Short Ciorcuit detector doesn't work

greenspun.com : LUSENET : Wiring for DCC : One Thread

I have divided my railroad into four booster areas. Each area is divided into 10 to 12 sections. One booster area has been wired using the 1156 auto bulb in series with each section on the "hot" side. When a short occurs, the bulb for that section does not light and the whole booster area goes down. I am using Easy DCC system which I don't think should matter.

-- Peter-N Butler (cpb2@eticomm.net), July 27, 2000

Answers

Peter, you have run into a problem which is not explained by those who advocate using the light bulb. The short circuit protection in some boosters operates instantaneously upon detecting a short circuit. A light bulb takes quite a few milliseconds to get the filament hot before it limits the current flow to it's rated value. (I believe it is in the 200 millisecond range but do not have an authoritative source.) During the time the filament is heating up, the current flow is much highrer than the steady state on current. So what happens is that you get a short in your track, the current flow goes way up and before the bulb can get hot and limit the current, the booster detects the short and shuts down. (You can verify this by shorting out the short circuit detection resistor on the Easy DCC booster and causing a short on your track. The bulb will light.)

I know for sure that this is the way the Easy DCC booster works because I have one. I also know for sure that the PS II cicuit breaker that TTX used to sell (now replaced by the PS III) also works this way (I have 2 of them). I also know for sure that the Digitrax boosters (I have one) have a built in and adjustable delay for the short circuit protection and the light bulb will work as expected with them. The Digitrax PM4 short circuit protector/autoreverser should also work OK but I haven't tried it.

I do not know about other manufacturers products but I expect that boosters that will autoreverse may work as you expect with a light bulb if the autoreverse is enabled.

However, if the booster is running with a heavy load under normal conditions and you get a short, the light bulb may not prevent a shutdown even if there is a delay in short circuit detection.

Also, even with the light bulb and even if the booster does not shutdown on a short, you will notice a blink in the loco lights and maybe even a jerk in motion when a short occurs for any other loco in the same section. This happens because of the high in rush current to the cold light bulb filament. The track voltage drops dramatically until the bulb starts to heat up. It is not a long period probably 200 milliseconds or so but long enough to be perceptible.

Side Track here. Our local transit system is replacing conventional lights (stop lights, running lights, turn signals) on the their buses with LED based indicators. You can easily identify the LED devices even at considerable distance if you can see a bus that has some LEDs and some conventional lights. The LED seems to snap on and off while the conventional lights ramp up and down and there is a delay at both the on and off time. The delay is due to the finite time it takes to heat and cool the bulb filament. You also see the same phenomenon in the newer cars with the LED CHMSLs. Check it out next time you are driving. Just thought I would throw this in for those who might not believe my previous explanation for the behaviour you are seeing with your booster.

Anyway, I hope these explanations based on my experience are helpful. As to what to do about the, there is no easy answer that I have found. You can email me if you want more information about what I have done and why I did it.

Dale Gloer daleg@interlog.com or dgloer@ca.ibm.com

-- Dale Gloer (dgloer@ca.ibm.com), July 27, 2000.


Dale,

Thanks for the great response to Peter. I went back and checked to see what I had under the section on the light bulb. I know that elsewhere in my page I talk about fast boosters, but not a single bit under the section for the light bulb itself. This section was written before the Chief came out; which was the first fast booster that was encountered. I definitely need to do a bit of updating here.

If I am reading your response correctly, the EasyDCC system is also fast when compared to a 1156. Correct me if I am wrong.

EVERYONE:

This is a good time for me to ask owners of other systems, which ones are too fast for the bulb? Let me know if the system has a way to slow it's response down, like the Chief does, so that a bulb can be used. I realize that one of these days, all systems may be too fast for the bulb.

Thank you one and all.

Allan Gartner

-- Allan Gartner (WiringForDCC@augustmail.com), July 28, 2000.


Allan, you assumption is correct. The Easy DCC booster is a fast booster. Also the TTX Power Shield II electronic circuit breaker is fast. I am using the Digitrax DCS 100 (Chief system) and it has a default delay of 1/8 second which can be changed to 1/2 second. With a light railroad load, the 1/8 second setting is fine with an 1156 bulb.

I built the Easy DCC booster from a kit so I could examine the circuit and see that there is no delay in the short circuit detection. I discussed the PS II design with the designer who confirmed that it has no delay either.

Dale.

-- Dale Gloer (daleg@interlog.com), July 28, 2000.


EasyDCC booster uses a simple RC time constant to eliminate false tripping. The approximate delay is about 0.25 seconds.

This time constant can be changed to make it longer. For the Booster3, the component is C5, the 33uF capacitor. To make the delay longer (longer before tripping) make the capacitor larger.

However, there is a downside. Longer duration short circuits will result in more heating of the power transistors and a higher risk of failure.

kgg

-- Keith Gutierrez (kggcvp@aol.com), September 06, 2000.


Moderation questions? read the FAQ