Don Kleinschnitz , Anthony Bolgar , Yuusuf Sallahuddin , Peter van der Walt .

@donkjr , @funinthefalls , @Yuusuf_Sallahuddin_Y , +Peter van der Walt . Here’s my first (incomplete) crack at a “Fisher Price” wiring schematic of the smoothie setup. It’s based on the information (one wire, no level shifter) I’ve seen posted. I would gladly incorporate your comments and ideas. I could also make a separate one for the other main LPS type. I’d need some photos of it, though.

EDIT: See updated schematic below.
missing/deleted image from Google+

Great job. Crystal clear. I thought @donkjr ​ suggested not to use the PSU 5v.

@donkjr , is the ground wire going to the power terminal between P2.5 and P2.7 required, or is one ground wire from the LPS to the main smoothie power connector sufficient?

Also, there’s an error in my labels on the LPS. The wires in the top left corner should be labeled L-, FG, AC, AC with FG going to your house ground.

I’ll wait for your comments before reposting a new image.

Thanks, @Ariel_Yahni_UniKpty ! If we’re not supposed to use the +5V from the LPS, it makes perfect sense to have a second ground. Why wouldn’t you use the ground on the main power connector, though?

@Ulf_Stahmer ​ this is what I last saw from him https://plus.google.com/113684285877323403487/posts/99hJJ1Vw32P

In my case I use another board but I power my board from the PSU 24v/GND and just L to the PIN

Yes, thanks @Ariel_Yahni_UniKpty . That’s what I used as my reference as I have another board, too :slight_smile:

@Ulf_Stahmer ​ still your drawing is very clear, just I would change the format from png to something else as you loose the background that the black line seems to disappear

Will do in the next revision.

@Ulf_Stahmer ​​​ theoretically using the power terminal ground will work. Especially when you are powering the smoothie from the lps 24v.

To avoid gnd loops I always accompany signals with their sources gnd. Hence L and gnd in one cable.

This may or may not be important dependent on the supply configuration (external or LPS).

For that reason I would leave it in (two wire) so it’s independent of the PS config. I.e. like my original dwg.

I do not recommend using the DC from the LPS as I think the capacity would be marginal. Someone needs to calculate a budget from actual measures to be sure. If these supplies droop in operation weird motor and control problems can occur.

Looks like we need separate Power Wiring Fisher Price drawings for various PS configs.

Sorry for delay responding …on vacation :slight_smile:

@donkjr ​ If I understand your last statement, if I connect my board directly from the LPS could it impact the power to the tube?

So if I understand you correctly @donkjr , you are recommending a completely separate power supply for the smoothie. Then the only wires from the LPS to the smoothie would be the PWM from L to P2.4 and the ground from the LPS to the smoothie to avoid the ground loops. Have I got it right? Can you recommend a power supply for the smoothie? I imagine that I would install it in the K40 case and run it off the AC available inside.

I wont be much help on this issue, other than to say if the diagram makes sense to a total electronics noob.

From looking at the pic it at least makes sense to me :slight_smile: Only thing that I wonder on is the G wire going to 2 places… as far as I am aware it’s only going to 1 place on my smoothie & the L is not going where you are sending it… I’m using @Scott_Marshall 's ACR though for simplifying the whole lot (since I have no idea what I’m doing).

Scott might be able to weigh in on this discussion when he gets time.

@Yuusuf_Sallahuddin_Y this is not a configuration intended for use with ACR.

@Yuusuf_Sallahuddin_Y , as @donkjr pointed out, if you are using the LPS to power the smoothie, only one ground is needed. I based my schematic on his post https://plus.google.com/113684285877323403487/posts/99hJJ1Vw32P in which he uses a separate power supply to power the smoothie (I wasn’t aware of that when I drew the schematic). In this case, the ground wires of both power supplies need to be connected to the smoothie to ensure a common ground between them. I will update the schematic showing both configurations.
https://plus.google.com/113684285877323403487/posts/99hJJ1Vw32P

What’s up with the ground wire that’s connected to the large mosfet?

According to the smoothie layout, it’s a “power in” between the two big mosfets. Don is just using if for a common ground. I’ll update the schematic showing both one and two power supply options.

@Ulf_Stahmer ideally I would reccomend a cable that had L and gnd from smoothie to the LPS any time you are wiring the L signal directly to the supply. Although it may seem redundant it helps to eliminate gnd loops.

I discovered a couple of weeks ago that the 2.4 MOSFET is grounded (I was writing an article for Ariel Yahni’s Blog on different ways of triggering the Laser and had one of thode “doc brown” on the toilet moments)

The K40 PSU has the positive side of it’s optical isolator tied to +5v internally, so all you have to do to fire the laser is pull it to ground. The current to do this is very low since the +5v has already satisfied the Optoisolator LEDs forward drop requirement. (I did it with a wet finger once just to see how sensitive it really is - it fires)
If you hook to the - pin of PW0. (it’s the drain of Q8), Q8 pulls it to ground, firing the laser. The gate of Q8 is already tied to pin 2.4 internally.

The downside of this is the lack of a positive shutdown. There’s nothing preventing the pin from going low and firing the laser. That’s just not safe.

It’s works, but it’s not considered a good engineering practice beacause of this.

With safety such a serious issue, and the fact that the Smoothieboard leaves the line uncontrolled during it’s boot period, I’m not comfortable telling people to do something you’re taught in engineering school isn’t a safe design practice. Leaving a logic input floating is a “Never do this” example that’s even shown in the TI TTL Data Book.
You can improve the safety level by using a nice low value pullup resistor on the line. (the FET can handle a bit of current no problem)

With a good quality level shifter coming in at under 2 bucks, and the few wires needed to run it, I think it’s a safer option.

While the FET hookup works, the potential for an unexpected discharge exists and considering the low quality of the K40 PSU and controls, I just can’t feel ok with recommending a circuit that adds the the risk of someone getting hit by a laser or shocked by a power supply they thought was off. It’s just not worth the 2 or 3 bucks and 5 minutes of wiring you save to do it right.

If you use this method, be aware, it’s akin to leaving out the neutral safety switch on your family car. While it might not happen right away, sooner or later somebody who doesn’t even know what a neutral safety switch is will start the car without pushing in the clutch and run into or over something.

It’s just not worth the risk.

In interest of full disclosure, I do sell products that compete with this arrangement, but have never witheld information or assistance from anyone ‘rolling their own’, and in fact rather enjoy helping people get their lasers running. I’ve been a hobbist since before I was 12 and have been building homemade stuff since. (I was granted a General License at 13) and worked at a TV repair shop by 14.
I truly support the do-it yourselfers, but don’t want someone to assemble a hazardous device because they saw it done on the internet.

I recently started making a plug in level shifter which uses an optiisolator and MOSFET to link the Smoothieboard to the K40PSU. It’s self-powered, offers 4500v of isolation, and is plug and play.

This device (and my new ACR Mark II) use positive logic, pulling the K40 PSU to +5V anytime the signal is not positively controlled (Like during the boot period of the Smoothieboard)
It requires that you remove the ! from the Pin designator in the Config file, but it’s worth it.
The design requires power to fire, so is fail-safe, unlike negative fire systems (including th stock K40 arrangement)

I’ll send anyone who asks for it a drawing on my shifter and a copy of the EZ~Config V6.0 which has the change implimented. Just send me your email.

My level shifters are also available as a kit. (they’ve been delayed slightly because I screwed up the board design by working when I was too tired to pay attention)

Please be careful out there, I never want to come onto our forum to find someone has been blinded or worse.

Scott

@Scott_Marshall what TTL level are you suggesting is floating. Non of these interfaces are TTL and none floating that I can see?
The L pin as you suggest is the cathode side of the optical isolators led and is connected to Q8’s drain. When Q8 is off it provides a high impedance to ground and no current flows in the led but it’s not floating. When it is turned on it is low impedance and current flows.
The led anode is tied to 5v through a 1k resistor which means that 5ma flows when Q8 conducts.

I don’t see the problem yet?

Certainly any time a low enough impedance is connected to the L pin directly it will fire no matter what way it’s connected.

That said …
If the smoothie uncontrolably turns on Q8’s gate during reset transitions that is a problem. Note: Q8 has a pulldown on its gate.
I don’t see how a level shifter ( the sparkfun is a source follower) would resolve that since an ON glitch would also propagate the shifter.

I haven’t tested resets yet however if smoothie allows drivers to be in random ON states during reset then there could be other unsafe condition like motors moving.

Failsafe systems that use processors do not usualy rely on a single pin. The ideal way is for the enable of the PS to be held off during resets and during power down with a por signal…

In this configuration there are two levels of protection since I am assuming interlocks are employed. To get hurt by a random PWM pulse, the enable switch has to be engaged and interlocks have to be disabled.
If your not using interlocks all bets are off… because there are multiple ways this machine can be unsafe.