Review of schematic

Projects
126

Revised version of the schematic:

  1. Addition of the MOSFET (PHP79NQ08LT)
  2. Added a temperature sensor (DS18S20). GPIO4 seems to be the input for the 1 wire connections on the RPi.
  3. Changed F1 (polyfuse) value to trip at 10A . When selecting the fuse, the trip current is 10A but the fuse is rated as 5A. This does not seem to make sense to me…perhaps I’m reading it wrong?
  4. Changed inputs on some circuits to 3V3.

Thanks

Adam

saw_V7.pdf (95.5 KB)

127

You have selected a wrong symbol for the MOSFET. The symbol you use has D@3 while the PHP… has D@2. That’s what I meant when alerting you to be careful which symbol you choose.

You need the IRLB8721PBF as a symbol. Note that the symbol has a footprint for vertical mounting pre-assigned. In case you want to mount the PHP… flat on the board, you have to assign a TO-220 footprint instead.

Also note that the PHP… has D on the tab, so be careful with this later when you do the board in case of horizontal (flat on the board) mounting.

128

Thanks. Will change.

Adam

129

The trip current is the current at which the fuse is guaranteed to trigger. The hold current is the current the fuse is guaranteed to sustain without being triggered. But this info and more is readily available when Googling “polyfuse function”.

2 Likes
130

It’s not going to work very well wired like that.

No you haven’t. There is no connection to 3V3.

131

Hi Rob,

Thanks. Attached is a revised schematic.

All the best,saw_v8.pdf (100.3 KB)

Adam

132

If you mirrored or edited the temp sensor it would clean up the appearance a little. You wouldn’t have the traces crossing. The pins don’t have to be in order for the schematic.

133

You need to fix up your reference ids but other than that I don’t see any problems at the moment.

A couple of comments:

  1. I don’t know if you plan to mount the temperature sensor directly on the board but you could have one or more three pin connectors to allow multiple off board sensors, such as on each motor. You could add the connectors or at least footprints for future use. The DS18S20 allows having multiple sensors in parallel so you would only need the one GPIO.
  2. Those fault outputs from the stepper drivers are open drain so you could tie them all together and connect them to a GPIO.
  3. I would move the LED on the incoming power to after the diode (D2).

Otherwise fix up the reference ids and run DRC and if it’s all good start planning your board.

134

If you mirrored or edited the temp sensor it would clean up the appearance a little. You wouldn’t have the traces crossing. The pins don’t have to be in order for the schematic.

Thanks Hermit. Done.

  1. I don’t know if you plan to mount the temperature sensor directly on the board but you could have one or more three pin connectors to allow multiple off board sensors, such as on each motor. You could add the connectors or at least footprints for future use. The DS18S20 allows having multiple sensors in parallel so you would only need the one GPIO.

Great idea. I’d like to use one to monitor the temp of the motor driving the saw. I’ve seen these as remote sensors enclosed and with connections.

I was also wondering if there is a good way to determine if the saw motor (1/3 HP AC LEESON ) has stalled (not turning when it should be). My first idea was based on magnetic sensors to make sure the drive shaft is turning. Another was based on current (which I assumed would be high) if the saw stalled. Monitoring shaft speed seems direct and its a single speed saw.

2) Those fault outputs from the stepper drivers are open drain so you could tie them all together and connect them to a GPIO.

Is that to monitor them? I could do that or do they simply need to be connected to ground with through a 100k resistor?

3) I would move the LED on the incoming power to after the diode (D2).

Done

135

Changes:

  1. Added 2 connections for 2 remote relays.
  2. Added 2 additional connections for remote temperature sensors
  3. Added weak pull up for fault pins

Q - I read up on the fault pin and the logic level is “low” if there is a fault. I found it could be weakly “pulled up” (not sure if this resets after an error?). Does pulling it up prevent it from floating to a fault state and impeding the operation of the driver chip? FYI - I’m not doing this at present on my prototypes and they seem to work.

saw_v9.pdf (120.7 KB)

Thanks

Adam

136

Yes, when a motor stalls the amps go up. It would be quite surprising if the motor wasn’t internally protected to shut off in that case though. So, after a surge of high amps it should go to zero quite quickly. Ideally you want the saw on an independent circuit with appropriate breaker as a backup to this. I’m not sure what you want to final setup to be but you could have a small breaker box with the electronics in a box tied to it. You could use the knockout plugs and some conduit fittings to make it one unit.

137

Thanks. I have something like this set up but will add a breaker just for the saw (instead of the entire machine).

138

You might want to double check them.

I meant to only connect them together if you wanted to connect them to one of the Pi’s GPIO pins in order to monitor them. Ideally you would connect each to it’s own GPIO but since you seem to be running out I suggested you tie them together so you could at least detect a fault even if you can’t detect which driver caused the fault. Not need for a pull up, if you connect it to the Pi then the Pi can provide the pull up, if not then just leave them unconnected.

Your power LED is not connected. Give the zener a Dn reference id. I usually give LEDs a Dn reference as well but that’s just a personal preference.

139

Thanks again Rob.

  1. I changed the remote relays to be driven by 24v. Used the same MOSFET based circuit as above.
  2. Rewired LEDs on remote relays (thanks)
  3. Changed reference on zener

saw_v10.pdf (116.6 KB)

All the best,

Adam

140

For every step forward you take two steps back.

By Dn I meant replace n with a number, ie. D1, D2 …

You have more than one zener.

Your FET circuits on your relays will not work. You’re better off using 5V. Why do you want two remote relays, you already have two free ones on the board?

141

Hi Rob,

The remote relays are b/c I wanted to keep the high voltage systems separate from the low voltage (the internal housings are also located in different areas of the saw). When I built my first prototype I found some interference from the AC current as well. Also makes the housings cleaner and less cluttered and fits the saw internals better.

I may be missing something but the FETS are driven by 5V from the UNL2803A and the 24V is sent to the relay coil. I wanted to avoid having a hot wire through the coil and waiting for the ground to open. I figured I could use the other side of the relay coil connected to ground as the ground. Do the FETs have to be placed as a switch opening and closing flow to ground after the laod?

All the best,

Adam

142

Hi Rob,

Appreciate the patience. Founds the error in the weak pull down connection which should have been to the gate and not the after the source.

If my thinking is correct the 2 versions below should work…

791×701 127 KB

Thanks

Adam

143

The ULN2803 sinks current to ground through NPN transistors, not 5V. The FET needs a positive voltage on the gate relative to the “source”, how can your source switch 24V in that case? If you used a pull up resistor to pull the gate up to 5V and let the ULN2803 turn it off it would mean the FET would turn on every time you powered up until the Pi booted and turned them off. Not ideal. You need to use the same circuit you used for the fans and drive the gate with a Pi GPIO pin.

144

Revised version with the FETs connected to the GPIOs

Thanks!

saw_V11.pdf (122.1 KB)

Adam

145

You still haven’t changed the ref id of the second zener or the regulator.

LED9 will not work as it is. Remember the ULN2803 provides a connection to ground.

To keep your LEDs a consistent brightness (assuming you use all the same LEDs) the current limiting resistor for LEDs powered by 5V should be 1K and the resistors for the LEDs powered by 24V should be 4K7. All LEDs will then get about 5ma.

All of your connectors have a ref id starting with J except for the Pause Button which starts with P. Not that it matters, you might have done that for a reason, if not you might want to be consistent.