Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Dariush]

After 6 hours of wasted time, I've finally gotten an infinite oscillator that runs on nothing but water and pumps to work!
Z1:

Z2:

In case it isn't clear (and I know it isn't), here a scheme and an explanation:
Z0:

Code: [Select]

######
#^####
######
######
####_#
######Z1:

Code: [Select]

######
#.%%X#
######
######
#+%%.#
######Z2:

Code: [Select]

______
_._*._
_%=*%_
_%_|%_
_._*._
______X = floodgate
+ = drawbridge (not retracting one!)
|,= = axles
# = wall
_ = floor
. = open space
% = pump (general direction is clockwise)
* = gear assemblies
^ = pressure plate connected to the floodgate and the bridge and whatever the oscillator needs to power
Initially, water is placed on the pressure plate on Z0, then the power is connected. Floodgate on Z1 opens and the bridge raises, pumps on Z1 pump water into the tile behind the opened floodgate, where it is sucked up by pumps on Z2 and dumped back onto Z0. The water isn't pumped until 100 steps pass and the bridge lowers, after which it is pumped back onto the pressure plate. The whole process repeats. I'm planning to use this to control a bridge that will be repeatedly raised and lowered to crush enemies, since this oscillator provides the needed 100-step delay. Hope this helps someone!

[DrKillPatient]

Can it be modular? Say, putting another pressure plate and floodgate set might allow you to delay for 200?

[Dariush]

Yes, if you connect only one plate out of two to the needed object, and by extending the design to the left and right (putting more pumps on the Z1) it can be extended even farther than that (if you are willing to ignore the veeeery small delay introduced by water moving through pumps).

[Komus]

I like! This ones going in the scrapbook

[Dariush]

Uh... a bit more testing revealed that the channels on Z2 must be walled around or the water spills. I'll edit the design in first post.

[Neyvn]

So what we have here is a method to make a repeater that's effective and independent of Dwarven actions after its been activated, correct?

[Lav]

First, doesn't this system lose water with time?

Second, do you need to connect the target device to just one or both pressure plates to achieve a 100-tick delay?

[Dariush]

Well, a bit more testing and several more hours later I discovered that the current design pretty much sucks. No, it doesn't lose water, but once both floodgates are open they never close since water never leaves pressure plates for 100 steps. To get 100-step delay you have to connect both plates, but since I'm remaking this, that's pretty much irrelevant. Wait for updates in first post.

[Dariush]

All right, new version completed around 100 loops and continues working perfectly. The only slight drawback is that the use of drawbridge makes FPS cry in anguish, but I can't see any alternative.

[Lav]

Why not just build a classic mini-repeater? Two pumps and two pressure plates, nothing else, each plate sends ON and OFF signals intermittently every 100 ticks.

[Dariush]

The pumps will suck up the water before it had the chance to trigger the plates, and even if the plates were somehow triggered, the water will never stay off them for 100 steps.

[Lav]

What does it matter if water does not stay somewhere for 100 steps? What is important is that pressure plates send a signal every 100 steps.

And if your pumps are sucking water before pressure plates have a chance to trigger, this simply means you've made a mistake somewhere. :-)

This is how mini-repeater looks from the side. Pressure plate A is connected to a gear feeding power to pump A, and pressure plate B is connected to a gear feeding power to pump B.



This repeater schematic works flawlessly and requires the absolute minimum of work and power (it can actually be fed with a single windmill at 40 wind or by two windmills at 20 wind).

[Dariush]

The whole cycle on your schematic will be completed practically instantly (maybe a couple of steps to account for water moving), while my goal was to introduce 100-step delay in the loop, and that was the hardest part.

[Lav]

Absolutely wrong. This repeater is working with the frequency of pressure plate recovery, which is exactly 100 steps.

Really, man, this design and it's many variants have been in extensive use since 40d days.

Just visit the Wiki and check everything for yourself. :-)

Actually I found the wiki link describing 2-step repeater.

[Dariush]

Holy shit. I didn't know that pressure plates don't recover instantly.