall motor switch machines. Unlike some of the other stationary decoders, however, there is no provision for manual control of the switch motors, and there is no way to restore factory default addresses should you end up loosing you secret address decoder ring.

Connections

All connections to the unit are via the included 7" wire harness. On a practical basis, it is unlikely that you will have four switch motors within 14" of each other. You will likely have to either extended the length of the wires by splicing or connect the unit to a terminal block to allow longer wire connections.

Feedback

There is no provision for switch position feedback to the cab bus.

Programming

Programming can be done in the block mode or the random address mode.

In the block mode, addresses are assumed to be sequential in blocks of four. Thus, 1-4 is one block, the next is 5-8, and so on. A table of block addresses is included in the instructions. To program in the block mode, find the White wire in the switch motor harness. Connect the White wire to the Black track power wire and turn on track power. Choose the block you want to use, and use your throttle to send any one of the addresses in the block. The DS44 will be programmed to the selected block. The A output will be the first address, the B output the second, etc. Once programmed, remove the White wire and carefully stow it so that it will not contact any of the other wires.

If you want to program the unit in the random address mode, connect the White wire in the switch motor harness to the Red track power wire and turn on the track power. Use your throttle to access the four addresses you want to use. The first address sent by the throttle will be the A output address, the second will be the B output, and so forth. In this way, you can assign each decoder output to any accessory address that you desire. Once programmed, remove the White wire and carefully stow it so that it will not contact any of the other wires.

Manual

The manual consists of a double-sided printed card included with the decoder. It is adequate to allow connection and operation of the DS44, but is lacking in all but the most basic information.

Performance

The DS44 performed its intended function quite well. The stall motor switch machine ran at normal speed with good torque for moving reticent switches. The run voltage was 10.6 volts, while the stall voltage was 9.8 volts. Both of these values are quite sufficient to ensure proper operation of stall motor switch machines. The measured run current was 2.4mA while the stall current was 13.9mA. These values are sufficient to allow the use of an LED in series with the switch motor to indicate switch position. I used a red/green dual LED in my testing, and both red and green were bright and easily visible. (If you are unfamiliar with a dual LED, they are designed to illuminate one color when the current flows through in one direction, and a different color when current flows in the opposite direction. Since current through the switch machine is in opposite directions for the two positions, a dual LED can be used to indicate the position of the switch machine.)

I also tested the DS44 using a 470 ohm ¼ watt resistor in series with an LED. The LED was bright and drew approximately 16.7mA. Based on this test, the DS44 could be used to control signal lights if LED’s are used in the signaling system. Since the unit is physically small, another application possibility is to control interior, marker, and drumhead lighting in your passenger cars. All of these lights must be LED’s since the DS44 does not have sufficient current/voltage capability to operate incandescent lamps. With the four sections, a separate address could be assigned to each of the lighting circuits.

Recommendation

The DS44 performs its intended function quite well. At an MSRP of approximately $10 per switch motor, it is quite cost effective. When comparing to other stationary decoders, you must decide if the lack of manual controls and lack of direct terminal block connections is an acceptable trade-off for lower cost per switch.

Product Comparison

The Digitrax DS44 is designed as a direct competitor of the NCE Switch-It, so I did some additional testing to directly compare the capabilities of the two devices. The first note of interest is that the DS44 and the NCE Switch-It use the same microprocessor (MicroChip 12CE519) and output driver (LM324). Switch-It has one LM324 to drive two switch motors, while the DS44 has two LM324 to drive four switch motors. Since the basic stationary decoder consists of a microprocessor (to receive and decode the DCC packets) and a switch motor driver (to supply operating and stall current to the switch motor), it would appear that the DS44 and Switch-It are identical. Performance measurements indicate that this is not true, and that the Switch-It is the superior performer.

For this series of tests, I used a Digitrax DCS 100 and selected "N", "HO", or "O/G" track voltage as required.

For the first test, I used resistive loads to simulate various current conditions, and measured the temperature of the driver package for a single loaded output. Based on the measured temperature rise, I calculated the case temperature of the driver for both outputs loaded to the same current value. The results are shown below:

Unit Tested
Scale Setting
Load (ohms)
Volts Out
MA Current
Case Deg C
DS44
HO
470
9.33
19.9
45.4
HO
260
7.32
28.2
58.8
O/G
600
8.7
14.5
87.6
O/G
260
4.41
17.0
105.6
Switch-It
HO
470
9.41
20.0
35.2
HO
260
7.18
27.6
44.6
O/G
600
13.93
23.2
39.8
O/G
260
8.76
33.7
58.6

Temperatures are given in degrees Celsius. Remember that 100 degrees Celsius is the boiling point of water. The maximum chip die temperature is specified at 125 degrees Celsius. The condition marked in red for the DS44 will definitely have a die temperature in excess of the maximum allowable, while the condition above it at 87.6 degrees case temperature may exceed the maximum allowable die temperature. All conditions for the Switch-It are within the maximum allowable power dissipation and die temperature for the LM324. A word of caution, however: 58.6 degrees Celsius will feel VERY hot to the touch, possibly hot enough to cause a burn.

As a follow up test, I connected each decoder to a Tortoise™ switch machine and tried operating the machine on each scale setting. I got the following results:

Unit Tested
Scale Setting
Response
DS44
N
Slow
HO
OK
O/G
Fast with Oscillations
Switch-It
N
Adequate
HO
OK
O/G
Fast with One Overshoot

On the N setting, both units were slow, but the Switch-It was slightly better. Both worked fine on the HO setting. On O/G, both units ran faster than you would like for a slow motion machine, but the DS44 exhibited another problem. It would run hard into the end stop, the output drive would crash so that the Tortoise™ bounced off the end stop and back the way it had come, the output drive would pop back up, and the Tortoise™ would run hard into the stop again. It continued this oscillation for as long as track power was applied. This would be unacceptable operation if the unit were in control of switch points. In contrast, the Switch-It ran hard into the stop, bounced back a small amount, and then settled into stable locked stall position.

Finally, I left each unit attached to the Tortoise™ in the stall position (for the DS44, I ran it to stall on "HO" and then changed the scale setting to "O/G") and monitored the driver chip case temperature. The results are shown below:

Unit Tested
Scale Setting
Current (ma)
Case deg C
DS44
HO
13.98
32.9
O
18.2
50 (Runaway)
Switch-It
HO
13.50
28.7
O
20.68
31.3

In this test, I reported the case temperature for just one active output. The DS44 went into thermal runaway. I stopped that test after the temperature passed 50 degrees Celsius and was still climbing. Two active outputs at this point would have had the die close to its maximum allowed temperature. Thermal runaway is a condition in which the output current heats up the die, and as the die gets hotter, the current output increases, which increases the temperature, etc. The die temperature continues to increase until the unit is destroyed.

Based on the data above, it appears that the DS44 should not be used with the scale set to "O/G". It appears to work acceptably well on "N" and "HO". The Switch-It can be used on any scale setting with no problems. It also appears that the Switch-It will generally run cooler than the DS44, all other conditions being equal. Temperature is the bane of semiconductor devices. Their failure rate increases as the cube (third power) of the temperature increase. Keeping things cooler will result in more reliable operation (i.e. fewer decoder failures).

  Phone: 800-978-3472 or 802-878-5005.
The entire site © 1997-2014 by Tony's Train Exchange. All rights reserved.
Can't find it? Try Tony's Sitemap.