Automated Wood Factory

[Mrchaim]

Thought you chaps might appreciate this idea - an automated wood processing factory, that takes in a stack of Wood blocks and eventually churns out all the possible results (From wood planks -> to gears).

Simply put, the idea works as follows: Use a Block Dispenser to place wood in front of a saw. the results are collected and then fed into a turn-table mounted Dispenser, which is setup to rotate and spit out it's contents between two directions. One direction is a stream of water leading to a hopper connected to a chest for storage, the other stream leads to a hopper feeding into another block dispenser, starting the entire process again.

I've not got any video recording software/nor editing software, so rather than worry about a video demonstrating it, i'll instead simply furnish you guys with a link to the save file where I've built this machine - http://www.mediafire.com/?qonk60th5kovk95

Edit:

Okay, here's the fully completed version in all it's glory. It's not quite perfect - the sorting is slightly uneven, and sometimes a few blocks get shot into odd places, but in general? it werks! and if i do say so myself, it works very well. The resource requirements for a legitimate world are probably a little hideous however, given the 5 block dispensers, 10 turn tables, 5 saws, and god knows how many gearboxes, axels and the ilk. Oh, and the gigantic scar you'll need to tear in the earth to make it. But hey! you'll find plenty of iron and coal, right?

Spoiler

Show

The save file link has also been updated to the now fully completed version. You will need to go around to the side and reset the pulser that lets the dispensers fire - but that's easy enough, just remove a piece of red-stone from the circuit, put it back, and quickly toggle the leaver on and off - Bamf, you've got wooden blocks being hacksawed.

Oh and, uh, one final piece of advice: be really careful you dont manage to drop a non-saw-able block into the machine. the results are... really annoying, shall we go with that?

[Gdnite]

Very nice, and yeah this is quite the build. Very good job =D

[gftweek]

Looks good. Have you thought about just reusing the first stage to reprocess the blocks? So the water stream from the item dispenser on the turntable leads to a piston or elevator item lifter that leads back to the original hopper.

This means you could use a single saw and block dispenser. You would also have to sort the results with filters on the hoppers though, but it should still reduce the footprint and resource requirements a lot.

I'm planning to build something similar soon, I need to move my tree farm though, so I can automate the processing of the tree and the current tree farm is in too small a location.

[finite8]

Sorry, beat you to it :)

And here it is:



Will annotate it all when i get the chance. Will put some in now though to explain briefly.

EDIT: Have annotated the Tree production process. The Milling part is on my to-do-list.

My Final Version



Does EVERYTHING from Wood, Panels, Moulding, Corners and Gears. Once it is fully supplied with Wood Planks, excess flows to the secondary processing centre where the supply is randomly fed through the same process. I think about 70% goes straight out into storage and the remainder gets fed through the process again.

[gftweek]

I still think I can simplify your tree harvester design finite8, I think the bulk of the timing circuits can be replaced with 2-3 RS-NOR latches, that trigger sequentially based on the signal received by the detector block. I just need to find some time to work on testing designs.

The item dispenser on a turntable does simplify your 25%-75% split though (not sure how you are currently doing it), by having 3 directions items are output going to one water stream and the other direction is the reprocess water stream, should be quite accurate too.

[grimper12341]

There's 3 RS-NOR latches in the replica I made of finite's design. Still needs 3 clocks tho, small fast one to power the pistons, long one to switch between up/down cycle, and a turntable for the automatic replanting of saplings. I suppose I could make the big clock bigger/slower and add an extra output for the auto replanting. Actually that might make for a smaller footprint than my current setup now that I think about it.

viewtopic.php?f=3&t=258&p=4975#p4716
viewtopic.php?f=3&t=258&p=4975#p4975

EDIT:
/facepalm
...much smaller, dunno why I didn't think of it before >.<
Would cut out an RS-NOR latch, 2 pulse generators and the turntable.

[gftweek]

By using a push down tree-harvester, you can detect when the tree is fully pushed down by seeing when the gravel block falls instead of being pushed by the piston. I use the same detector block to see when the last gravel block has disappeared into the block dispenser to replant the sapling.

With the upwards pushing harvester, you may be able to use a detector block to spot the lack of gravel stack (so last log has been pushed up), instead of the long timer, to switch between up and down cycles (only check it's output after the gravel pusher has started), but this may require a delay to avoid false positives from leaves. So timing may be simpler in this case.

[finite8]

I still think I can simplify your tree harvester design finite8, I think the bulk of the timing circuits can be replaced with 2-3 RS-NOR latches, that trigger sequentially based on the signal received by the detector block. I just need to find some time to work on testing designs.

The item dispenser on a turntable does simplify your 25%-75% split though (not sure how you are currently doing it), by having 3 directions items are output going to one water stream and the other direction is the reprocess water stream, should be quite accurate too.

Yeah, the reprocessing stream was messy as i kind of got myself stuck with limited space. Ideally, a way to establish the split would:
- Have all output from the re processor go straight into a hopper that in turn flows into an Item Dispenser.
- Have a piston / BD water stream control that flows to either storage or reprocessing.
- Use a clock to control every pulse of the Dispenser. Items are dispensed into the water flow. Based on the current clock cycle, block either the path to storage or to reprocessing. This way, you will have 100% accuracy of the ratio due to the water flow control and the fact that you are injecting only 1 item at a time.

My system simply uses the detection of the item lifting platform to control flow. When the platform is down, a block is placed forcing all produced items to flow into storage. When it is up, the block is removed and items can flow into storage. As it keeps the Block detector active approx 50% of the time, and the pistons i have underneath the saw has a 50/50 split (items are pushed to the nearest non-occupied block) i have an almost 75/25 split as when the Platform is down, one side is blocked forcing all items to flow into storage. Crude, but effective.

I found it more preferable to use pistons and a turntable than redstone circuits as it saved on a lot of space (redstone isn't so nice when adjacent to eachother. Redstone circuits are great, but the space needed always catches me out)

My design isn't too complex, it just seems it due to the fact that the wires kind of go everywhere. If i get time, i will sit down and try to draw / film it all out to verify it's complexity.

I do love the design posted here however. Would like to see a vid of it all in action (i can't be stuffed downloading the world).

[finite8]

By using a push down tree-harvester, you can detect when the tree is fully pushed down by seeing when the gravel block falls instead of being pushed by the piston. I use the same detector block to see when the last gravel block has disappeared into the block dispenser to replant the sapling.

With the upwards pushing harvester, you may be able to use a detector block to spot the lack of gravel stack (so last log has been pushed up), instead of the long timer, to switch between up and down cycles (only check it's output after the gravel pusher has started), but this may require a delay to avoid false positives from leaves. So timing may be simpler in this case.

I still don't understand how sustainable a "push down" tree harvester is as you would have to add a whole new complex process to get the gravel back up to the top to push down again. I really want to see a vid of this as it is doing my head in trying to figure out how it would work.

Trying to detect when the gravel falls back down is okay but you may find it even messier to get working. With my design, I only have one redstone line running up which synchronizes the Trunk moving process into the mill. If you want to keep the timing / clock logic down below, you would also need to run the required detector from the top down to the bottom to switch modes. You would ALSO need to modify the detection process to work by the following:
1- There is nothing after the sticky retracts... this is expected as the trunk hasn't reached us yet.
2- Now there is something after the sticky retracts... we must then have a trunk.
3- Now there is nothing after the sticky retracts. We can now assume that all trunks have passed and we are picking up and dropping gravel. Send a signal to retract the gravel.
4- Cycle has finished, so reset back to normal.

I'm not sure how to do that redstone logic in a simple way and keep the grow area unobstructed. This also runs on the assumption that a BD detects a block instantly. I have discovered scenarios where the BD has taken a second to respond. In this time, the wood may be pushed out of the way before the BD has been given a chance to detect its presence so you may not be able to rely on the results. On slow timings, this is okay but with the Saw being broken by moving trunks fixed, my system works great. I can run it at maximum speed without a hiccup and as the cycle does not rely on a BD, there are no issues. I optimized by shortening the duration of the push-up cycle to reduce the number of gravel-drop cycles. I will do a recording and time how long it takes between a tree Growing and a tree being replanted. I think you will find my system pretty damn quick now.

[grimper12341]

With the upwards pushing harvester, you may be able to use a detector block to spot the lack of gravel stack (so last log has been pushed up), instead of the long timer

...

Blah.

Ah well ish a learning experience.

[gftweek]

Yeah, there is that with the push down design, with the BD not outputting inventory when swallowed by another BD, there is no way to get the gravel back up. However my test design only uses 9 blocks of gravel (for a 4 high tree + 3 leaves, plus one before detector kicks in, plus one due to delay), and you can fill a BD with 576 blocks of gravel, so it can process 64 trees before it needs to be topped up again, will use more for pine trees as they are taller, but it's still decent.

Hmm yes, I did struggle to get the timing sorted and redstone all working, and I need to optimise it, but it seems simpler to me, since without a circuit diagram I had a hard time replicating your design exactly, so tried my own approach. When I get a chance I'll try and compress it as much as possible, but it should be smaller and use less resources (I hope), I don't think it will be any more efficient at tree gathering though, probably a bit worse if I'm using standard trees (haven't found pine yet), I should try it with birch, as like pine I think the possible trees that can grow are less in variety.

The speed of it is why I want to build something similar. I currently have a tree farm with two trees and saws next to where the trunk will grow, but it takes so long before it is allowed to grow a tree of the required size and shape, I seldom get any wood out of it that I haven't force grown with bonemeal.

Hmm, according to the wiki, birch trees may grow faster than regular oak trees too!