UC300ETH: UCBB Vs UB1 breakout board ?

[routercnc]

I am considering a control system overhaul.

The PC which controls my CNC machine has been having boot up issues over the past week or so. It dates from around 2008 and was pretty old then. A while ago I had overheating problems (my fault on the CPU install) which I then fixed, but recently it won't get started into winXP (the background and some icons appear but that is it) and often trips back around to rebooting. Sometimes I get a blue screen full of lots of text and it lists out the n4_disp.dll which is the graphics driver. I can't see how the file became corrupted as it is not online and is just used to load gcode from a memory stick. As they are pretty cheap I've just dropped a replacement NVidia FX5200 graphics card (used, but working) into the PC but it still has the same problem. I could try a windows repair disc in case it really is a corrupted file, but I am wondering if this is an opportunity. Even more so as my Wife seems vaguely supportive!

Given the age of the machine, the fact I'm using the parallel port into a cheap / free BoB, and have various work-arounds in the control cabinet to prevent false trips, I think it is time for an overhaul.

So skipping past USB and onto Ethernet has led me to Cs-labs or UCCNC. UCCNC is winning at the moment but there is a lot of information to gather up to make an informed choice, especially for the breakout boards. I have been skim reading manuals and forums to get up to speed. The single UCSB looks too small, and the UCBB is much better and not much more money and has more connections. Even better there is an offer from CNCdrive at the moment for opt 1:

opt1:
UC300ETH + UCBB + UCCNC software for £170 (offer until 31st Aug 2021)

opt2:
UC300ETH + UB1 + UCCNC software for £289 [ note that UB1 is made by a 3rd party, although all reports say they work well together ]

So it comes down to UCBB Vs UB1. I like the compact form of the UB1 and the screwless terminals, and may pay a bit of a premium for that, but is there any other functional difference ? I'm thinking in/outs, speed control etc. ?

Thanks

[JAZZCNC]

So skipping past USB and onto Ethernet has led me to Cs-labs or UCCNC. UCCNC is winning at the moment but there is a lot of information to gather up to make an informed choice, especially for the breakout boards.

No brainer in terms of which Software is the better, UCCNC wins hands down compared to the Cslabs options. In terms of electronics and neatness then Cslabs takes some beating as its quality, that said the UC300 and UCBB are very well made also and certainly up to the task just not quite as refined or neat as Cslabs.



I have been skim reading manuals and forums to get up to speed. The single UCSB looks too small, and the UCBB is much better and not much more money and has more connections. Even better there is an offer from CNCdrive at the moment for opt 1:

opt1:
UC300ETH + UCBB + UCCNC software for £170 (offer until 31st Aug 2021)

opt2:
UC300ETH + UB1 + UCCNC software for £289 [ note that UB1 is made by a 3rd party, although all reports say they work well together ]

Careful here because you need to factor in UK VAT and postage. You will get charged VAT and a fee from the shipping agent which I believe CncDrive uses FedEx now as the shipping agent they used before no longer delivers to Uk since we left the EU.


So it comes down to UCBB Vs UB1. I like the compact form of the UB1 and the screwless terminals, and may pay a bit of a premium for that, but is there any other functional difference ? I'm thinking in/outs, speed control etc. ?

I'm probably one of the few that can answer this with any certainty because I use both setups.!
The UB1 is the better board for a few reasons BUT not the connectors which in my opinion actually spoil the board as they are small and not very nice to use.

The UB1 wins because it's neater with no messing around with ribbon cables, etc.
Another reason is that it provides a protected 0-10V Analog Output which the UCBB doesn't have. If you need 0-10V then you have to take it directly from the UC300 pins.!!
Also, It only needs a 24Vdc PSU, not 5v and 24Vdc PSU's. The UC300 needs 5V and feeds the UCBB with 5V thru pin 26 of the IDC cable. This is a little crazy to me because the UCBB needs 24Vdc and actually generates a 5Vdc which makes it available on some of its terminals so why not feed the UC300 and do away with a 5Vdc PSU.? Which is exactly what the UB1 does.!

The UCBB is a nice well made board but it's not as clean n tidy as the UB1 and you have the other factors mentioned. That said I do use them both and keep the UCBB in stock, which I don't the UB1, but this is mostly down to cost. The UB1 features are nice but £100+ saving with no real loss of features can't be ignored.!

You won't be disappointed with either.

However, If you don't need loads of I/O then I'd give the AXBB-E a good look over because they are a fantastic neat little package, I fit these to nearly every machine I build these days except machines with ATC, etc that require lots of I/O.
In performance terms, it's exactly the same motion controller as the UC300, just with fewer ports. It also has a 0-10V analog output, unlike the UCBB.

Also, If you don't want to buy direct from Cncdrive then I have both in stock and can provide the license file for UCCNC.

[routercnc]

Thank you Dean, that was very helpful.




Careful hear because you need to factor in UK VAT and postage. You will get charged VAT and a fee from the shipping agent which I believe CncDrive uses FedEx now as the shipping agent they used before no longer delivers to Uk since we left the EU.


Ah yes, good to know !




However, If you don't need loads of I/O then I'd give the AXBB-E a good look over because they are a fantastic neat little package, I fit these to nearly every machine I build these days except machines with ATC, etc that require lots of I/O.
In performance terms, it's exactly the same motion controller as the UC300, just with fewer ports. It also has a 0-10V analog output, unlike the UCBB.


I did look at the AXBB briefly but didn't go into the detail. Looks interesting. I need 5 axis control (2X, 1Y, 2Z) and this offers 4 by default. It says another 2 can use configurable I/O pins to make 6 in total. I'm trying to work out if I lose too much by doing this and am going through the manual.

Aside from the steppers the other inputs/outputs on my CNC machine are:

2off prox for homing X axis both sides
1off prox for homing Y axis
1off prox for homing Z axis

1off limit prox on X axis
1off limit prox on Y axis
(no limits on Z)

2 estop switches

Probe input

Pendant controller, but I don't really use it so not a big deal if it can't be used

In/out not required:
I have a direct manual start/stop switch and pot for the VFD control mounted on the main cabinet panel, direct to the VFD. I don't really go for software/BoB control for these but that is my personal choice and so is not really needed on the BoB.

Also no plan for ATC any time soon so that is not required for I/O


AXBB-E also appears to have an additional connector for something like the UCSB to give more in/out, but I would like to see if the above requirements can be met without it.



Also, If you don't want to buy direct from Cncdrive then I have both in stock and can provide the license file for UCCNC.

OK, good to know. I need to check the above, then will email for all in price. Thanks

[Ger21]

Regarding the connectors on the UB1, I use wire ferrules for 20 gauge wire that are .059" diameter, and they are a tight, but perfect fit. Not sure what the equivalent size would be for you guys over there.

[routercnc]

Ok thanks Gerry. If I go UB1 route I'll bear that in mind.

For the AXBB-E - reading around a bit and my concerns about how to drive 5 axes appear to be justified. The manual suggests the additional axes (anything more than 4) can use the spare I/O channels. But these are isolated outputs which can't drive step/dir ( I believe).

This means I would need to add a ribbon cable connector to the front of the unit, and adding an expansion board such as UCSB or UCBB to a second board to get access to the other non-isolated outputs for the 5th axis. I think this means I am no better off than using the UC300ETH + UCBB in terms of space and general layout.

Does anyone have a different experience of getting more than 4 axes out of the AXBB as otherwise it looks like a great little unit ?

There is the naughty approach of pairing up step/dir to the 2 Zaxis drivers, but of course I would never do that . . .

[Kitwn]

I can't see why the isolated outputs of the AXBB-E cannot drive step and direction inputs. These are normally optocouplers much like the input of an SSR as shown on page 22 of the manual. If you use a 5V supply then no resistor should be required.

[routercnc]

I can't see why the isolated outputs of the AXBB-E cannot drive step and direction inputs. These are normally optocouplers much like the input of an SSR as shown on page 22 of the manual. If you use a 5V supply then no resistor should be required.

Thank you Kitwin. OK they were probably not supplying 5V to the isolated outputs so it was not working for them. Looks like that option is back on again. I’m away from the PC so when I get back I’ll look at that page and get it clear in my mind.

On a general note how are other people provisioning for when multiple supplies are required such as lots of 24V or lots of 5V. My method has always been a single cable from the BOB then once into the conduit solder on additional cables to supply each device. Seems like there must be a neater way. I may go for terminal blocks and link them together with a clip in bus bar (Dinkle supply some nice purpose made DIN terminals which can be linked). I guess space on the BOB means they never supply repeated 24V or 5V outputs ?

[dazp1976]

Regarding AXBB-E. People over on cnczone haven't been able to add 5th or 6th axis straight onto the additional IO fitted to the board. They are mainly all designed for switching. The way to do it is via port#3 with an additional UCSB or other bob.

I have the UC300eth with a UCBB. This setup limits you to 16 out and 18 in through 2 ports. You need an extra IDC16 cable and a terminal block for use of analoge port. Or you can plug directly with dupont cabling/connectors.
I prefer to use seperate components because if 1 fails it's cheaper to replace.

UB1 uses 3 ports + the analoge as standard giving more IO over the UCBB.
For the other 2 ports (4+5) there is the UD1 daughterboard. My view is that if something fails it will be a large cost to replace it again.

[routercnc]

OK thank you Daz. That was my concern. Having read all the manuals, lots of searching and the comments here it is all making sense, so thanks to all the comments so far it has really helped. Each of the options has pros and cons and I'm trying to get together the neatest overall solution. More to think about . . .

[JAZZCNC]

Regarding AXBB-E. People over on cnczone haven't been able to add 5th or 6th axis straight onto the additional IO fitted to the board. They are mainly all designed for switching. The way to do it is via port#3 with an additional UCSB or other bob.

That's because it's not designed to do that so why they are trying is beyond me, don't they think if it could then CncDrive would promote that it as 5 or 6 Axis onboard.? Only port 3 provides the extra I/O needed and they clearly state that only 8 pins are wired for Step/Dir on the board in the manual.

To me, the AXBB-E is perfect if you have a machine that only requires the standard I/O it offers but with the added bonus that if at a future time you need a little more then you have the option to add-on.
If you know you need lots of I/O from the start then go with the UC300 + BOB option.

Also, another plus for the UB1 which can't believe I forgot mention was the fact it offers differential outputs which can be a BIG deal for some setups, esp if using Servos.!

[routercnc]

I can see the attraction of the AXBB-E for 4 axis systems with modest I/O, plus the inclusion of analogue outputs to control the VFD speed. It even has a jumper to allow either 5V or 10V range for the speed. I think for the majority of hobby users it would be a perfect little unit.

But I need 5 stepper motor control, so the extra space for the additional board just to driver the 5th axis makes it similar to the UC300ETH+UCBB in terms of footprint. More thinking . . . .

[dazp1976]

I can see the attraction of the AXBB-E for 4 axis systems with modest I/O, plus the inclusion of analogue outputs to control the VFD speed. It even has a jumper to allow either 5V or 10V range for the speed. I think for the majority of hobby users it would be a perfect little unit.

But I need 5 stepper motor control, so the extra space for the additional board just to driver the 5th axis makes it similar to the UC300ETH+UCBB in terms of footprint. More thinking . . . .

Well. I have mine stacked. I have uprights through the UCBB mount holes (3mm threaded rods + 4mm od brass tubes) going up to 2 ally straps. The UC300 is then fitted on top.
I can still get to all the screw terminals of the UCBB. Looks a little bodged on mine but it saved a lot of space.

[routercnc]

OK yes I can see how that would work. For info, I have been emailing Balazs at CNCdrive today and he has been replying pretty quickly. Impressive given it is Sunday afternoon/evening, and a bank holiday (in the UK at least) ! Anyway I asked what would happen if 2 stepper drivers were connected to a single step/dir output E.g. if I went with the compact AXBB-E route which can drive 4 stepper drivers as standard, and connected both of my 2 Z step/dir to the same output to drive a total of 5 steppers. This saves the additional BoB, ribbon cable etc. In this scenario I would have:

1: X axis
2: Y axis
3: Z axis & 2nd Z axis from same step/dir outputs
4: A axis (slave for X axis)

The answers from him were when using one output to drive 2 stepper drivers:

Homing could not be controlled independently - well that is OK for me as they are either side of the Z axis which can't rack

Issues for rack and pinion as often one side needs to rotate in the opposite direction - again that is OK as both turn the same way for me (plus I could swap 2 of the 4 motor cables)

In theory one output on the AXBB-E could drive 2 step inputs on 2 stepper drives since the output can drive 20mA current max. and usually one stepper drive requires only 5-10mA of current for the step optocoupler, so 20mA should be OK. - this was my main concern but sounds like this will probably work


So in summary I'm prepared to give the AXBB-E a go, double up on one step/dir output to drive 2 Z steppers, and if I do run into problems then I'll have to add the extra BoB for the 5th driver.

[JAZZCNC]

I have been emailing Balazs at CNCdrive today and he has been replying pretty quickly. Impressive given it is Sunday afternoon/evening, and a bank holiday (in the UK at least) !

I regularly email and get replies from Balazs late on an evening, even at weekends, he's always very very helpful and certainly knows his stuff so if he's ok with it then I'd say certainly worth a try.

[routercnc]

OK thanks. Although he has said this should work, I feel it is still on me as it is unusual.

Thanks for your help too, and the others in this post. I've sent you an email for some prices, but no need to reply until back at work (Tues?).

[JAZZCNC]

I've sent you an email for some prices, but no need to reply until back at work (Tues?).

No problem and we don't do Bank holidays or 9 to 5 around here so will email you tomorrow...:thumsup:

[routercnc]

OK great thank you

[dazp1976]

OK yes I can see how that would work. For info, I have been emailing Balazs at CNCdrive today and he has been replying pretty quickly. Impressive given it is Sunday afternoon/evening, and a bank holiday (in the UK at least) ! Anyway I asked what would happen if 2 stepper drivers were connected to a single step/dir output E.g. if I went with the compact AXBB-E route which can drive 4 stepper drivers as standard, and connected both of my 2 Z step/dir to the same output to drive a total of 5 steppers. This saves the additional BoB, ribbon cable etc. In this scenario I would have:

1: X axis
2: Y axis
3: Z axis & 2nd Z axis from same step/dir outputs
4: A axis (slave for X axis)

The answers from him were when using one output to drive 2 stepper drivers:


Beings you're talking about a Z axis. My assumption is you won't need it ridiculously fast.
It is my understanding that another option is you can run 2 steppers from a siingle driver as long as you follow ohms law.
So example if you have 2 steppers rated at 4A, you can safely run both together using a 8A capable driver so you can up the running voltage Thus only needing to use one axis output from controller.

A little research needed on this other option.

These diagrams are in series. Apparently run better with higher voltages due to it being shared.
They look slightly different so I assume one will spin both in same direction, one will spin them in opposite directions.
Can also run parallel which shares current so then needs higher amp settings.

https://content.instructables.com/ORIG/FRR/ILFP/IQW1GRRY/FRRILFPIQW1GRRY.jpg
https://i.imgur.com/JwKg3uj.jpg

[Kitwn]

Beings you're talking about a Z axis. My assumption is you won't need it ridiculously fast.
It is my understanding that another option is you can run 2 steppers from a siingle driver as long as you follow ohms law.
So example if you have 2 steppers rated at 4A, you can safely run both together using a 8A capable driver so you can up the running voltage Thus only needing to use one axis output from controller.

A little research needed on this other option.

These diagrams are in series. Apparently run better with higher voltages due to it being shared.
They look slightly different so I assume one will spin both in same direction, one will spin them in opposite directions.
Can also run parallel which shares current so then needs higher amp settings.

https://content.instructables.com/ORIG/FRR/ILFP/IQW1GRRY/FRRILFPIQW1GRRY.jpg
https://i.imgur.com/JwKg3uj.jpg

Looks good to me. The top diagram will run the motors in opposite directions, the bottom one will not.

With the coils in series you only need a 4A rated driver for two 4A motors, it's the same four amps in both coils. Problem is they only get half the volts each so you won't get as much torque at speed (you can also interpret this as twice the inductance for the same voltage). An 8A driver would allow the motors to be connected in parallel but you cannot guarantee exact current sharing and so the risk of one motor stalling while the other still runs is possibly increased. My money would be on separate drivers being fed the same step and dir signals. If you're worried about having enough current then the signals could be buffered easily enough. A quid's worth of MOSFETs and a resistor or two should do the trick. This leaves you without auto-squaring, which you say you don't need anyway, but seems no different from using a slave axis but only one homing sensor.

The other option is to use one big driver and one big motor driving both ballscrews with a belt.

Kit

[Neale]

I don't know if this is just an academic quibble, but is there an issue with the concept of two steppers driving a single, very rigid, axis with no possibility of adjusting their relative position? I'm thinking that the typical micro-stepping driver (although possibly not the latest and greatest) will reset to the nearest full step at switch-on, and there is no reason to suppose that the "full-step" positions on both steppers will align. In the case of the typical both-ends-of-a-gantry setup, there is enough give to allow for this (and we are taking about quite small mis-alignments - a full step with a 5mm pitch screw is only 0.025mm) but auto-squaring takes this out anyway. But if this dual-drive Z axis is going to be much stiffer, is this an issue? Maybe this will be addressed at initial setup?

I believe that some newer stepper drivers will maintain micro-step position across switch off/on, which might help.

[JAZZCNC]

It is my understanding that another option is you can run 2 steppers from a siingle driver as long as you follow ohms law.
So example if you have 2 steppers rated at 4A, you can safely run both together using a 8A capable driver so you can up the running voltage Thus only needing to use one axis output from controller.

Not a good idea because while the drive can handle the loads they can't handle other aspects of driving the motor like resonance compensation, the drive doesn't and can't know which motor is causing resonance so, therefore, can't apply any correction to the correct motor, also the correction applied would be all wrong because it's getting mixed signals. The upshot is a very rough running motor's.

[routercnc]

Not so keen on 2 motors off one driver.

Also for step alignment I will loosen one coupling, turn them on, then tighten the coupling. I think this will get the motors aligned internally.

Here is the machine as a reminder:
30378

30379

[JAZZCNC]

I don't know if this is just an academic quibble, but is there an issue with the concept of two steppers driving a single, very rigid, axis with no possibility of adjusting their relative position? I'm thinking that the typical micro-stepping driver (although possibly not the latest and greatest) will reset to the nearest full step at switch-on, and there is no reason to suppose that the "full-step" positions on both steppers will align.

I agree with Neale on this (even though I'm probably talking myself out of a sale...Lol), the motors will always jump to the nearest full step on startup or when drive powers down and you won't know if they both go the same way. Over time this will accumulate and twist up the axis. Loosening the coupler will only work that one time and without some way to home each motor to a known location, you will need to loosen the coupler every time to be accurate or at least every other power down/up.!.

[routercnc]

OK sounds like I will have to do some testing on the steppers and see how they behave over multiple startups (w/o couplers connected).

[routercnc]

Hang on. Is this any different to how I have been running this machine for several years? I have 2 identical stepper motors wired the same way to identical stepper drivers. They have their own step/dir signals but as these are slaved they should be the same. When I first built it I aligned the motors to full step with loose couplings and have not touched it since. Z axis sounds smooth as the day it was built. Only difference with the update is running off the same dir and step outputs.

[dazp1976]

I've only just really started wiring mine up. Hasn't turned out as well as I wanted. Everything is bigger than I thought and the box I got is a bit tight. Uc300 stacked on Ucbb.

30389

30390


I also have a second Ucbb that will go on the stilts between the existing bb and 300. This is in case I end up going to full servo control on all axis.

[JAZZCNC]

Hang on. Is this any different to how I have been running this machine for several years? I have 2 identical stepper motors wired the same way to identical stepper drivers. They have their own step/dir signals but as these are slaved they should be the same. When I first built it I aligned the motors to full step with loose couplings and have not touched it since. Z axis sounds smooth as the day it was built. Only difference with the update is running off the same dir and step outputs.

If you haven't been homing each motor then Yes, it's exactly the same. I can't see how the motors won't have jumped in opposite directions at some point and if you haven't got reference marks to check against then it could be very hard to see. It would be a random occurrence and you could have just been lucky that it's not twisted up enough to show up.? Still, it wouldn't be my first choice if I'm honest.!

[routercnc]

@dazp
Wow that is busy ! Funny how you always seem to need more space than you originally think.

@Dean
I hear you. I might have been lucky all these years but I could always add a pointer on the rear output shaft of each motor and periodically check they are in the same position after homing.

[Kitwn]

If you haven't been homing each motor then Yes, it's exactly the same. I can't see how the motors won't have jumped in opposite directions at some point and if you haven't got reference marks to check against then it could be very hard to see. It would be a random occurrence and you could have just been lucky that it's not twisted up enough to show up.? Still, it wouldn't be my first choice if I'm honest.!

Surely the mechanism will relax mechanically every time the machine is switched off and turn the motors back to the same starting positions. There may be a small twist at switch-on but it won't accumulate over multiple on/off cycles.

routercnc,
Thanks for the reminder pictures.
Sounds like you're talking yourself into a redesign to have a single motor and belt for Z as you have for X (or s it Y?). Was there a good reason for not using the same technique?

Kit