gokhan
Junior Member
Posts: 89
|
Post by gokhan on Dec 2, 2016 14:47:25 GMT
You have to watch on + 5v if you also have parasites.At home, there are also parasites on the + 5v.I think it should also filter, or isolate the power + 5v USB.Filter 5V power use a self-powered USB hub?I am trying to redo the connections at the moment with shielded cables,Soon my tests is your VFD box shielded too ? I will check 5v with osc. VFD box is not shielded, VFDs are in it and all the cable connections are open pin connections without shield.
|
|
gokhan
Junior Member
Posts: 89
|
Post by gokhan on Dec 2, 2016 14:56:27 GMT
You have to watch on + 5v if you also have parasites.At home, there are also parasites on the + 5v.I think it should also filter, or isolate the power + 5v USB.Filter 5V power use a self-powered USB hub?I am trying to redo the connections at the moment with shielded cables,Soon my tests is your VFD box shielded too ? I will check 5v with osc. VFD box is not shielded, VFDs are in it and all the cable connections are open pin connections without shield. I tested 5v and the noice is more than the Vout of the sensors. Is it really a problem or Osc. picks up everyhting even if it doesn' t effect the performance that much? here is the video.
|
|
|
Post by tronicgr on Dec 2, 2016 16:31:50 GMT
I will check 5v with osc. VFD box is not shielded, VFDs are in it and all the cable connections are open pin connections without shield. I tested 5v and the noice is more than the Vout of the sensors. Is it really a problem or Osc. picks up everyhting even if it doesn' t effect the performance that much? here is the video. You should try filter the noise from the 5v. This looks bad!! On the amc1280usb board there are two lines of 5v and Gnd. You can attach on them a few electrolytic capacitors to cover a wide range of noise: 0.1uF 1uF 10uF 100uF 1000uF Watch the polarity of electrolytic caps to be correct. You can solder them on female headers and plug them one at a time to find the best combination with less or no noise.
|
|
|
Post by paulg100 on Dec 2, 2016 18:29:17 GMT
"I also found out that the main reason for my low freq oscillation was min speed setting. It was set to 2, now 5. It stopped oscillation."
Are you sure a min motor speed of 5 is not going to make the platform lag vs motion cue?
You have the exact same issues and similar oscope readings as me, sorry I havnt had a chance to makes some videos to compare, I havnt forgotten, I just have to wait till xmas vacation in a few weeks so I have some good time to experiment.
I have some CAT 7 cable waiting to test
Did you have the 0.1uf caps on in those videos? those make a big difference in smoothing the signal.
|
|
gokhan
Junior Member
Posts: 89
|
Post by gokhan on Dec 3, 2016 12:34:09 GMT
I tested 5v and the noice is more than the Vout of the sensors. Is it really a problem or Osc. picks up everyhting even if it doesn' t effect the performance that much? here is the video. You should try filter the noise from the 5v. This looks bad!! On the amc1280usb board there are two lines of 5v and Gnd. You can attach on them a few electrolytic capacitors to cover a wide range of noise: 0.1uF 1uF 10uF 100uF 1000uF Watch the polarity of electrolytic caps to be correct. You can solder them on female headers and plug them one at a time to find the best combination with less or no noise. I did as Thanos described. I tried 0.1, 0.22, 1, 10, 100 micro farad electrolytic condensators one by one. No significant effect. 1000 mf killed the power so I didn' t try it again. Strangely, noice dissappears if I connect the osc to the 5v and ground at the extra power pins near the usb connector. Noice increases if I connect the osc to the sensor pins on the AMC board. And maximizes at the back of the sensor connectors on the wall of the box. here is the video. I don't want to occupy the forum more on this. I am not sure if this is a critical issue? I may still try an external 5v power supply
|
|
|
Post by riton39 on Dec 5, 2016 19:29:34 GMT
Hello I started changing my wiring. I found several things to reduce the noise. Check the mass chassis, mass of the sockets, USB cables, at home, very bad ... 24v power supply: (small Chinese power supply) The ground is not grounded (24V ground) If you connect earth to ground 24v it is much better! Less noise! The ground 5V is not connected to ground 24v It must be connected. I will give you my next tests;)
|
|
|
Post by riton39 on Dec 5, 2016 20:15:37 GMT
|
|
gokhan
Junior Member
Posts: 89
|
Post by gokhan on Dec 5, 2016 20:53:12 GMT
So you connect 5v and 24 grounds together and all the noice stops? Is that correct?
|
|
|
Post by riton39 on Dec 5, 2016 21:34:47 GMT
yes , and connect at the earth / frame of the box
There is a bad mass on the USB "frame", check on your USB cables. The 24 V ground is not grounded to the 24V power supply
There is no ground connection between the AMC64 card and the expansion card 6 DOf (to check on the AMC 128, but I think it is identical)
Normally the ground must be grounded
|
|
|
Post by clyevo on Dec 7, 2016 4:43:01 GMT
You should try filter the noise from the 5v. This looks bad!! On the amc1280usb board there are two lines of 5v and Gnd. You can attach on them a few electrolytic capacitors to cover a wide range of noise: 0.1uF 1uF 10uF 100uF 1000uF Watch the polarity of electrolytic caps to be correct. You can solder them on female headers and plug them one at a time to find the best combination with less or no noise. is it still safe to try out 1000uf since gokhan is experiencing shorting out? So you connect 5v and 24 grounds together and all the noice stops? Is that correct? Did you try connecting the ground of the 24v and 5v to earth like riton39 suggested. I always wanted to try this but worried of ground loop and the leaking current might burn out the power source. I am not sure of how it will happen but i am no electronic engineer so it is just one of my silly thoughts. Can someone explain to me how is it always safe to connect dc ground to earth ground? Thanks.
|
|
|
Post by riton39 on Dec 7, 2016 20:14:46 GMT
www.maltep.com/fr/c9-la-masse-%C3%A9lectriqueI think you'll find information in English with Google. The earthing of the masses is "conventional" on the devices. The metal covers are connected to ground, and these surfaces are connected to earth to protect people in the event of isolation faults, for example It also allows to shield the cables, components ... cage of faraday. If the masses are not common, there may be problems, noise ....
|
|
|
Post by tronicgr on Dec 7, 2016 21:46:31 GMT
You should try filter the noise from the 5v. This looks bad!! On the amc1280usb board there are two lines of 5v and Gnd. You can attach on them a few electrolytic capacitors to cover a wide range of noise: 0.1uF 1uF 10uF 100uF 1000uF Watch the polarity of electrolytic caps to be correct. You can solder them on female headers and plug them one at a time to find the best combination with less or no noise. is it still safe to try out 1000uf since gokhan is experiencing shorting out? So you connect 5v and 24 grounds together and all the noice stops? Is that correct? Did you try connecting the ground of the 24v and 5v to earth like riton39 suggested. I always wanted to try this but worried of ground loop and the leaking current might burn out the power source. I am not sure of how it will happen but i am no electronic engineer so it is just one of my silly thoughts. Can someone explain to me how is it always safe to connect dc ground to earth ground? Thanks. I used up to 2200uF before, to stabilize the 5v for a small motor that was powered via the 5v of the USB. You should have no problems unless your USB port on the computer side cannot handle that much capacitance (some usb ports turn off the power if they see too deep voltage drop during plugin). You should always connect all ground together. Some computers do not have proper ground, for example laptops where their psu connects to wall outlet with only two legs... So the only other way to get good ground for the AMC1280USB is using the ground connection of the VFDs or other motor drives that actually connect to ground of mains (mass). Thanks Thanos
|
|
gokhan
Junior Member
Posts: 89
|
Post by gokhan on Dec 9, 2016 16:31:06 GMT
Well, I decided to recable everything. I started with the AMC box, then encoder cables, then inverter cables and then control cabling in the inverter box. At this stage I was soldering the tips of the speed control cables "puff" something blew out!.. I forgat 24 dc power connected on the extension board and touching the speed wire with grounded solder gun shorted the extension board Fortunately I had spares. I replaced everything and completed the recabling finally with the shielded power cables. Now every cable is shielded and every chasis and shield is properly grounded to the same ground bar. I also tried 1000mF electrolytic condensator without loosing the usb power, but it still doesn't filter the noice. It doesn't look too bad but the noice still exists, so I don't know what to think. Here is the latest video of noice on 5v and one of the encoder Vout, after all recabling.
|
|
|
Post by riton39 on Dec 9, 2016 19:19:04 GMT
- Grounding the shielding of the mini USB connector of the AMC 64 (it was not there)
-Replacement of the USB cable by a gold plated double shield. (The old cables had a very bad ground connection ... tested with 2 other cables)
-GND 24v Power Grounding
Cables internal shielding:
-Alim 24v
-Potentiometer
-Analog signals: variator speed and direction
-RJ45 female connector on all links
-shielded motor cable
At the moment I have cables networks recovery with a simple shielding, I wait for my Cat 7 SSTP.
It remains for me to shield and fix the small end of cable that goes from the potar to the RJ45 on the engines
The difference is important, the simulator moves almost more.
The AMC 64 does not have the firmware of the 128, and the dead zone in the center is present, which I think is the cause of my problem now ....
The PID looks to fit differently, but I have not tested too much to say more.
I think then test condensators on the inputs of the potar.
And see if closed box I still have an improvement.
In any case I am very happy with the current result, I still think improve the thing until almost perfect.
I think the result in games is going to be really interesting!
|
|
|
Post by riton39 on Dec 9, 2016 23:13:53 GMT
can I see a picture of your AMC box and the wiring? best quality image to see in details the wiring You must have a problem with wiring, if I manage to reduce my parasites, you must also be able to reduce your noise what is the division with Volts in your oscilloscope ? 0.5v ? 1 V
|
|