Mini kossel with FSRs and heated bed with RAMPS board under the bed?

I'm working on my kossel and I have it setup without the heated bed so far. Does anyone out there have a printer with this configuration? I'm not quite sure how I will construct my heated bed yet. I see two potential issues to work around:

1) I assume the FSR's need to be completely insulated from heat, so I think I will have to mount the bed heater + aluminum spreader + glass bed combo to a piece of wood and put the FSR's under the wood. I don't think I could use cardboard because if it flexes then it might not give reliable readings to the FSRs. So this is doable, but will take some work. Any other ideas?

2) If the wood (above) insulated the bed well enough, this might not be an issue but has anyone had issues having a heated bed directly above the RAMPS board? I can't be good for having everything run cool if the bed is too close. I can mount a fan or two around the ramps, or print a bracket to mount it above if I raise the bed to get some clearance. I could just move the ramps out from under the bed but I kinda like it there and if I move it then I have to build some kind of mounting place or case. Any suggestions?

Thanks for any help!

My suggestions - forget the FSR and make the machine robust enough so that calibration, once done, persists for a while. I just built a Kossel mini and found quite a few things in the design that make calibration inaccurate.
Screwing bolts into plastic is a very bad idea. It means that you cannot tighten things up fully and they will loosen quite quickly with temperature fluctuations.

Electronics failures are often related to heat. Trying to cool down electronics under a heat bed is not an optimum situation especially given that many people will be using cheap boards with underrated FETs without heatsinks.

I suggest building an enclosure for the electronics out from under the heatbed. I consider auto levelling as a bit of a flavor of the month approach that really isn't needed if one designs the machine for stability of calibration in the first place and spends the time needed for accurate calibration.

Wood is not a stable material especially near heat. It contains a lot of moisture and will warp.

Quote
Ralph.Hilton
My suggestions - forget the FSR and make the machine robust enough so that calibration, once done, persists for a while. I just built a Kossel mini and found quite a few things in the design that make calibration inaccurate.
Screwing bolts into plastic is a very bad idea. It means that you cannot tighten things up fully and they will loosen quite quickly with temperature fluctuations.

Electronics failures are often related to heat. Trying to cool down electronics under a heat bed is not an optimum situation especially given that many people will be using cheap boards with underrated FETs without heatsinks.

I suggest building an enclosure for the electronics out from under the heatbed. I consider auto levelling as a bit of a flavor of the month approach that really isn't needed if one designs the machine for stability of calibration in the first place and spends the time needed for accurate calibration.

Wood is not a stable material especially near heat. It contains a lot of moisture and will warp.

Thanks for the suggestion but I am determined that I need some type of auto leveling. I use an E3D and even if you try to always tighten the nozzle the same, there is bound to be slight differences. There is no way I want to re calibrate the height every time I change a nozzle. Using the FSRs is a great solution for this. As far as kossels not being accuarate due to screwing thru plastic... sure, its not perfect but there seems to be plenty of well calibrated kossels all over youtube. So it is working for a lot of people. However, it's looking likely that I will have to move the RAMPS board from under the bed.

While I agree to some extent on Ralph's idea of making the machine stout enough to maintain calibration (every machine should whether FSR's are used or not), it's hard to do on a printer with the stiffness of a wet noodle. While I greatly admire what Johann has done, his designs often leave a bit to be desired in terms of stiffness.

FSR's are heat sensitive, but somewhat heat tolerant.
If you use Johann's code, it will be a problem and they will need to be completely insulated. If you use FSR's with a Trinket, it will not. You will just need to reset the Trinket before you calibrate. This is because Johann's is triggered based on set numbers, the Trinket systems trigger simply based on any change of resistance.

You will be okay with FSR's mounted normal, without insulation but with a Trinket, IF you only do PLA (in which case is sort of pointless). If you plan on heating the bed high enough for ABS you really should insulate it at least some. If you only use a bit of plywood as insulation, you will slowly cook your Ramps while doing ABS. My Rostock used to get quite hot underneath and it had better airflow than your Kossel. A layer of cork makes a HUGE difference. Unfortunately, while wood is stiff and a bad insulator, cork flexes and will cause resolution loss with FSR's.

If you think that's a problem, you haven't worked with FSR's enough. Attaching FSR's is one of the biggest headaches you will face, they are a pain in the neck! They are not quite the simple solution forums have made them out to be (same for magnetic rod ends!). If you read forums they all proclaim how great they are, and they are great for calibrating, but few are telling you the other half of it, like the glass falling off, or tipping while calibrating and throwing the whole calibration out of whack. Everyone I have met with them in person has commented that while great, they aren't quite what they expected. One even said we (Griffin) were foolish even attempting to put them in a production machine, and I can't entirely say he was wrong. We have spent far more time trying to get them right than we ever expected and are only now really narrowing in on what we think will be the final solution for us. That's before we even attach a heated bed, which many on the Kickstarter wanted. While we think we can do it, we need more time for that than we had for the Kickstarter.

If you really want to go down this path, be prepared for a long fight on getting things working how you expect, even if you put the Ramps elsewhere.

I'm not quite sure what a trinket does and I really don't follow how using a different code (firmware) could possibly change how much I should insulate around the FSR's if I want to use a heated bed. Do you know of a page where I can do some reading to learn about the trinket?

Also, how is it hard to attach the FSRs? Everything I have seen says just secure them with a drop of glue. Seems pretty simple. Not trying to be smart, just want to learn. Thanks.

[www.liming.org]
This page may help