It’s usually the thickness of a single layer. 0.5mm is about twice what I use as an offset.
PVA has a good print temperature of around 190 - 200C max. If you try to run it at 255C it will quickly swell, tar up, and clog your nozzle. Purging at such a wide variance in temperature will likely be a disaster as any residue of PVA left in the nozzle when you switch back will caramelize and jam.
I have seen traces of filament of a previous color show up randomly in a print 30 min after a good purge with another. I have a nozzle dedicated to PVA for these reasons.
If I use the gap between support layers and model trick I put the gap at 0.05 more than my layer thickness. Works like a charm for me on single head extruders.
@Jeff_Parish I’m currently using S3D and unfortunately it only gives the option for number of layers between support and part but I assume I’ll have to switch over to the Prusa edition of Slic3r for the multi material upgrade.
Would using a 3rd material to transition work? I’ve got a ton of MakerGeeks PLA that prints at 235 C. It would just be a bit of a nuisance to add in the 3rd color.
Purging with an intermediate material MIGHT work, but I would still be worried about residual crud stuck in nooks and crannies in the hot end.
It sounds like a number of people wih Y splitters took a very long time to get the two filaments to switch out without one clogging the nozzle. Each filament switch will impose a retraction time and a heat up or cool down time. Adding time to a layer is not exactly what people with 3D printers like to do…unless they are getting paid for it.
Not to be rude @NathanielStenzel but none of your advise has really been helpful, and some of it just plain false (support Z offset). Most of it was criticizing the Y switching setup and judging from your comments you haven’t even had any experience with it. I understand switching materials/colors with it is going to cause print time to go up, but in my opinion having the ability to use multiple filaments is worth it, and I still have the ability to use just a single spool if I choose. I also believe it’s far better than a dual nozzle setup and all the calibration difficulty it causes, not to mention the loss of build area. Besides you usually need an ooze shield anyway which would also increase print time. Please try to keep further comments on topic of support material for PETG filament. By the way PETG is a horrible support for itself because of how well it sticks to itself. That’s not to say you can’t print PETG with supports, it just makes cleanup that much more difficult.
@Adam_Steinmark hey, have you not heard of the problems people have had with Y adapters? At any rate, you asked for opinions and you got opinions. 
Read through this post and try telling me that I am the only person saying that a Y adapter with the those two materials was a bad idea.
@NathanielStenzel Let’s please stay on topic. Yes, the consensus was made that PVA would not work as a good support with a Y splitter setup, but we’re still discussing the possibility of using HIPS or a 3rd transitionary material. The difference was you badmouthed the entire concept of a Y splitter, ignoring the fact I asked for opinions on support material for PETG.
Most of the issues people are having with Y splitters is with improper retraction leaving a blob at the end of the filament, causing the filament to jam when it’s automatically re-fed. Prusa delayed shipping their kit to fix this specific issue. I’d really like to discuss the question I asked, and not the pros and cons of a Y splitter setup.
I googled it, found PETG edge and scaffold from E3D and nothing else. Nothing. I am not sure if people used the combo in a Y splitter. I think your chances of using stuff s3ems more likely if your thermistor gives a higher temperature than what it really is. Edge seems to print at a lower temperature than you mentioned, but that is just based on product description. 240C for Edge and 215C for Scaffold. Scaffold is 45…whatever monetary unit that is for 500g of filament. You would have to ask them if that combo would work with a Y splitter.
You could wait for new filament to come out, print PETG at a temperature that makes it behave poorly or dope one spool of it using a filament oiler/dyer going into the Y splitter (I have no idea if that would work).
@Adam_Steinmark You also might not find a water soluble option at these print temperatures. If you do please let me know about it. Anyway, other possible options…
HIPS might be an option as was mentioned before but as you already questioned I’m not sure of the side effects of Limonene on PETG when dissolving it out. There is some evidence that it weakens ABS but I’ve not verified that.
What if you used ABS as a support for the PETG? They are close to the same melt temperature. PETG doesn’t dissolve in acetone (as far as I’m aware) while ABS does. Still not water soluble but neither is HIPS so other than a different chemical you’re basically in the same boat either way. Not sure if you can get PETG to lay down on top of ABS or HIPS.
@Jeff_Parish I saw one post from a guy asking about ABS as a support material for nylon. He did not get a solid yes or no.
Nylon, ABS and PETG seem to be the only things in that temperature range. Feel free to mention others. Maybe looking for support materials for nylon is a possible way to find your answer.
@NathanielStenzel You could add HIPS to your list. 
Same dead end with support material for Nylon so far. PVA and HIPS are both mentioned with HIPS having a slight edge due to the more compatible print temperatures. Nothing on using ABS other than your mention above.
Every water soluble support material I find has the same temperature limits and caramelizes above 215 C.
Maybe with the right tweaks (speed, layer thickness, etc.) ABS will work?
@Jeff_Parish I did not make it up. https://www.reddit.com/r/3Dprinting/comments/52ms60/using_abs_as_dissolvable_support_material_with/ As I said, he did not get a solid answer.
Matching support materials is hard. Lots of chemistry involved in getting the right amount of stick and compatible build plate / chamber temps. Adding the additional complexity of running them through a single nozzle is piling up the difficulty.
HIPS and ABS is the only GOOD option I can think of for single-nozzle printing because they’re damn near the same material so they’re fully miscible in the melt pool (good purging) and can print at the same temp.
And yes, limonene does weaken ABS when dissolving the HIPS. How much it weakens it depends greatly on the ABS blend, specifically how high the acrylonitrile content is. No way to tell except by experimenting.
Oh, and pro tip… when somebody posts that they’re going to try something and then never follow up with results, that almost always means it didn’t work well.
@NathanielStenzel Didn’t think you did. It just didn’t show up for me in the searches. Your reference to it was the only one I’d seen so far. All that tells me is that it is not a commonly mentioned/used solution. Either not fully explored (more likely) or has some major drawbacks that are not being reported (less likely as they would be mentioned I would think).
I’m afraid a high temperature water soluble support material available to our crowd is not yet avaialble. Stratisys SR materials might be higher temperature but I don’t know much about their printing properties.
Stratasys soluble support filament pairs great with ABS in separate nozzles… not sure about same nozzle. It’s not strictly water-soluble though, you need to add their expensive-ass tablets or a particular type of dishwashing power. Some kind of base, I forget what offhand though.
Yep, the caustic detergent they add is what leaves that white powdery coating on your parts if they are not VERY thoroughly rinsed off, drained, washed, rinsed, repeat, repeat.
Is it possible to use it in our machines? No doubt expensive. From your comment it sounds like you have some experience with it?
@Jeff_Parish a regular poster on the 3D Printing Tips, Tricks, and Reviews Google group (Enginwiz) has experimented with it a lot while figuring out how to run a Mojo more cost-effectively.
By the way, I made a separate thread for the idea of a PETG only or PETG with filament dyeing/oiling attachment solutions that I hope may get useful replies. You know…since Adam can’t handle thinking about alternative options on his thread.