I have a very odd situation with one of my control panels that has 3 - DS64s on it and a pair of crossover turnouts still on a manual toggle switch. The problem has existed since I built the panel and I finally have decided to figure out what is happening. The only reason for the toggle swicth is that I needed one more output and didn't want to add another DS64 for a single output.
The problem is that when I flip the toggle switch to throw the turnouts, all of the DS64 outputs on all 3 DS64s go to zero volts for a second and then come back on. It causes all of the indicator lights to go out and come back on. There is nothing in common with the toggle switch and the DS64s except the 12V supply. I've tried powering the DS64s off both the 12V supply and track power with the same issue. I've monitored the 12V at AX1/AX2 when I flip the toggle switch and nothing changes. The toggle switch is simply a DPDT switch cross wired to provide 12V to a pair of Tortoises with 2 LEDs in series for panel indicators.
The oddest part is trying to figure out what would cause all of the outputs on the DS64 to go to zero volts (they don't flip in the opposite direction/polarity) momentarily. The DS64 inputs have nothing connected on any of them. I've tried both track power and AX1/AX2 power and the same thing occurs.
It hasn't caused an operational problem because the lights come back on and all of the turnout are controlled by Tortoises. I am scratching my head to understand what would cause the outputs to drop like this. I've monitored Loconet with Decoder Pro and there are no signals being generated when I flip the toggle switch for the manually controlled turounts.
Any thoughts ?
Engineer Jeff NS Nut Visit my layout at: http://www.thebinks.com/trains/
Perhaps your toggle is make before break so it momentarily shorts out when you flip it. As a test try the DS64's on track power and the 12v supply only for the toggle. If you have an analog meter available, you could also turn off all power and connect the meter in ohms mode to the input of the toggle then flip it. A digital meter won;t react fast enough, but an analog meter you'd see the needle jump if there is a short as the toggle is thrown.
I'll lay odds that's what the problem is though.
--Randy
Modeling the Reading Railroad in the 1950's
Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.
I agree with both of you but I am pretty sure that I removed the 12V power off of AX1/AX2 last night and it still did it, while running off track power. I'll get the analog meter out and do some more testing tonight. Another option I have considered is adding a 47 ohm resistor in series with the input side of the toggle switch, from the power supply. My power supply has auto shutdown in case of over current conditions. It may be fast enough to cause a quick glitch on the DS64 power feeds and then restore power when the short is removed. It's based upon an LM338K adjustable voltage regulator. The 47 ohm resistor would reduce the amount of current during the suspected short, which would not interfere with the normal Tortoise operation (would drop 1.5V for 2 Tortoises) but would keep a dead short from drawing enough current to trip the over current condition. I'll poke around more with it tonight and let you know what I find. It may end up being a defective toggle switch. I've used these same switches on my other panel with no issues.
Well the mystery deepens but I was able to find a tactical fix. I pulled out the analog VOM and measured the input voltage at AX1/AX2 while throwing the toggle switch. The needle was dead still but the panel indicator lights on all 12 DS64 controlled Tortoises still went off and back on. Next I tried removing AX1/AX2 and running off track power. The same result. Then I went the other way and tried to run everything off 12V. Same thing. I placed a 10 ohm 10W resistor at the input of AX1/AX2 just to ensure that the power supply would supply at least 1A+ of current without dropping the voltage. It stayed right at 13V.
So next I tried removing the output from the toggle switch to the panel LEDs, which are in series with the 2 Tortoises. This would tell me if there was a short in the switch, since 12V was still connected to the switch, just no load. I really didn't think this was the problem, since the 12V at AX1/AX2 didn't appear to be dropping. When I took the load off, the DS64 outputs did not drop but my 2 turnouts were now not working.
Next, I inserted a 100 ohm resistor in series from the toggle switch to the LEDs. I threw the toggle switch, the Tortoises moved and the lights did not blink on the DS64 outputs.
So the mystery is what is causing all outputs across all DS64s to drop ? It doesn't appear to be the input voltage on AX1/AX2 and the problem still occurs, even when the DS64s run off track power.
Are you SURE there is no common wiring anywhere? Common ground runs, anything liek that? Does that power supply run ANYTHIGN but the DS64's and that one toggle switch? Like say, UP5's or other components? Are your LEDs all wires with 2 wires, or did you connect either the anodes or cathodes in common? The clue is disconnecting the LED from the toggle - the DS64's then worked but disconnected the LED stopped the Tortoise from working? Are they just series LEDs with the Tortoise motors? You've got a common connection there somewhere, either with the LEDs or the Tortoises.
rrinker Are you SURE there is no common wiring anywhere? Common ground runs, anything liek that? Does that power supply run ANYTHIGN but the DS64's and that one toggle switch? Like say, UP5's or other components? Are your LEDs all wires with 2 wires, or did you connect either the anodes or cathodes in common? The clue is disconnecting the LED from the toggle - the DS64's then worked but disconnected the LED stopped the Tortoise from working? Are they just series LEDs with the Tortoise motors? You've got a common connection there somewhere, either with the LEDs or the Tortoises. --Randy
Yes, the LEDs are just in series with the Tortoises. I am going to trace some wires tonight on the 12V bus to see what else it may be powering. It might be powering the UP5 panel near the control panel. However, the question is still, regardless of what it is powering, why would it cause all DS64 outputs to drop ? Especially when the input voltage isn't moving and track power is still applied to the DS64s, even if I pull AX1/AX2. Nothing I can think of should cause the DS64s to drop their outputs. There isn't even a command you can issue via Loconet to cause this.
Tonight I'll trace the toggle switch outputs to the Tortoises and the 12V power bus to see what else might be running off of it.
A suggestion I saw on another thread was to just run the LED's off of the relay contacts built into the tortoises.
Springfield PA
jbinkley60 rrinker Are you SURE there is no common wiring anywhere? Common ground runs, anything liek that? Does that power supply run ANYTHIGN but the DS64's and that one toggle switch? Like say, UP5's or other components? Are your LEDs all wires with 2 wires, or did you connect either the anodes or cathodes in common? The clue is disconnecting the LED from the toggle - the DS64's then worked but disconnected the LED stopped the Tortoise from working? Are they just series LEDs with the Tortoise motors? You've got a common connection there somewhere, either with the LEDs or the Tortoises. --Randy Yes, the LEDs are just in series with the Tortoises. I am going to trace some wires tonight on the 12V bus to see what else it may be powering. It might be powering the UP5 panel near the control panel. However, the question is still, regardless of what it is powering, why would it cause all DS64 outputs to drop ? Especially when the input voltage isn't moving and track power is still applied to the DS64s, even if I pull AX1/AX2. Nothing I can think of should cause the DS64s to drop their outputs. There isn't even a command you can issue via Loconet to cause this. Tonight I'll trace the toggle switch outputs to the Tortoises and the 12V power bus to see what else might be running off of it.
Tonight I began testing again. First I bypassed the resistor to ensure the problem returned and it did. Next I checked to see what else is powered off the 12V that feeds to DS64s and the answer is nothing. It only powers this panel. Next I tried a different power supply. Completely different technology and the like. The problem remained. I did some more testing including removing all Loconet connections, trying different combinations of track power vs. 12VDC and still nothing. Next I tried adding filtering capacitors to the 12V input to the panel. I tried everything from a .1uF up through a 220uF and still nothing.
Finally I tried a .01uF capacitor across the toggle switch outputs and suddenly the lights were not going out every time. They were going off about once out of every 3 times. I increased to a .1uF mylar capacitor and bingo, no more lights dropping on the DS64 outputs. Your initial inclination would likely be that an induced signal is being fed via the 12V power from the toggle switch to the DS64s. However, the problem occurs when the DS64s are powered from the track and there is no connection to Ax1/Ax2.
So, let the theories begin. My guess is that I have all of the outputs bundled together in a wire bundle with the 12V power feed. I suspect a signal is being induced across the DS64 output(s), from the toggle switch, which is causing this. I could cut apart the wire bundle and bypass the capacitor to be sure. Here's a picture:
The larger red/white twisted wires going to the 4 conductor European barrier block is the 12V feed coming from the power supply.
I find it hard to believe though that a single low current (it IS driving a Tortoise after all) 12V 'spike' in the same budnelof wires as the outputs from a bunch of DS64's is goign to make ALL DS64 outputs suddenly turn off. Not flip state, but actually power OFF. A spike or sage at the 12V power input to the DS64 could do this - and might not be visible without a scope, even an analog meter would be too slow. However, if the 12V supply was completely isolated fromt eh DS64's that woudl stop that from happening. The cap would stop it - much the same way that there's usually a .01uf cap per every so many logic gates in a circuit - the gates changing states induce sags ont he power input to the chip, the capacitor stops that and keeps the sag from propogatign to other chips in the circuit. Are you using any button inputs to the DS64s that might share a common ground with the 12V supply?
-Randy
rrinker I find it hard to believe though that a single low current (it IS driving a Tortoise after all) 12V 'spike' in the same budnelof wires as the outputs from a bunch of DS64's is goign to make ALL DS64 outputs suddenly turn off. Not flip state, but actually power OFF. A spike or sage at the 12V power input to the DS64 could do this - and might not be visible without a scope, even an analog meter would be too slow. However, if the 12V supply was completely isolated fromt eh DS64's that woudl stop that from happening. The cap would stop it - much the same way that there's usually a .01uf cap per every so many logic gates in a circuit - the gates changing states induce sags ont he power input to the chip, the capacitor stops that and keeps the sag from propogatign to other chips in the circuit. Are you using any button inputs to the DS64s that might share a common ground with the 12V supply? -Randy
I agree that it doesn't make sense. I have a scope. Maybe I will drag it out to continue this academic approach. No, nothing in common with button inputs to the DS64s because there are no buttons on this panel.
Here's the front of the panel.
The only thing on the panel are the LEDs from the DS64 outputs, 4 blocks power toggle switches and the toggle switch in the bottom right for the crossover.
All DS64s are controlled via the throttle. I may go back and add buttons for the yard routes in the future. No, isolating the 12V supply from the DS64s did not stop it from happening. That's ths strange part. Powering them off track power and having the 12V just power the toggle switch did not stop the problem. Do rememebr though that the induced voltage coming back from the Tortoises can be much larger than 12V, even if just for an instance. Again, I agree this sounds very odd but it's my theory for now.