For those of you with at least a little soldering knowledge, here is the solutionLS (LSI) 7083/7183They do all the work, and provide direct up/down pulse from any dual output incremental encoders.Just hook the two enc output to them, and use their two outputs for your hardware (FSBUS, IO, EPIC, whatever).7083 has normal mode and 4x mode, while 7183 has also 2x mode (which is suitable for my encoders that have a detent in the middle of a phase).Actually i find only one reseller carring themwww.usdigital.comFarnell, rs, mouser, elfa, distrelec all do not have those ICs.If anyone knows any european reseller, i'm all ears!Shipping them all over the atlantic is costly :(

Sweetness .. this stuff is super.Someone wants to order a pack of those 7183's and sell them on to the others ? This way they are a lot cheaper. Or should I order from them ?I use ALPS at the moment and I wire them up to SHIFT and a letter to get lower case letter one way, capital letter the other. It works though I must not turn to fast or I end up with capital letter all the time.I think with this I could turn as fast as I want ?

Depends on what kind of ALPS do you have.If you have the "knitter" type, outputting a different pulse per rotation direction; then these chips are no use.If you have a standard rotary encoder (leading phase) then these will do the trick.They also have the option to modify the pulse lenght (called TOW in the docs)

I have once opened one of mine and closed it back with succes and made a 3D max drawing of the inside. Looked fairly simple.What they do, I'm sure is:close A, close B, open A, open B, close A, close B, ..... one wayclose B, close A, open B, open A, close B, close A, ..... the other wayAnd that's exactly what those little chips want as input as I can see from the diagram on their site. I WANT THOSE ! :(As far as I remember from ordring the ALPS, I have 9 of them.

Well, you're halfway there.The Chips from USDigital work great.. they work even better if you use an optical encoder.There's one minor catch though.... the output from the 7083/7183 chips is roughly a 20 uS pulse (20 Microseconds). That's way too short for other circuitry to capture, such as an EPIC system which scans for line transitions every 20 MILLI-seconds, or a Keyboard encoder module which I think also scans every 20 mS. You can fit a lot of 20uS pulses inside a 20 mS timeframe... all of which will be missed by the scanning circuitry.You will need to stretch that pulse into something more substantial, so you could use a 555 timer configured as a one-shot or a 74LS221 dual one-shot.I also have a zillion of the 7083 chips. Ray

At least the do the quadrature, enlonging the pulse should be a fairly easy task compared to that.Anyway the 7183 have the possibility to define the pulse by resistor.Which LS chip do you have?If you have 7183 you may add some to my package...I'll mail you.

You can define the pulse width by resistor, but even maxed out, it's still not long enough to be captured by support circuitry.I've forgotten what I paid for them. I'll go home tonight after work and take a small inventory, then let you know :-)

Thank you, i'll wait for the info.

Hello,Philippe it seems to me that you have implemented Kevin Darling's idea of using differenial rotary encoders (one contact wired to letter and another to SHIFT key). I'm almost decided to do the same, but I'm not sure what exact type of encoder should I buy (Panasonic ones used by Kevin are not available in Poland where I live). I have found in Elfa catalogue ALPS encoder type EC16B-24204 (www.elfa.se/pdf/35/03584745.pdf). Is it the same type you are using in your cockpit ? The second question I have is - how fast can you turn without getting capital letter errors ? (I would like to use these encoders for my MCP to be used as HDG, CRS, SPD, ALT and V/S rotaries) ?Pawel

Hello Pawel,The encoder you point out is exactly what I have. Thnx for this btw cuz I lost the type number and it is not written on them. I needed that to order more if I chose to get the same type.Ok there are 2 things to say 'bout these:1) it is an advantage to use the "shift" + "key" system for the simple reason:for the whole pack of rotary encoders you'll use you need 1 key input per rotary + 1 "shift" which you wire up to each rotary.this means if you'd have 20 encoders, this would need 21 key inputs.if you'd have other knitters you'd need 40 key inputs !2) it is a quite big disadvantage that you have to mind your rotation speed for the thing not to screw up. That's it.Now, the question you ask "how fast for it to work well" .. that's tricky. "Fast" is relative ;)Lets keep it this simple:It's been a while since I've used them. I have unhoocked my Hagstrom card and was waiting for an other interface to arrive. Now I got my Photon but need to wire everything up again + write my software for it. So as far as I can remember and give you an indication, I'd say take about 1.5 to 2 seconds per 360

Philippe, thank you for your quick response, and I am glad that I've reminded you the type number :)Now back to encoder turning speed. Doing 360

Whow stop!h

Wow - thank youYou saved my keyboard controller (and maybe even my pc mainboard) :)Now I know how to connect these rotaries, so I can continue to hack my USB keyboard. Because of tiny controller contacts placed very close to each other (1 mm gap between them) it is extremely hard to connect them to wires, but PMDG 737 is definitely worth it :) cheers,Pawel Kobryn

Hi PawelWe guys from Poland have to support each other:) I wonder how many of use flight simmers are in this 40 million people's country:)I suggest you visiting Leo Lacava's page:http://www.betainnovations.com/He just released his Cyclotron module for reliable decoding inputs from rotaries. I know it's not very cheap (especially if you're paid in zlotys and not in Dollars or Euros) but I think it's worth to consider.Leo is very friendly guy, you can find many of his excellent posts on this forum.Piotr [email protected]