[mflux_gamblej]

I mistakenly gave the machine M03, to go to 4000 rpm, with the brake turned on.

The machine tried to move the spindle, and I recognized the issue within 1 to 2 seconds, and switched the brake off, the machine spun up, but it will not make it to speed now.

I've taken the brake off of the machine, thinking that it would be a brake issue. But it isn't. With the brake taken completely off the machine, the spindle will only make it to speed at around 2800 RPM.

The spindle makes a little bit more vibration than I remember. Also, if you turn the spindle by hand, in the CW direction, there is no noise, but in the CCW direction, there is a 'clicking' kind of noise.


Some initial thoughts that I have, 1.) the KTK spindle controller is set to 50 Hz, this could limit the reachable speed to 50/60 * 4000, which is around 3300 RPM, assuming that the drive board's tacho is dependent on the incoming AC frequency.



I was running the machine last night despite the trouble, and eventually the overheating overload throws, and the spindle shuts down.

The machine did appear to go to 4000 rpms before I did this brake on mistake.

Any thoughts?

[mflux_gamblej]

I consider this to be a machine crash. I cannot believe that the control system allows you to start the spindle with the damn brake on.

Whining aside, I am convinced that the spindle motor is damaged in one or more ways. Some symptoms / observations:

• When I was disassembling the brake, the lowermost plate that is bolted to the top of the DC spindle motor was 'clamped' on in there pretty good. It seems that the motor may have torqued the brake assay hard enough to deform the top of the motor!? I got it out, but had to use a flathead screw driver to lever it off.
• The spindle motor turns freely with no rubbing (a positive note) in the CW direction, but in CCW you can hear some clicking type noise. This may simply be the magnetic 'cogging' that you get with DC motors, or it might be some damaged commutator / brushes
• The KTK drive electronics and MTI board are behaving just fine. Interestingly, if you switch to indicator position 7 (?) on the KTK FAC1R board, it doesn't indicate that the motor is under maximum load, even though it cannot reach speed. The indicator would read 3 if it was under max load, but it reads back and forth between 1 and 2 I wonder if the tacho encoder took a mechanical jolt here?
• The spindle used to spin up to 4000 RPM without any problem, and made lesss vibrations than it does now, although the vibrations aren't particularly loud or ominous, and my memory might not be perfect.
• I did a sanity check of the tacho accuracy by setting the spindle speed to 40 RPM and counting the rotations for 30 seconds, it made 20 rotations in 30 seconds - that's 40 RPM all right..
• How could it be that the machine can operate just fine, but only up to 2800 RPM (out of the 4000 its suppposed to), but cannot reach speeds beyond this, but the motor drive indicates that it's not even trying all that hard?

[mflux_gamblej]

I ran the spindle at 1000 RPM in the reverse direction, and used the spindle override to vary the speed a little bit. The system seemed to be having trouble keeping up with the changes.

I programmed 1800 RPM, and the spindle was making the kind of growling vibrating noise that I've described, and it did not make it to speed.

I changed the speed to 40 RPM and the spindle simply stopped. Now it will not start at all - the MTI-1 board isn't powered up.

I've done some preliminary looking around, the LED on the MTI-1 board does not turn on. Both of the fuses (5A) located just above the transformer on the MTI-1 board are blowing a second or two after you turn on the machine isolator.

My plan is to use the interact 1 mk 2 maintenance manual to pretend that I installed a new drive - it talks you through setting some current limiting pots on the FAC1R, etc. Does anyone have commentary on why the MTI-1 would be blowing those fuses?

After that, I expect to be back where I was.. with a motor than cant get to speed after a 'spindle on command to max speed while electromagnetic brake was applied crash.'

I want to remove the motor from the machine to check the commutator and brushes. Does anyone have experience removing the spindle motor from an interact 1 mk 2? Is there anyway to check the brushes without moving the motor out the machine with a hoist / crane?

[mflux_gamblej]

I've read in a few places, people talking about the brake in 'Auto,' or 'Off' modes. The Interact 1 mk 2 has a switch on the machine that has symbols as follows:

(<-O->), and
->(O)<-

I assume this to be Off (above), and On (below).

The machine was in the 'On' (below) position when I gave the machine the M03 to 4000 RPM command on Th. night when the motor and/or drive crashed. The spindle thereafter will only reach 2800 rpm.

Since then I went and played around, in an uneducated way, with the pots on the FAC1R and MTI1 boards. Now the MTI-1 blows the two 5A fuses above it's small on-board transformer when I close the machine isolator.

Can some of you with experience on an interact explain the On/Off/Auto settings to me?



In terms of getting this situation repaired, so far I have inspected the motor's brush, which is short, but looks good. I expect that the armature or field windings were over-currented during the 'crash.' If the switches as described above are indeed off or auto, my machine has been tampered with in an undesirable way, or if they're simply off or on, the interact 1 mk2 must have been crashed in this same way an un-countable amount of times, and someone must have seen this happen before.

[keebler303]

I don't know anything about your machine but I have a Series 1 BOSS machine with pneumatically actuated brake. The brake is either "auto" or "off". In auto it retracts the brake when the spindle is started. The only way to turn on the spindle with the brake on is if the brake was already on and there is no air pressure to release the brake. On a Series 1, the belt will simply slip so no harm is done to anything.

I cannot imagine that your machine would allow this to occur. I think it may have been modified or some part is not functioning as it should.

The daily operation of the brake slamming on when the spindle is at 4000rpm would probably be just as likely to tighten it excessively as a single locked rotor event.

The motor should have brushes, you stated one brush. Make sure you check them all (multiple of 2 like 2,4,6,8). Is the commutator clean? Any evidence of arcing?

The field current is normally controlled by the drive. Even in a locked rotor situation, it should not change significantly. It is possible that the drive or motor was damaged due to an overcurrent in the armature windings. You can check the windings using an ohmmeter and opposite contacts on the commutator. Basically you are connecting the ohmmeter how the brushes would contact them. If you have any shorts or open circuits, your motor is cooked. These motors are generally pretty rugged and I'd be surprised if 2 seconds of overload damaged it.

Is it an SCR spindle drive? One or more SCR's could be blown open which would reduce the average armature voltage which would reduce max speed.

Does the MTI-1 have anything to do with gate drive for the SCR's? Could be blowing due to the blown SCR's.

Why in God's name would you play around with pots in an "uneducated manner"? They are surely controlling something of importance and I see no way how random fumbling would better the situation.

I am working on a large lathe right now with DC spindle and SCR drive. The previous owner crashed it pretty big time and it blew a couple SCR's in the drive. The field supply and the motor itself are fine.

Good Luck
Matt

[mark562]

The Interact 1 mk 2 has a magnetic actuated brake. There is a large electro-magnet coil on the top of the motor shaft. The switch you refereed to is an om/off for Manuel mode operation of the brake. there should be no way for the brake to not release in program-run mode. There must be an electrical problem between the control and the brakes coil (broken wire, bad relay or switch) your electrical schematic should have the relays marked.

Mark

[mflux_gamblej]

If the brake will auto dis-engage in program-run mode, I don't know. The machine was in manual operation, or hand-wheel mode when the crash happened.

Currently I've got the drive box (FAC1R, THY1R, and SOL1 boards in a big aluminum heat sink box) pulled out of the machine for SCR testing.

So far, the testing shows that the SCR's on the THY1R board are OK, although, since the SCR's come in a series package with two SCR's each, I'm unsure if my testing methods are adequate. I'll post about how to test SCR's when I'm sure I've got it right. They all give the same response though, which means they're all OK, or all dead.

Interestingly, the machine will overload even when the spindle is not spinning (cycle start has been pressed, spindle fan motor is spinning). The first time the machine overloaded, it did not trip the big breaker overload near the MTI-1 board. The spindle drive simply shut down. Several minutes later, I went back out and turned things back on, and the spindle ran again (but still with the 2800 RPM max speed). Eventually, after about an hour, the drive overloaded again, this time with the spindle running, and did trip the big overload breaker.

I've measured the SCR on the MTI-1 board. It gives the expected Diode behavior, but I need to pull that board, take the SCR off and measure it independently. I ordered a replacement SCR for this one, since its cheap.

Thanks for the support guys

[mflux_gamblej]

The interact 1 mk2 does have a way to remove the spindle motor which is convenient for maintenance. There are a pair of rollers that are bolted (in a temp storage location) to the top of the machine. When you want to take the motor out, you take off the brake assay, release the drive belt, attache the rollers, unbolt the motor maintenance rail, re-bolt it in the 'out' position, and slide the motor on the rollers out of the machine.

I'll need to do this motor extraction on rollers to check the tachometer. Currently, the fan is off of the motor which exposes the motor brush and commutator (all of which look good).

I have attempted to measure the resistance between poles on the commutator, but cannot - since the expected resistance is around 30 milli-ohm, you need a 4-probe milli-ohm meter to make the measurement. Even a fancy FLUKE dmm only reads down around 100 milli-ohm. Another option would be to apply some current and make voltage drop measurements. I am hesitant to do this. If I cannot find an electronics problem in the drive, I'll buy a milli-ohm meter and get back to testing the motor.

My opinion is that the motor is fine. There was no hot / burning smell and the locked rotor event lasted 1 to 2 seconds.

Additionally, since the drive will overload without the spindle even running, I am suspicious of broken electronics. Hopefully on the MTI-1 board, which is easy to rebuild.

There appears to be only 1 brush. The brush has two wires that attach to it, but they're connected together. The brush face has a clear line down its longitudinal axis, which might indicate a two brushes in 1 assembly? It is also possible that there is only 1 brush, or the other brush is somewhere else in the motor.

I've been taking lots of pictures along the way, and I wouldn't mind posting some of the instructive ones here - although, it seems difficult to get a picture to show up on this forum.

[gus]

I have never seen a brush motor with one brush. Current path needs two brushes to make a circuit

[keebler303]

Agreed, there are at least 2 brushes, they should be located 180 degrees apart.

Matt

[Al_The_Man]

Interestingly, the machine will overload even when the spindle is not spinning (cycle start has been pressed, spindle fan motor is spinning).
Thanks for the support guys

Try it without the motor connected, if a SCR or the drive is shorted it will overload/trip in this condition without load.

I've been taking lots of pictures along the way, and I wouldn't mind posting some of the instructive ones here - although, it seems difficult to get a picture to show up on this forum.

I usually have no problem, you can resize them with Irfanview if needed.
Al.

[mflux_gamblej]

I tested the SCR's on the THY1R board, they're all good. This is nice since a replacement is nearly impossible to source, and a substitute SCR power package is around $90.00 each and up.

Here's how to test SCR's:

1.) a picture of my test setup: 'SCR Test setup 001.jpg'
2.) a circuit diagram for the same: 'SCR Test Circuit Diagram.jpg'
3.) a close up of the connections on the THY1R board: 'SCR Test Setup 002.jpg'
4.) shows the SCR on, then off: 'SCR Test Setup 003.jpg'

An SCR works like a diode, until it is turned on. You use the Gate lead to turn on the SCR. The gate Current and Voltage are on the mfg.'s datasheet (try finding a datasheet for an SCR package from the 80s!) Do not overdo it, you could blow the gate, and that would be unfortunate.

Once the SCR is turned on by applying the required gate current, the SCR will conduct current from the anode to the cathode (labeled K) as long as the minimum 'hold-up' current is being supplied.

The SCR stays on after you take away the gate current. The gate is only a switch.

The SCR will turn off, when you break the current supply to the anode / cathode.

Then you have to turn it on again using the gate.


To test an SCR, you need a power supply, or a couple of batteries (and an LED or flashlight bulb). You connect up a circuit as shown. If you don't have a PSU which will read out the current, you can use a light bulb, it turns on when there is current - put it in series with the Anode to Cathode circuit. If you use an LED, you'll need a 3rd resistor to set the LED current. The resistors are to control the current in each piece of the circuit. The gate current is the most important, especially since we're talking about very high power SCR's, the gate current is pretty much the only sensitive one under benchtop testing. Use V = IR (ohms law) to calculate the value of the resistor, R1, that will decide the gate current, when using the mfg recommended gate voltage. The other resistor, R2, is to protect/limit the anode to cathode current, which isn't as sensitive in this case, but its a good safe-guard to hold the currents low.

When you apply a voltage across the anode to cathode, nothing happens (diode until gate is turned on remember).

Then, when you apply current to the gate, the device turns on. In 'SCR Test Setup 003.jpg' (upper image) I showed the effects after pressing the gate switch (SW2, NO) - the PSU reads around 250 mA. This is the current from the anode to the cathode. When you are pressing the gate switch, the current will read the sum of the gate and A-> K currents. In my case the gate current is 150mA, gate voltage is 2.5V. The lower image in that file shows after I have pressed the circuit interrupt switch (SW1, NC). NC means normally closed, NO means normally open, i.e. you apply gate current when you press the button, but not before.

If the current shuts off between the anode and cathode when you release the gate switch, make sure you've exceeded the 'holdup current' (also on mfg datasheet). If you have, and it still doesn't stay on, the SCR is dead.

If you press the gate switch and nothing happens, check your connections, and that you're applying enough current and try again, if still nothing, the SCR is dead.


In my case, on this THY1R board, there are two SCR's in each 'powerblock' assembly. There are 6 power blocks, making 12 thyristors (SCRs), which are used for rectifying (chopping only the peaks of current off of your 3 phase power oscillating through the rectifier input - this makes DC). In a 12 SCR DC generating config, this is enough to run the thyristor rectifyer in quad. You can do fwd, rev, speed control, and regen. braking (fast braking as far as we care about here).


I am dissapointed to find this really, because I'm not totally sure where to go next. I'll check the SCR on the MTI-1 board tomorrow, and replace it anyway when my new one arrives (it was only $6). If that SCR is shorted out, then this would explain the blowing fuses on the MTI-1. But it might not totally explain the can't get to full speed problem.

[gridley51]

Have you been playing with the max current pots?

[mflux_gamblej]

Have you been playing with the max current pots?

I made the comment about 'uneducated' messing with pots.

I know what each pot does, and gave the max voltage, max field each a very small CW adjustment to see if it would make the situation any different. This was after the crash, looking to see if the max attainable speed would change. It did not.

George sent me his KTK drive notes, which have a procedure on setting each pot. Tonight, or tomorrow I'll pull the MTI1 board, test it's SCR, replace it if necessary, and reassemble the drive electronics. Currently I still have the drive pulled as shown below.

After this, I am going to pull the motor to see if I can find the other brush, and to check the tacho brushes. I also would like to measure the resistance between places on the commutator, although, the setup requires a 4 pt probe. That's fine, I've got a couple of DMM's and a power supply, but it's not going to be easy to fit all that hardware into the inspection port on the motor. The expected resistance is ~ 30 milliohms. Your typical fluke 87 (their nicest) DMM can measure 100 +/- 100 milliohms. The contact resistance of the probes is simply too high. This measurement would confirm that the motor is OK, though, which I suspect that it is.


This failure / crash comes from 1 of 2 things: 1.) in manual and/or handwheel modes on the 151B, you're allowed to turn on the spindle with the brake engaged (ever heard of an interlock!?! BP?) or 2.) a previous owner has done something stupid to the mill and disabled the auto off, or a broken relay

Option 2 is more likely, in my opinion, but it's the least of my worries at the moment.


Gridley, do you have experience with this particular set of BP interact machines? Ever seen these symptoms before? I swore up and down that the SCRs on the THY1R board (at least 1 of them) would have been blown. But they're all OK.

[mflux_gamblej]

Indeed, the SCR on the MTI-1 board is shorted out. All leads (K, A, G) are shorted ~ 0.5 ohm, any combination.

I believe that this happened after I changed the current setting on the MTI1 board though.

Of interest, the motor could not keep up with my speed changes and only made it to around 1800 RPM in the reverse direction when the drive completely failed. This was after those 'uneducated' pot changes.

I am somewhat doubtful that the SCR on the MTI-1 board is used for fast switching, though. I expect that it's a high current switch that can deal with turning on and off during starting, braking, and reversing. We'll see.

I'm going to replace the SCR (already on it's way), go through George's KTK maintenance setup of the pots, and see what the results are. If still no full speed, which I am kind of expecting, I'll pull the motor / tacho to check it.

[keebler303]

Any chance that SCR powers the field winding? Excessive field current could prevent the motor from reaching full speed and cause overcurrent trips.

Matt

[mflux_gamblej]

Any chance that SCR powers the field winding? Excessive field current could prevent the motor from reaching full speed and cause overcurrent trips.

Matt

It is directly involved in the field control. Its the only SCR on the MTI1 board, shown in the attached (PDF) schematic. It is labeled TH1 BTW68 1200N. It's connected to 415V AC incoming to the MTI-1. I suppose it might rectify a single phase if you switch it fast enough. The gate is controlled by an amplifier and what looks like a triac.

I don't quite understand the incomplete fault, though. Why would it allow the motor to get to 2800 rpm but not higher?

It definately explains why the SCR is shorted now, after I messed with the max field pot, but I'm not sold on it being the reason why the motor speed was limited.

It does, though, fit with the overheating behavior. Since the O/L trips whether the motor is running or not, and then behaves after the machine is off, cooling, for a while, it could be that this SCR is the culprit (how nice would that be! a ~$6 repair).


Having thought about it a bit, I believe the motor field is powered whether the spindle is turning or not, which might explain the O/L that does not trip the actual circuit breaker O/L. Although, it could also be any old circuit getting hot I suppose...

I am pretty irritated with this machine right now, and it's damn MTI-1 board. It's got to be one of the worst after-thoughts of a custom motor drive interface board ever made.

[keebler303]

I don't know enough about the physics to explain it, but excessive field strength will hamper top speed. Google "field weakening" and you will find lots of info.

From Wikipedia: "Field weakening is used in some electronic controls to increase the top speed of an electric vehicle. The simplest form uses a contactor and field-weakening resistor; the electronic control monitors the motor current and switches the field weakening resistor into circuit when the motor current reduces below a preset value (this will be when the motor is at its full design speed). Once the resistor is in circuit, the motor will increase speed above its normal speed at its rated voltage. When motor current increases, the control will disconnect the resistor and low speed torque is made available."

Brushed DC electric motor - Wikipedia, the free encyclopedia

Matt

[Al_The_Man]

Higher field strength allows higher torque, but lower maxim speed for a given armature voltage, weakening the field will allow for higher rpm at the cost of torque.
Also if the field happens to open circuit while running, the motor can speed up and destroy itself if field loss protection is not in place.
Al.

[mflux_gamblej]

The BPort Interact series operator's manual does say that the brake switch is either off or auto modes.

That means that my machine is broken. Either the TNC, a relay, or the previous owner by-passed it (holy sh*t!).

So, once I've gotten my crashed machine up and running again, I get to solve yet another mystery. I merely want to run toolpaths into some metal! Why is it so difficult!?