AMC1280USB Motion Simulator Controller

The AMT-103V sensors are magnetic, and their resolution is 2048 CPR. The encoder2position modules could handle with no issues the 10000CPR quadrature encoders on real servomotors with no losing pulses in highest speed 500mm/s!!!

https://www.cui.com/product/resource/amt10.pdf

I'm sure its pull-up signal issue of your encoders or something. This is the reason I prefer the AMT-103V as they are pure TTL digital output with no rise or fall curves. Also its easy to add a 3D printed Axle-adapter to mount them on your actuators. I don't see how is going to be very hard.

Also on the back of the Encoder2Position modules, there are some jumpers marked as 1-B-0 and 1-A-0... make sure these are still jumped as B-0 and A-0. And make sure you are only connecting the A,B,Gnd,+5v...

The 1-B and 1-A jumper position are for differential connections using both A+,A-,B+ and B- connections.



Also I think the rise fall timing drops for the encoders you have since they are not rated TTL only but have range from 5v to 24v !! This is based on a similar optical encoder I seen on Amazon... I bet they won't lose pulses if powered with 24v! But don't! The encoder2position modules only use 5v sensors...
www.amazon.com/Encoder-Incremental-Rotary-Length-Measurement/dp/B0756WJTP7

If you have specs sheet of your encoders please share it to find out how to remedy this.

Hu @Thanos thanks for update will check the Jumpers when i get home later , the spec of the device is the same as the ones shown in your link and by all accounts it should do the Job ,

I will fit Pull up resistors to my Test encoder2pos and hope this gets rid of this problem .. it will boil down to me missing something in my setup but i can handle the embarrassment if i get this to finally work as it was intended .. i do remember that i had to fit pull up resistor when i was doing trials years ago with Kangeroo but thought that was kangeroo related

NO... these on Amazon link are NOT TTL OUTPUT encoders... they will not perform the same as true TTL encoders (like AMT-103v or HEDSxxxx). They are using analog photodiodes directly with no correction circuitry for filtering with Schmitt trigger for digital conversion.
https://en.m.wikipedia.org/wiki/Schmitt_trigger

PLEASE SHOW ME HOW DO YOU WIRED THEM EXACTLY! Drawing or photos with some labels on the wiring perhaps... did you connect +5v? Gnd? A, B? Where exactly? Did you connect pull up resistors? Where and what value...

Ps. If you want my opinion, these encoders will never work right with 5v, no matter what pull up value you try. They have slope rise and fall pulses shape, not clean square 90 degrees shift pulses and will confuse the encoder2position modules.

Hi @Thanos many thanks for this latest info , and I can see that some incorrect assumptions were made on my part regarding Optical encoder selection. This is good news Technically but bad for the wallet LOL

So in for a penny - In for a pound as they say , i'll organize getting a sample of the recommended Encoder and repeat these tests at that time

I will post my wiring diagram later as its on another PC , Again assumption on my part but the wiring was done as per the connection diagram ref Tindie Page info , the A + B encoder Outputs to Encoder2pos PCB , 0v and +5v shared for the encoder and the Limit switches High and Low .. I also confirmed they worked using the ARDUINO Monitor Program showed Pulse count increment and decrement with Encoder rotation and Limit switches being activated also shown on this Monitor Program .

Because I was able to do Calibration with repeatable results, using the AMC128USB and Sabertooth , I was looking elsewhere for a reason why I was loosing Position ...these tips you have give, are just what I needed as tis guy tends to get tunnel vision when setting things up.

Many thanks Thanos ...Now to hunt down a Uk supplier for the new sensors.. Google is My Friend LOL

Hi Again Guys this is my wiring for the test Actuator using Optical encoder with Encoder2Pos Interface Pcb Sabertooth & AMC128USB barebones setup

One obvious mistake I see on the wiring on the photos, is that you wired the B- not the B+ !!!
You need the A+, B+ connected if you are using the 8-pin connector!!

You can leave the A- and B- unconnected.

Hi, @Thanos, I just noticed there is a FW3.5 rev1b but I can not find much info about it, what are the differences of rev1a? You know I use AMC1280USB for sabertooth's, so I guess I need FW upgrade to rev1b. Does it have a benefit to use that? Can you point me to the description?

The Rev.1b has some fixes specifically for the sabertooth's, like for example it disables the automatic "Kill motors" function that would make the motors jump a bit while saving parameters on the LCD menu. Also it removes some delays to allow control of the sabertooth signals continuously, to avoid runaway motors. Give it a try, if you have any issues, let me know.

Thanks
Thanos

Hi @Thanos thanks for sorting that mistake as case of Cant see the wood for the trees syndrome, so your fresh eyes noticed something I thought was ok .. ordered 2 of the amt1o3/v sensors from digikey UK so hopefully they will arrive before I go to USA to visit family .. cherrs and thanks for putting me on the right track . Can' wait to try these .. will need to 3d printer a special housing as there are few places to mount this type on the actuators .. but I'll sort it

Hi @Thanos Sensors Ordered so should be here on Monday .. weird thing ! ordered them on UK digikey site and dam things are still being shipped form USA ???? any ways have found a solution to mounting these when they get here sett attached sketch of 3d printed parts ( 2 elements . top & bottom ) with 2 bearings inside to support shaft to which encoder will be fitted as shown and at one end and Toothed gear at the other end Plus a bridge piece as adjustable actuator mount The gear will be clamped to shaft and positioned to minimise any axial end float which may affect encoder.. dam Cad PC at home playing up so this will be my 1st Fusion 360 test to design and Create STL file for printing .

Thank god for 3D printers now... easy to make custom parts at home!

More boys toys, and have to admit this 3D printer does the jobs nicely, only downside wish it was quicker ( yeah I know complain complain, complain ) but 2K + to get faster speeds Hmmmmm i think ill go make tea and watch a movie or 2 during long prints, but it does add flexibility that the CNC router doesn't so Im loving this Age were in and my Boys toys Collection .. Once all sorted and operation verified i'll post Stl Files if anyone is interested in using this setup

Sometimes my rig becomes jerky when it's descending slowly and doesn't have much weight load on the legs. Its better when I use the PID P value relatively small, about 10, but in that case I lost tiny details of motion. So I use P15-20 for racing games and P10 for flying games. I think it happens because of backlash, so maybe it's normal, but im curious if I can do something more to reduce that? Maybe to set something with sabertooth.
I made a video to show you this situation, I move FlyPT sliders slowly when you see jerkying.
The PID settings are P15 I0 D0 (I and D values don't affect anithing in my case ...I tried many values but nothing changed)
This behavior is the same with FlyPT, HeXpod, or direct simtools control.
At the end of that video I'm setting the home postion on the AMC1280USB board, you can see it jerks too.
Now I changed the gravity center of rotation at Y axis 150mm to front to get more load the legs, it helped much.
Does anybode have an idea what I can do more?

Because wormgears are sliding ,its possible they make little jumps turning slow.

Agree. As stated in other thread worms have higher friction and therefore can "stick" at slow speeds. Once the power is great enough to overcome friction the worm breaks loose but then it may surge. Due to backlash this surging can cause an oscillation to begin. The only way I know of fixing this is mechanically not electrically.

Yes, I thought the same. Much more comfortable when the gravity center point is 150mm on the front and I sit on it.
What else could I do mechanicaly?

Gas spring, bungee, etc. or new gearbox.

Looks bad.
How much LPF you have on it?

Also it’s important you do this test loaded

Once the rig is unloaded the weight is not Centered

Edit:

Jump in the rig and while testing increase the amount of LPF until it desapears.

Do not use simtools washout

Post How much it will be?

I'm sure it's a mechanical issue as @BlazinH and @adgun said. Watch the video from 2:35 you can see when I'm decreasing start position on the AMC board directly and motors are jerky too.
But I will make an attempt with LPF and of course jump on it

Not sure. My rig jerks also when empty while changing the start pos. (No filtering). Not as much as yours but still.

Edit: anlog or digital encoders ?