FragBum

Okay I noticed that after the last Windows 10 update DCS performance has suffered however overall CPU usage is up significantly. I am currently running a 9600K build which I did about a month ago and testing shows a several % increase in CPU usage and at the same time lower DCS performance, aka this hardware is less likely to hit 90FPS in Rift. Has anyone else had a similar experience? It may or may not be the windows update but I am testing this hardware with DCS stable version so nothing has changed there and no hardware configuration changes.

You may have to experiment perhaps with a higher resolution rotary encoder something like this. Not this specific unit. Now this unit has 200ppr not 16ppr so it's like more like 14 times the pulse rate! Now if it works with the Leo Bodnar USB and it should but the higher ppr could be filtered in the 836 software or there may even a pulse rate limit within DCS, I just haven't used one with the 836. <edit> The 836 appears to set the Pin to a pull up state so you would need to have an A6AZ-CWZ3C type which would be open collector 90deg bi-phase the unit listed wouldn't work unless you could change the state to Pin, perhaps this could be achieved with the setup software, but I don't know ATM. The Pin mode can be changed at the programming level.

However looking at the Leo Bodner site rotary encoders are fairly straight forward, you connect the common GND and the *2 signal wires to 2 inputs and use the config software. Leo Bodner encoders. Pretty much any bi-phase rotary encoder should work. Use the config software to tell the USB interface that the 2 inputs you connected are in fact a rotary encoder then assign them in DCS as the up or down button repeat for units tens hundreds as required. I think that's what Mr Burns was aluding to :) *Some encoders have an additional momentary push switch that can also be used and assigned in DCS Nice pit BTW. :thumbup:

Yes you can simply invert axis, works fine and Yes the pin numbers are correct but the pin numbering on the 10 turn pot is different. :) Just to be sure the terminal numbers from the shaft end is 3,1,2 different to a "normal" pot. Red +5V, Black ground, Yellow (Green in your case) analogue in on your Leo Bodnar board. Check that PDF I linked previous and a handy picture here. Or if you prefer. Images modified for illustration from,. http://www.leobodnar.com/shop/ And https://www.bourns.com/pdfs/3590.pdf

The KA 50 engines are not made to produce thrust they drive a shaft as opposed to a jet that is designed to produce thrust. In both cases air density is a factor. :)

If you look at the picture it has a diagram of the pinouts on the body of the pot. Click on the picture to enlarge it. The pin designation is from the shaft end #3 #1 #2 where #3 and #1 are the either end of the resistance and # 2 (furthermost away from the shaft) is the wiper. you'l see the little arrow pointing towards pin#3 and the letters CW. Think #3 and #1 as the top and bottom connection of a normal pot and #2 as the wiper or center pin. For Leo Bodnar boards the connection to an analogues input would be:- +V to pin closest to the shaft end. #3. GND (-V) to the next pin #1. And ANA in to the bottom pin #2. According to the marking that should get you increasing voltage input for clockwise rotation. You can verify this with a DMM pins #3 and #1 will always be about 10Kr +/- a couple of % Specs for the pot link here. Soldering shouldn't be difficult it's a learned skill so some practise first. :D <edit> My question is can you actually use an analogue input to control the radio frequency, usually it would be a switch?

Yeah, I'm sure some of the rivets on the F/A 18 still ain't just, you know, right yet. :P :D Heaven forbid we'ed like the DCS VR engine and maps, you know stuff we bought sometime ago worked on,.. :thumbup:

You do have a Pilots Log that logs the hours in each air frame and landings and kills and those really hard landings ah crashes. :thumbup:

This about sums up where Rift is heading note the front page headlines image in the article, very telling. Luckey eh!..:music_whistling:

I so resisted to respond to this topic, I did,.. :lol: :D But for O/P it could be more that the added graphics are taxing your system and the newer A/C require more resources. Have you checked CPU and GPU usage?

And that sir becomes a major pain. I had an ear speaker fail and it become apparent just how utilitarian those built in ear speakers are when you have to use a pair of headphones in addition to the Rift, fortunately the HMD got replaced under warranty. Aside, I think Rift has lost the plot or more to it "we" aren't buying enough software to keep Mark & Co in a manner they've become accustomed to because instead of developing a better Rift they've come out with a better sales gadget. IMHO. My next VR HMD wont be a Rift. :music_whistling:

Thank you borchi_2b. Standing by. :thumbup:

Yeah 1080Ti to RTX2080Ti gave some improvement I also wouldn't think going to an RTX Titan would be worth it. Your money of course. You might get better gains with CPU/Z390 upgrade even with your existing GPU or maybe add 2080Ti. I also wouldn't rule out Ryzen 3000 or Radeon VII but we'll have to wait for new Ryzen CPUs. I'm testing a 9600K stock CPU clock build with my 2080Ti and the only aspect the 5930K@4.7Ghz build in sig, is better at, is CPU bench mark, oh and the Intel 750 series PCIe NVME SSD but only 16 PCIe lanes. I sent my mobo back for a warranty claim. :cry: The intended GPU for the 9600K build is my old 1080Ti which also tested very well with Rift. YMMV :thumbup:

I think this guy is putting something very similar to your build and I ave recently also done a 9600K build, I put my specs up on post 11. Link to another 9600K build here Long story short the 9600K build turned out pretty good in fact I have tested it with my 2080Ti and the only area that my 5930K rig beats it is CPU! My 9600K build components.

True but DCS modules will end up like steam and modules will be locked to your user name. I'd say the DRM interface is not well thought out it's impossible to tell what the state of a module is e.g. activated or deactivated when in module manager and don't start me on serial numbers when you buy say all the WW2 fighters as a "pack". Your left with a reference to the 'pack" on your ED profile page with 4 or 5 serial numbers and no way to tell which serial is for which aircraft. Starforce itself isn't that bad not taking into account the somewhat over zealous hardware limitations, for me it's the user interface. Oh and do not attempt to deactivate or activate in VR always switch to 2D screen first. :thumbup:

Same here mostly happens exiting from MP use, quit the MP session goes back to the hanger no menu and I have to use taskmanger to kill the DCS process. Also the switching between MP chat seems dodgy. I'm currently on the DCS 2.5.4.27430 stable version. For those that can't task switch you need to have another app running in background I usually have chrome running so I can task switch to that then run task manager and end DCS task.

Yes I swear at the model given that swapping out and in SSD's RAM and so forth for simply testing as I do often at least when everything goes over to user name that will be helpful. I even think I got pinged when Windows did an upgrade a couple of days ago, you can tell when DCS starts up and it asks for you user name and password again,...

And that is the point if you O/C the CPU and GPU as the OP stated they intend to do you need to allow 180W for CPU and 300W+ for GPU, that's ball park . Then you'll want some headroom on top of that. Lets add a minimum 100W for SSD's mother board and a couple of fans, so that's around the 580W mark, if your into efficiency most PSU's will be at the highest efficiency when loaded to somewhere around the 50 to 70% I wouldn't be putting a 600Watter in the build personal choice of course. ;)

Just to give the OP a bit of an idea. I recently did a 9600K build for DCS, given I already had some hardware lying around, it was a minimal spend as I wanted to include Rift. 9600K Asus Z390 Prime *32GB 2666 XMP to 2800 *1080Ti *intel 545 series 500GB SATA SSD *Noctua NH-D15 nice air cooler especially if you intend to O/C. :thumbup: 750W EVGA *Phanteks Enthoo Evolv case *stuff I had already,.. Performance on 2560*1600 was 80 to 100FPS+ with out of the box DCS settings, sorry I could only handle a minute or so of 2D :helpsmilie: but if your resolution is that or lower it should work well for you. :) Having said that it also works great in Rift VR 45 to 90FPS but suffers a bit on PG and Normandy mostly 45FPS fairly solid, except PG with weird FPS drops. IMHO You got RAM covered 16GB see how that goes going 32 will help for MP however I regularly see VRAM go to 10+GB on MP food for thought? Listen to demon re PSU's Really 2006?? For my budget it was a great build, chalangiing a bit due to glass on both sides of the case but a fun build. :D If I were to re-think, I'd look at better GPU with maybe more VRAM look at Radeon VII some feedback on this forum. link here. Although surprisingly my 1080Ti is fine at this level. :thumbup: I honestly can't recommend the 2080Ti as value for money, yes it's better but not value for money IMO. Next maybe 9700K, I don't think the 9900k is value for money same deal it's better but it's your money, right? Maybe save bit and go SATA SSD there really isn't much HDD access once DCS is loaded and the difference in load times doesn't appear to be great. Both are great choices.

Just keep in mind for Starforce DRM stuff changing RAM will cost demerit points towards your activation count. :music_whistling:

Well hind sight would have been to de-activate each DRM-ed module before you upgraded. :cry: If you go ahead a re-activate the DRM-ed module(s) it will now cost one activation for each module you activate unfortunately. You can check the status of your modules on your ED profile page.

Maybe ditch the onboard graphics or at least ditch it an check performance for DCS.

I think that is correct re DCS AFAIK however Windows 10 can spread the core load out I don't know how but I'm getting usage over 50% with DCS using about the same DCS usage hitting 35+ on peaks, total usage is now hitting over 50% (That total usage is up from maybe 40% after the latest Win update Win 9 build 1030!) . I'm currently using a 9600K 32GB build with 2080Ti and DCS 2.5.4.27430 Stable, in NTTR and Caucasus maps I'm getting 45 to 90FPS with the best looking DCS Rift graphics I've seen in DCS for some time now. I avoid the PG and Normandy maps as currently there rubbish as far as VR performance goes and are way too resource hungry and the PG map just has plain weirdness built into it. IMHO As those maps are paid content they really need some work done to them.

This is what I have from "TM 1-1520-Mi-17-10" Dept of the Army 2007 The only other difference I see is this note here also from the same publication. I'll just add I am not claiming one or the other to be right or wrong just that there is a difference everything else agrees as to operation, I only included that detail as an observation of available data. :)

Great job, having proper inputs makes the immersion that much better. :thumbup: I'm a big fan of the low cost motor cycle steering dampers for such tasks as collective and pedals. For those that can do the mechanical side there are after market USB controllers available if the thought of programming scares you from such a project.