Just a thought... Maybe you have a bad encoder or connection to it.

Does the motor get the cutting head to the proper position at low speeds even whenyou have vibration? If it does that would seem to show that the encoder(encoders) is working. I wonder if there are two encoders -- one for the servo loop of the motor and one for the final positioning of the cutting head. If there are two encoders either one could be bad.

 

It occurs to me that, at those very low speeds, the motor will behave more or less as if its rotor is locked. It will provide virtually no back-emf so the current draw will be very high. That could be causing the 18201T H-bridge to go into self-protect mode, where it keeps switching off and then trying again to switch on (as per the Current Limiting section of the datasheet), hence the vibration you experience. A possible solution would be to find some way to reduce the motor stall current.

 

RichardO, Thanks, I didn't really realize that there were circuit boards with more than 2 layers. That would explain why in a few areas I can see a path that is not on the top or back. Can't really "see" between the layers much, even with a bright point of light, as the solid layers of resist or copper block it. I had suspicions though! That would also explain why a few of the pins of the H bridge seem to not be connected to anything, and especially the power input pin 6. If so, these 3 or 4 layer boards are a piece of art, there is no visible seam or joint to tell there is an inner layer, and it looks like one board in cross-section. Thanks again, Jeff

 

Thanks AlecT.
That would make sense to me too (I have limited electronic experience though!)
Could I put a resistor in one of the leads to the motor to cut back on some current?
Also, only has one encoder on each motor, a single enclosed disc opposite the output shaft, no other encoders.

 

I think I'll go and put a ammeter on the motors lead and see what it's drawing at very slow speed and medium speeds.

 

Also, only has one encoder on each motor, a single enclosed disc opposite the output shaft, no other encoders.

There might be additional encoders closer, mechanically, to the cutting head. These would be used to tell the internal computer the position of the head.

 

I re-installed the board to measure motor amps., it draws about .28-.38 depending on speed, etc. If I put some resistance on a pulley, it will go to 1 amp. I didn't try to load it heavily. The laser head carriage is not hard to move. Strange thing is that this time around with the testing that I did, it seems to be working fine now. I've had this board in and out a number of times, the latest a few days ago when I checked the large caps, and just before I removed it, it was still slowing down when it shouldn't. It wasn't vibrating like it did in the past, but I wasn't running it for an extended time either. This amp testing time, it was quite smooth. I won't say it is completely "fixed" yet without running it for a more extended time and over the course of a few days. Strange!!!

 

It is quite possible that you've had a so called "dry joint" in the region you soldered the capacitor and that by doing the soldering, have fixed the "dry Joint"! Happened to me a few times while fault finding on older tv sets.

 

I re-installed the board to measure motor amps., it draws about .28-.38 depending on speed, etc. If I put some resistance on a pulley, it will go to 1 amp. I didn't try to load it heavily. The laser head carriage is not hard to move. Strange thing is that this time around with the testing that I did, it seems to be working fine now. I've had this board in and out a number of times, the latest a few days ago when I checked the large caps, and just before I removed it, it was still slowing down when it shouldn't. It wasn't vibrating like it did in the past, but I wasn't running it for an extended time either. This amp testing time, it was quite smooth. I won't say it is completely "fixed" yet without running it for a more extended time and over the course of a few days. Strange!!!

Most strangeness in a circuit like that one can be attributed to false contacts (or dry joints, as Johan has just mentioned)
I once "fixed" a malfunctioning circuit by simply re-flowing every solder joint in it with the tip of a soldering iron.

 

Another problem is it is probably not evident as to what kind of servo control and feedback it has, if the motor has a tach also, then it is the older velocity controlled system as well as PID loop via encoder.
It sounds to me as though there is a fault in the gain control of the drive amplifier, with a decent servo set up, the PID control will keep the position to the least input increment of the encoder, regardless of commanded speed.
Max.

 

brush type dc motors have a problem called "cogging" as the brushes and the comutator segment gaps line up.

 

Strange thing is that this time around with the testing that I did, it seems to be working fine now.
...
I won't say it is completely "fixed" yet without running it for a more extended time and over the course of a few days. Strange!!!

I call this method of repair "laying on of hands". Sometimes it works for a long period of time and sometimes it doesn't. A typical cause is oxide/corrosion on a connector or socket pin.

An example:
In the early days of personal computing I had a memory chip go bad. Fortunately, the computer worked good enough that that I could run a memory diagnostic to find the bad part. I had a replacement part and was up and running fairly quickly. A few years later I had another memory failure. I ran the memory test again and it said the same part was bad. This time I just removed and reinstalled the same chip. Problem solved. Never failed again before I retired that computer.

 

brush type dc motors have a problem called "cogging" as the brushes and the comutator segment gaps line up.

Although BLDC have more of a cogging effect than DC brushed when used open loop, with a proper Torque mode feedback you can eliminate cogging completely with both DC and BLDC using a high frequency response to encoder feedback..
But a decent DC servo not only has a higher pole count than a BLDC, it also has skewed lamination's of the rotor.
Max.

 

I call this method of repair "laying on of hands". Sometimes it works for a long period of time and sometimes it doesn't. A typical cause is oxide/corrosion on a connector or socket pin.

An example:
In the early days of personal computing I had a memory chip go bad. Fortunately, the computer worked good enough that that I could run a memory diagnostic to find the bad part. I had a replacement part and was up and running fairly quickly. A few years later I had another memory failure. I ran the memory test again and it said the same part was bad. This time I just removed and reinstalled the same chip. Problem solved. Never failed again before I retired that computer.

I have a machine running with an MCU that's inserted in a ZIF. This machine is located in a very humid environment. Every year and a half or so strange things start to happen. All I have to do is remove that chip, scrape its pins lightly with a knife, and put it back in its socket. Problem solved.

 

Thanks guys, after some more extensive testing, I still have the slowing down problem although it seems like it is better than it was and less often. It's hard to quantify since it is not very consistent.
Amp draw varies quite a bit, but the highest I could see was about 1.5 or so for short periods of a few seconds. Most of the time it's around .3 to .8 amp. sometimes it is near 0, using a basic digital multi-meter.
I have not had the vibration problem with the motor as much lately, just the short slow downs.
I do think the cap I soldered on the backside to replace the SM cap that blew has helped some. I'm wondering if I should try a larger value, like a 1uf or a 10uf or maybe even 100uf. The data sheet doesn't show a cap between pin 6 and 8 (which is where mine is from factory), but does talk about supply bypassing between pin 6 supply power and pin 7 power ground. It kind of looks like pins 7 and 8 both eventually connect to ground. Data sheet does talk about 100uf per amp for bypass cap.
Any more thoughts on this, please?
Thanks.

 

how about mechanical backlash? is there any in the gears, belts or whatever between the motor and the laser mount?

 

Hello, no backlash, all bearings are great, it uses small cogged timing belts that are tightened to reasonable tension. I have tried slightly less tight and tighter, but no change in motor stalling/slowing in small spots while moving on the Y axis. No tight spots in linear axis slide. Thanks.