Page 6 of 12

Re: IPGregory Build and Progress Thread

Posted: Mon Nov 28, 2022 1:15 am
by IPGregory
Thx Uwe, I will check that. I've gone through every other setting from TB 270 but didn't try manually setting the handshaking. Windows is fully up to date on updates and it's a Gig onboard NIC with the latest Dell Drivers. If the Acorn only needs 100Mbs at Half Duplex, I'll turn it down to that.

Re: IPGregory Build and Progress Thread

Posted: Mon Nov 28, 2022 2:03 am
by suntravel
That depends on the networkchip, I would try different settings.

Starting Acorn first, then the PC helps also in some cases.

Uwe

Re: IPGregory Build and Progress Thread

Posted: Mon Nov 28, 2022 2:18 am
by IPGregory
Good to know. I will try some different settings and see what happens.

Since I am still heavily in set up and config mode, I am starting and stopping acorn quite a bit, but the PC normally stays on through the resets, etc. so that might be a factor. I'll try and see.

Thanks again for the help.

Re: IPGregory Build and Progress Thread

Posted: Thu Dec 01, 2022 3:11 pm
by IPGregory
Away on a trip at the moment so haven't had chance to work on the comms issues above yet. Will update when I do.

In the meantime I noticed that V4.82 has been released. Not sure what the 'release candidate' status means though? Is this a type of beta or something? Is this something I should be updating to, or should I avoid and stay on v4.80 until I have my machine up and running in a stable condition?

Re: IPGregory Build and Progress Thread

Posted: Thu Dec 01, 2022 3:18 pm
by suntravel
It is a minor bugfix, and some new features.

If have updated today, I would in your case also update and spend not more time configuring 4.80

Uwe

Re: IPGregory Build and Progress Thread

Posted: Thu Dec 01, 2022 3:21 pm
by IPGregory
Thx Uwe, I will grab a copy then.

Re: IPGregory Build and Progress Thread

Posted: Thu Dec 01, 2022 8:28 pm
by ShawnM
A quick question for you, are those vertical black plates on either side of the gantry made from aluminum or steel? It appears you are using those as “flags” for your prox switches. If they are aluminum they won’t be accurate. Prox switches need ferrous metal for accurate detection. I know you mentioned something about moving them but wherever you put them use a steel flag like the Z axis if you want repeatability.

Is every wire that carries a signal shielded and grounded? Spindle wire shielded an grounded?

As Uwe stated, the NIC can cause issues, is it an Intel NIC?

Re: IPGregory Build and Progress Thread

Posted: Thu Dec 01, 2022 8:58 pm
by IPGregory
Hi Shawn, the whole sensor setup is up for change. Its working and I am using it for set up, etc. but it's not how I want it long term for robustness and 'adjustability'. I need some new brackets and some sheet steel to do it again properly. The side brackets are aluminum. Good to know on the sensors working better with ferrous. I'll keep that in mind when I re-do them. I have auto homing running great, but I don't see the point in setting up Auto Axis Squaring until I re-do it and can accurately measure and set the setback of the Main Axis Sensor vs the Paired Axis Sensor, etc.

The spindle cable is braid shielded and bonded at both ends. I have confirmed the ground connection in the spindle and checked continuity of ground from spindle shaft to board ground bus. The Encoder cables from the motors are braid shielded and bonded at the drive ends. The step/dir. cables are braid shielded and bonded at the drive ends even though they are only a few inches long. The Spindle analogue cable is foil pair shielded and is bonded at the Acorn end. The Prox sensor cables are braid shielded and bonded at the board end. All of the bonds go to the central ground bus. The motor power cables are braid shielded cable but not bonded. I have a ground wire from the bus ran to the machine but haven't connected it. I thought about it and am not sure it won't create a potential ground loop with the spindle being grounded.

I have the Eth cable pulled out of the board and direct to the PC at the moment to eliminate any noise from the comms problem. I am pretty sure it's a PC or cable problem but haven't had chance to try any of the items Uwe suggested yet as I am away at the moment. Not sure on the brand of the NIC for the same reason. Since its onboard, I presume it's an Intel chip but will check when I get back.

Re: IPGregory Build and Progress Thread

Posted: Thu Dec 01, 2022 9:37 pm
by IPGregory
Correction, 3 of the motor power cables are factory Stepperonline cables and are unshielded. The 4th is a cable I added to extend the Chinese motor short cable. It has a shield, but I didn't bond it as the other 3 aren't. All 4 cables are separately routed in the board from the encoder and step/dir. cables and don't go past the acorn.

Re: IPGregory Build and Progress Thread

Posted: Fri Dec 02, 2022 12:03 am
by ShawnM
Good to know. When you get back you can dig deeper.

Not to mislead you, they do make proximity switches specifically designed for non-ferrous metals like aluminum or even brass but the cheap Chinese ones that we use are designed for ferrous metals. Aluminum will trip them but I wouldn't trust it for repeatability, that's all I wanted to mention.