Failure when mounted sounds a lot like a broken trace on the PCB or inside a part. Once I traced a failure to a diode with one lead loose inside the plastic.
So now it is time to grab a strong magnifier and a bright light and figure what area gets flexed when the board is mounted. AND still, it may be a failed solder connection.

 

I didn't see it right away, but the FWD/REV function is done through the 3P3T switch on the front panel.

Makes sense!

 

I was thinking about adding a power switch to this unit. It actually doesn't have one; the control board is powered as long as the cord is plugged in. The board not having transformer isolation scares me...Can I put the switch in the Hot line? Seems like sending the HOT wire to a front panel switch is dangerous.

A reminder: The Hot and Neutral come in from mains to a terminal block on the board and are rectified to DC with four diodes. There is no mains isolation.

 

Technically you should have a proper E-stop circuit, generally the circuit is LV and in the event of a E-stop, the main power is disconnected via a suitable relay or contactor, In your case, in the supply.
The start button takes it out of E-stop
In large systems, it is just the motorized mechanisms that are Estopped

 

I can switch the motor off with the REV/STOP/FWD switch, but I'd like to also have a power switch so that it's not always on. Currently, I have to plug/unplug the power feed every time I use the mill.

 

I can switch the motor off with the REV/STOP/FWD switch, but I'd like to also have a power switch so that it's not always on. Currently, I have to plug/unplug the power feed every time I use the mill.

Hence the E-stop circuit, provides two uses, Safety stop and periodic stop/start.

 

I just looked at the schematic for my bench power supply and the power switch is on the AC Hot line. Any reason I can't do the same thing for the transformerless power feed board?

 

I can switch the motor off with the REV/STOP/FWD switch, but I'd like to also have a power switch so that it's not always on. Currently, I have to plug/unplug the power feed every time I use the mill.

T recommend a double pole switch rated for more than the nameplate current listed for the original unit. So probably ten amps.
The reasons for breaking both sides are: First, when it is off even if the line and neutral got swapped, and second, the circuit is still off even if one contact stuck on.

 

Also you need to ensure that all metallic parts of the machine and system are properly earth grounded for safety.
Especially If this mini-mill is of Chinese origin, and also would explain the lack of E-stop.

 

Both good points. There is no ground line coming in from mains, only Hot and Neutral. So, yeah, they could get swapped. And, actually, they are!...I noticed that the mains Hot wire is connected to the N terminal on the board, and the mains Neutral wire is connected to the L terminal.

The power feed resides in a plastic housing. The two ways the power feed is possibly electrically connected to the mini-mill is that the shaft of the motor connects to the mill table leadscrew and the housing of the motor is connected to a metal plate that fastens to the mill table. The mill motor and cast iron base are grounded through to mains ground.

I'm now concerned about how the power feed could be conductive to the mill, given that there's no transformer isolation. Grounding anything on the power feed side could be problematic, no?

 

Certainly no part of the power circuit should be connected to the mill's frame. I suggest, to see if you have a problem already, to use a voltmeter to check for any voltage between the mil frame sections and the external safety ground. Hopefully there will be no leaked voltage. Then reconnect the safety grounding wire.
I do not recall any "E-Stop" control on any of the mills that I have used, only the regular start/stop buttons. Of course, any milling machine accidents would be a over before a fast hand could even move towards a button, so it does not matter, from where I see it. Mill accidents happen very fast.

 

E-Stops are required by code in N.A. and EU, But many DIY M/C's lack it, also machines of Chinese origin frequently omit it.
The Earth GND should be bonded to all metallic parts of the M/C.
I have never put together a machine without it.

 

Does the PWM rate affect the motor startup time? One thing that has bothered me for a while is that at low speed settings it takes a couple seconds before the motor starts turning. Running the motor from a bench power supply has no lag at all; the motor turns on immediately. Even at 2 volts! I was wondering if this might be due to the somewhat low PWM rate, which is about 1000Hz. I could try put a capacitor across the motor terminals.

 

I assume you mean that when you run from the bench supply it is direct? i.e. NO PWM?
If so , it depends on the PWM rate you have it set for initially.
Does the controller have a built in ramp-up rate?
You need to either 'scope or voltage-measure it.

 

Oh, I didn't think the controller would do that...I'll check.

 

Well, without even checking I can answer this. I'm not turning the controller from Off to On. The controller is already powered on and the motor direction switch is set to Neutral (motor disconnected) and I'm just setting the direction to Forward or Reverse. The PWM is already running for the set speed.

 

To confirm what is happening it would need as i suggested either a 'scope or a meter at least on the motor leads.

 

Okay, I got off my lazy horse and scoped it. Yeah, the motor doesn't receive any PWM for like four seconds after switching to Forward/Reverse. So probably not fixable, huh?

 

Must be a feature of the drive?
Could be some prevention of going directly from FWD to REV?

 

I think it's detecting no load and shuts off. I had to remember: It *is* a switching power supply afterall. This Fairchild Power Switch was(is?) a common part in switching power supplies for products.