two months of x4 experiences - design flaws, safety issues, and performance problems

[rickturner646]

I've had one of the new Syil x4+ machines for about two months now and I thought that I'd feed back my experiences so that y'all “out there” can perhaps benefit from the learning curve that I have had to go up, and have some information from a user rather than from marketing literature. Disclaimer: these are of course personal opinions based on my own experiences.

At the start of this year I was looking for a small CNC mill to do small scale component prototypes for engineering purposes, and secondly to assist with my hobby of building 3.5 + 5 inch gauge live steam locomotives. The new Syil x4+ seemed a perfect fit, so I ordered one up from China and waited with baited breath for it to arrive. Two months ago, the big day as the machine arrived, and I got started assembling it. I have to say that ever since then, things have gone downhill.

Overall I am very disappointed with the poor performance of the machine, with the bad build quality, and with the lack of support that I have (not) received. There are also some serious operational issues that potential users need to be aware of as they cause real safety issues – the machine that I have makes unpredictable (and uncommanded) high speed moves when cutting, which is incredibly dangerous.

Performance
Despite the size and weight of the machine it struggles with anything but the lightest of cuts and vibrates badly. Chatter is noticeable even with feed rates as low as F50 and a cut depth of 1mm both in brass and ordinary EN1A bright mild steel. Cutter geometry seems to make little difference – I have found that industrial indexed cutters work best for roughing, and multi-flute carbide cutters for finishing.

If you increase the speed much beyond this the vibration becomes bad enough to shake the whole machine. The practical limit is probably F100 or perhaps a bit less. With cutting speeds that have to be this low, it takes a long time to finish a part and therefore this machine is probably not suitable for any sort of production application – even light ones, and it would seem to be best in the 'pure hobby' category.

I will attach some images to this post that show the (results of the) problems that I am describing and you will be able to clearly see the chatter in some of them. When when I sent a couple of these images to Syil in China, I was first told “I am cutting too fast” despite the F50 feed rate. When I sent them the g-code program used to cut the part, the response that I got back was “sorry, this machine is not meant to cut steel, only aluminium”. This is of course not stated anywhere in the marketing material, so be warned if you are expecting to work with any other metal.

As an aside, as I can machine exactly the same parts with exactly the same cutters in exactly the same material manually on my little Sieg X2 without any problems and – for simple parts such as the brackets illustrated here – in less time than the x4 needs. This would suggest that there is nothing particularly 'difficult' about the parts that are causing problems for the x4. Given these circumstances, I am inclined to think that the lack of rigidity is due to a design problem – my current best guess is that the way that the vertical column is fixed to the base is inadequate (it's just four small bolts).

Build Quality
The machine was shipped to me in pieces in the crate and I had to assemble it myself. Internal parts are not marked and the instructions are less than illuminating, so be prepared for a serious challenge here.

On arrival, the 4th axis unit was found to be inoperable due to excessive vibration and overheating of the motor. It took nearly a month of exchanging emails before I got this replaced, and the new unit is not much better. It still vibrates quite badly, though at least the motor doesn't heat up quite as fast.

The 4th axis unit is permanently wired into the guts of the machine and cannot be removed, even if you are not using it. You'll need to find somewhere safe close to the machine to put this down when not in active use. The cables are not long enough to reach more than a couple of feet away, and it's both heavy and an awkward shape. A simple four pin DIN plug would solve this issue, but I got no response to my suggestion on this to Syil. It's an easy enough thing to do for yourself, however.

The display panel on my machine flashed irregularly on and off when I got it – turned out to be due to a couple of loose connections in the electronics behind the panel itself (you have to take the top front cover off to get into these).

Serious Operational Problems – and more Design Flaws?
The worst problem that I have found however is of serious concern from a safety perspective and one that certainly makes me worried.

For reasons unknown, my x4 is liable to randomly make uncommanded moves whilst cutting, often at high speed and with great force. One of these unexpected moves broke a 12mm carbide cutter in half, which gives you some idea of the forces at work.

The most heart-stopping instances occur when the machine decides to move laterally whilst the cutter is embedded in a slot or drilling a hole – the resulting noise is kinda loud and frequently accompanied by bits of broken metal flying through the air. I've literally lost count of the number of cutters that have broken and parts that have been ruined by this.

Most annoying of all, I've had uncommanded plunge cuts go through the bottom of the part that I am cutting, and have had fixtures damaged by this. Even though I am leaving 3-5mm of extra metal as a clearance space. Two images of this are attached to this document – on shows a bracket where a lateral cut suddenly turned into a plunge, and the other shows the back of a piece of brass from which I was machining a small gear – you can see how the 1mm deep cut gradually starts to penetrate the bottom of the 2mm brass sheet. I stopped this running as the second pass would have gone right through the backing plate into the table.

The safety implications are obvious – unpredictable movement of spinning cutters is extremely dangerous. I have reported this to Syil more than once, but have gotten no explanation or acknowledgement of the problem, much less a fix or support (see below). From a personal perspective, I would feel really bad if I heard that somebody had gotten hurt by one of these moves so I would be remiss if I were not to voice my concerns about this issue here. If you're considering using one of these machines, you need to be aware of the potential danger.

One other thing to be aware of is that hitting the e-stop button on the machine wont stop the machine for maybe 5 seconds, perhaps a little bit longer. This applies to the soft e-stop button in Mach too. It's probably legally allowable but five seconds can be an awful long time if something is going wrong.

Smaller Moves
The precision and repeatability of the x4 leave a lot to be desired too. I suspect that it is simply a less dramatic example of the uncommanded movement referred to above, but reasons unknown there are random and all too frequent positioning errors.

Midway through a programme, the machine will sometimes move further than it is instructed to, with the result that features that are supposed to be cut co-linearly are offset to one side. If you look at the images of the bracket with the slot, you'll see several things:

• the cuts get progressively deeper – 1mm at the top and nearly 3mm at the base, where the cutter went through the programmed bottom (safe) level
• the slot is off centre, the result of a positioning error
• chatter on the outside, caused by machine vibration even though machining speed was F50 with a 1mm cut depth.

I also attach a picture of partly machined main rod segments (for a locomotive). The lower one has the centre slot offset towards the camera – this happened midway through a cut. A more illustrative example is the upper one; if you look at the left hand end of the rod just next to the wider end piece, you'll see a marked dogleg in the cutting path. This is where the machine 'jumped' down and left (as the photo is oriented) midway through the cut. The cut is therefore both offset and too long.

Whatever the cause is, it's a problem on the x4. I've put a breakout box on the parallel port and Mach is behaving – if you tell the machine to move (say) 6mm Mach will send the right number of pulses, and the Mach position indicators will say “6mm” - but the machine will have moved 8.5mm. Or 7mm. Or 4.169 mm. This seems to affect the Z axis most but I have seen it on all three linear axes. I also suspect that there is a similar problem with the 4th axis unit too but I have not gotten to the bottom of this yet. The position here is that the 4th axis does not always move correctly on command either. Funnily, all the odd moves seem to be too far, not too little, so the explanation is not that it's missing Mach pulses either.

The positioning problem also appears in a different guise - the machine appears to 'lose track' of offsets and zero points. After one of these random positioning errors one or more of the x/y/z zero (origin) points are in effect displaced by the amount of the position error. This means that you have to stop the job and reset all the G54 offsets and machine zero. This is a silent error and if you don't notice it, the first warning you will get is when the next tool or cut prangs.

Spindle Speed
There are problems with the way that the x4 responds to spindle speed commands. If I run a piece of g-code that has a setting of say 'S500', the mill will start up at say 740 rpm. If I put S900 into the code, I still get 740 rpm. If I then use the manual override in Mach, and boost the speed, it will do nothing for two or three clicks and I will end up with a situation where Mach says 'normal speed 900', 'speed override 1140', and a mill still doing 740 rpm...... At some point, the speed override does kick in but incorrectly - in this case, upping it again to say 1200 manually will make the actual mill speed jump from 740 to maybe 1420 rpm.


Support
I have raised all of these problems with Syil but they have not been able to provide any information or fixes. In fact, the usual first response is to tell me that I am “doing something wrong” or “not understanding” their machine, or they will blame “somebody else's” software.

I've run pretty exhaustive tests and I am certain that the software is not at fault, however. The driver software for the x4 is Mach3, which is pretty well a standard and used extensively without problems elsewhere. Mach's own diagnostics suggest that nothing untoward is going on, and where I have checked this externally (e.g. with the breakout box) it confirms that Mach is working correctly.

Ditto the CAM software that I am using, which is SprutCAM; this *is* an industry standard and in daily use for serious production applications. The g-code that it generates is fine – paths are sensible and they work on other machines. I had somebody cut a couple of the parts on a Haas machine just to be sure – they went ok first time without any issues.

My current view is that the root cause of all these issues is that there are either electronic design flaws or component tolerance issues in the x4 electronics – since the g-code is ok, and Mach is behaving, the problems must lie downstream of these, which means it has to be in the on-board processors and electronics on the x4.

As for me personally “not understanding”, well, I am an engineer by training with 30 years experience of CAD/CAM and CAE, and engineering. I think that I know what I am doing by now.

The net result of all this is that in the two months that I have had the machine, I've only been able to successfully complete perhaps ten simple parts, but have had to throw away at least four times that number which have been damaged by the unpredictable behaviour of the machine. I've not yet managed to complete any complicated parts, because something inevitably goes wrong in the process. I've run through about 50 cutters, all broken by uncommanded moves.

I have had precisely zero support from Syil UK – who seem to be little more than an importer, to be fair to them – and the support from China is perhaps best described as “well meaning but ineffectual”. There are significant language difficulties, and it simply doesn't work. Support in your country may well be better than it is here, but it's something that you need to consider.

It's possible that I have a “Monday machine” and that you will be lucky enough to get one that works perfectly out of the box. I hope so – if you get one like mine, you're going to end up just as frustrated as I am.

Regards to all, and happy machining!

Rick

[Al_The_Man]

Originally Posted by rickturner646 The safety implications are obvious – unpredictable movement of spinning cutters is extremely dangerous. One other thing to be aware of is that hitting the e-stop button on the machine wont stop the machine for maybe 5 seconds, perhaps a little bit longer. This applies to the soft e-stop button in Mach too. It's probably legally allowable but five seconds can be an awful long time if something is going wrong. Rick

It sounds like they are not conforming to standard safety regulations, here in Canada the machine would have to be CSA approved, IIRC the British Standard is BSI.
There may be another now under the EU.
A 5 second delay in an E-stop would normally be considered unacceptable by any standard.
The E-stop sounds like it is software implemented instead of hardware operated.
Al.

[Al_The_Man]

BTW, regarding the chatter, did you snug the gibs up enough?
I assume you did, but thought I would ask.
Al.

[rickturner646]

Yep, I did tighten up the gibs etc. I ran it as it arrived for a couple of days to see how it went then when I discovered the issues I put it through what you might call a "10000 mile service" as I would with any machine. Made no difference.

e-stop: the machine sports a set of CE marks (the EU equivalent of your CSA) but I am surprised that it passed their checks with this behaviour. I've begun following up on this here - my personal view is that the machines are very dangerous to operate in their present form. It's not just my machine either - there is another post in this same forum from somebody who has the same behaviour.

Rick

[iGG1e]

I'm sorry to hear about all of your problems. I thought I was getting a bit sick of fettling the SX3 version to get the results I, in my opinion, should've been seeing.

There's no point in having an emergency stop if it doesn't immediately halt the machine. You might as well just stand there with your hand on the mains plug ready to pull it.

Have you been all over the wiring and basically re-crimping and tightening everything in sight? Checked the current supply to the steppers etc. Monday morning is one way to look at it...I'd go with rushing to knock off early on Friday as well.

The spindle speed control is nothing short of useless. I've got a spreadsheet I use to put in a fudge figure that'll hopefully end up with the mill running close to the speed I want. Forget trying to request a speed of below 700RPM, they need to be done on the front panel. I really want to scrap their controller and use something that works. In fact I now regret not just getting a standard mill and converting it myself.


I hope things get better for you very soon.

[syil canada]

Good day,

I am the Canadian distributor for Syil in Canada.

i would like to post my personal x4 experiences, we are carrying the x4+ (the yellow one), i have tried to replicate some of the issues that you are having with the x4. I am assuming that you have the x4 (the blue one), as i have not been able to replicate the same problems.
As the Canadian distributor we make sure to test all machines prior to send them out to the customer, during this process we even have caught a few complications. which as the distributor our problem to solve not the customers, I am unsure why you would have to assemble the machine, as it should have been fairly plug and play.
As of current with the x4+ we are cutting titanium, stainless, and aluminum. I have not experienced any of the same problems, but you pictures that you posted defiantly show a big problem. Through my experienced with being the syil distributor we have encounter as similar problem, the problem we discovered was in mach3.
Here is the story, as eager distributors’ when our first shipment arrived we were very excited to set up our x3+. We wanted everything to be perfect, so i went out and bought a kickin computer, huge ram, huge processer, and running windows XP. I hooked the mill to the computer and attempted to do some basic cuts, things went south in a quick hurry “literally”.
I found that i was having a hard time moving both axis at the same , it was also making un programmed movements?? The First thing i tried was checking the gibbs to see if he were way to tight, they were good. Then i checked the ball screws, they were also good. Checked the mach3 config file, there was a few quarks, but it was fine.
I then checked the serial cable, it was fine. I as a last resort i checked the parallel card, it was also functioning properly????
The only thing left was the x3+ its self, i ripped apart the machine and started replacing parts, i turned on the machine and attempted the same cuts. Unfortunately the same thing happened!!!!!
I went home to bed “very frustrated” i awoke in the middle of the night and thought “the computer”, so I stopped on the way to work and picked up a $350.00 acer desktop computer. I swapped out the computers and presto, it worked. This same machine has now 1000+ hrs of machine time on it?

So what was the problem? the processer!
I found out the hard way that mach 3 does not like 64bit processor; this was later confirmed to me at the mach3 convention in Knoxville. Believe it or not mach3 does not also like laptop computers; the serial port on al laptop is not always capable of constantly outputting the proper voltage signal.

After that i suggest to all of my clients not to use 64bit computers, all of my machines now run 32 bit processors. As i was told that the “Syil cnc will only do what they are told to do by the computer”. Any e stop functions are on the pin config controlled by the computer, it the computer is too busy doing different processes it can take a few seconds to register.
All of my computers are stripped bare of all non essential software and are not hooked to the internet. I learnt the hard way turn off the automatic windows updater, i was running a program that i had run a hundred times before, the windows updated screen popped up and my machine crashed hard!

If you are finding that the spindle speed is incorrect, this can be adjusted in mach3, this is done by
CONFIG menu>
Spindle pulleys
You will see a spot that the ratio can be adjusted, it is usually set to (1) try playing with number (ie).997 ect. This number can be changed up or down, and with some playing you can have mach3 and the machine aligned with the same rpm number.

I am not sure how you program, as everyone does it different, so pardon me if i state the obvious. When i am running a program i will use a g28 to sent all axis home to its switches, i find that this keeps the repeatability of my parts accurate. As the machine will only know where the part is based on the work piece coordinates (g54) it does this by knowing how far it is from its switches. I will also in mach3 click the machine coordinates button , and REF ALL AXIS button. I do this when i start up the machine every time.
In regards to tool chatter, i am unsure whether you are working in mm or inches, my calculations, in mild steel, would be,
4 Flute .250 (hss) endmill = 3.11 imp or 79 mm. @ a rpm of 1390
Or
2 flute.250 (hss) endmill = 1.56 imp or 39.36 mm @ a rpm of 1390

A super handy tool for these calculations can be found @ http://www.cnczone.com.com/autonc.htm it is a free program and the machinist calculator is great.

I hope that this will help you, if you have any other questions feel free to email me @ keith@syil.ca + I can send our imperial .xml file (sorry no metric)


keith

[syil]

yes,rick machine was belong frist shipment which still use small wood carton.
so,need him to wire the cable which we both think a trouble work.

also,we still improve the manual,it is a big challenge for us,but we have been ask some office to help us.

i`m not sure about keith regards computer cpu was relative the mach.becuase in syil,we have many computer in diffferent ``generation``
total working well.some it is a test pc..
we are here in shanghai machine show,and becuase of that,i book a dell desktop pc with small computer case and have original paralle port.working well.

with Rick problem issue we are very very sorry.take so much time and very poor support for him but anyway,we currectly have been make the deal in common.
So.hope other customers also can provide more good advise for our machine.
Syil China.Xushuo

[pointcloud]

Just courious but what were your feedrates? What kind of cutter? Diameter? Coolant? How many flutes? Climb cutting of Conventinal? Depth of cut? step distance?

Also how did you find the parts center?

On the uncommanded plunge did you check to see if that was a mechanical issue? another words was Z0 still in the same place? Did the machine fall and Z0 now looks like Z-? Did you check to see if the CUTTER slipped down?

[Al_The_Man]

Originally Posted by syil canada After that i suggest to all of my clients not to use 64bit computers, all of my machines now run 32 bit processors. As i was told that the “Syil cnc will only do what they are told to do by the computer”. Any e stop functions are on the pin config controlled by the computer, it the computer is too busy doing different processes it can take a few seconds to register.

I am puzzled how a machine is able to be sold with this condition to perform an unsafe E-stop as it appears to contravene the Electrical code and NFPA79 recommendations.
This is why Safety Relays are replacing normal E-stop relay practice and will probably become mandatory in N.A. also.
Al.

[highspeedmazak]

The computer is the first thing that I would replace and replace it with one that is dedicated for that machine only no onlining or anything else so that it does not have any conflics, I dont even know why they would even bother selling a laptop to run a cnc? Then try cutting say .050 deep aluminum 3/8 cutter anywhere around 2 to 4000 rpm depending on the cutter at say feed of 8. inches a minute. that would be a good start for a good finish and dont think that it is a mazak so be patient cutting. Steel (slow your cuts down) That machine shoud do just fine I have seen them cut very nice now I have never seen the blue one? The E stop well that is not good??????? Hope things get better for you on that machine.

[syil canada]

Originally Posted by Al_The_Man I am puzzled how a machine is able to be sold with this condition to perform an unsafe E-stop as it appears to contravene the Electrical code and NFPA79 recommendations. This is why Safety Relays are replacing normal E-stop relay practice and will probably become mandatory in N.A. also. Al.

This would depend on the way the the e stop would be applied, sorry but there was not enough information posted. I am referencing if the stop is done in mach3, at that point the computer is sending the signal. if the estop button is depressed on the machine it should be almost instant, and you computer load will not effect the estop.

I now have sold many of these x4+ mills, and have had them in stock for months. I have never encountered any estop malfunctions, as we test every one of our machines prior to sending them out.

my concern for the the fellow in the uk is; if the depth of cut is large and fast the amount of chatter that it is producing could damage the machine and the electronics.

As you can seen on the step down on the part in the photo, this is a large depth of cut and if done at the posted feed rate, you would be hard pressed to do that even on a large CNC mill.

As of current i am throughly impressed with the x4+, i personally own 2. these machines are a workhorse, but like any mill the depth of cut and feed are limited to the size and rigidity of the machine.

keith

[Al_The_Man]

Originally Posted by syil canada This would depend on the way the the e stop would be applied, sorry but there was not enough information posted. I am referencing if the stop is done in mach3, at that point the computer is sending the signal. if the estop button is depressed on the machine it should be almost instant, and you computer load will not effect the estop. keith

Personally I think it is a bad idea to have a screen or PC operated E-stop Button at all, even if there is also a parallel hard-wired one fitted.
It leads to confusion and may give the impression to an owner/operator that they operate in the same way.
If I were to implement a machine in this way, I would be concerned about injury related law suits.
The original poster mentions 'hitting the E-stop' so this implies he initiated the E-stop and not the PC detecting a fault.
Al.