Photos 2 & 4 out of 6 - gold balls soldered in place from underneath.
I tips me hat to you for that one. Brilliant.
Cheers
Roger
Photos 2 & 4 out of 6 - gold balls soldered in place from underneath.
I tips me hat to you for that one. Brilliant.
Cheers
Roger
The initial design was using a cage like most (if not all) are using made out of composite (POM I was hoping) but then I was getting pretty severe thermal expansion on those sim models results. Those composites can thermally expand as much as 0.05mm from early morning to having the afternoon sun shining on the probe. So that idea got scrapped pretty quick :/ Turned out to be an easy way out of this issue!
BREAKING NEWS - We Have Contact! Slipped-in a 5P Micro USB connector on the PCB and will machine a pocket on the side of the body with the 4th axis to accommodate this. So... no more wicked wires pocking out of the side of that probe when in storage or during auto-tool change. That'll go in when I've revised the PCB/Circuit and receive the V1.2 of it. Same size...all very minuscule but sexy items. All good fun!
Four signals I plan for that MicroUSB are "0V, 5V, NO, NC" (In case I need to have it set to NC... for whatever reason that might be, good to have it there).
Yes, on the probe switch, you have a 1k resistor to ground. That will work, but it is creating a voltage divider between it and the internal pullup of the UC. That will work, but it reduces the signal noise immunity. With the resistor in series, and the switch closed the signal will be around .5V instead of 0. You might think about removing that resistor completely so that low = 0V and high = 5V.
I have a couple other recommendations if you are interested. Everyone does circuits differently, so this is just me thinking out loud.
1. Use a separate dedicated UC output signal connected outside the probe. With no external cable connected to the switch signal, there is little chance for noise coupling onto the line and tricking the UC about when the switch is open and closed.
2. I liked _britt 's idea of having a second signal to stop over travel. Instead of outputting a NO/NC, use the second signal to alarm for over travel. You could program a timer so that if the probe switch is open for too long, send the alarm.
3. As far as the connector, you could use a 6 pin connector for output of the signals. Maybe a standard SPI programming connector that connects VCC, GND, MISO, MOSI, SCK and RST. This will let you reprogram it while assembled.
4. You could us a SMD LED mounted at the edge of the PCB and a clear lens through the probe body so you can see it on the outside.
I really do like this probe. The small size and the versatility of having a UC inside is the bomb.
Thanks Max, too kind. Have a look at what I have at the moment. It is the simplest circuit ever designed by man-kind, which is the main challenge for right now. We'll expand features later on.
Tossed out the R for the switch, will just use that freebie internal pull-up instead. Dragged back the R on the led's anode, and linked up the remaining available digital pins to that 5P Micro USB (which I have in stock + the footprint spot-on for, used it before). As for what those extra signals (will) do, whether lora/ir/nc/no/whatever, I'll have plenty of time to give them a life later on in the built. You are right by spotting the advantages of that built-in MCU ;-)
I'm in the train for 3 hrs hours tomorrow and the plan is to look at connector alternatives, if any. A 10P type-C or anything else of 'micro' size is what I'm hunting for. The challenge really is to have it full surface mount (no TH) as I may have to do the termination/soldering while the PCB is in the body. We'll see...
All excellent recommendations, btw. I'll consider for sure, let me have a look at what I can do with this.
I'll take you on the idea of replacing this fugly 1966 era LED with a super bright 1206 smd. Now that we have a bit more room to fit stuff on that pcb.
Found that clear cover (PC?) in OD4mm. Snap and whala.
Good idea...
Parts on hand are always better than having to buy new.
What if you were to use the shield pin as the ground? Then you can include all the programming pins on the connector and do away with all the test points completely?
Take a look at the schematic here. If you add a schottky diode between pins 5 and the shield, plus a cap and resistor, you can still connect a USB charging cable to the probe and get the functionality out that you have in your original design.
BTW, I think it is funny, the difference in our schematics. Yours is new school, mine is old school.
This schematic has the required components.
You are an Electrical Engineer, that's why. I just have a nicer/newer/kooler Soldering Station than yours, and that's all ;-)
Let me pin this to a breadboard when I'm back in town early Sat. Looks promising and doing more than I wished for (thanks!)
From a land constraint perspective, looks like that could work flawlessly (that fat horizontal 1206 being that bright'er led).
Maybe newer and kooler, but my reflow station can heat up a hot pocket lunch.
Heartless, how can you show that to someone who has to blow away components with his hot air/soldering station to get a pcb done.
I'll let you off for this one, but just because you've been more than kind to help me out with the design
That is one kool (and useful) reflow kit you got there. You win this round, I'm jealous
Oh, maxspongebob... I do hope you've gone 100% lead free!
So... you've got the right reflow profile programmed in for mozzarella, right?
Mecanix - If you do use a USB connector but have an incompatible pinout... to prevent unfortunate accidents, be sure to mark clearly on it 'NOT USB'... perhaps with the USB logo in one of those universal "no" circles.
Edited to remove non-functional unicode symbol... the forum ate it.
Meh, lead free solder sucks. I may not be able to use it for work but I sure as hell gonna keep using it for home projects.
Good point on the labelling of the USB. Better yet, set it up so at least the power lines are on USB power and ground pins and your NO/NC outputs are electrically & non destructively compatible with the USB data lines.
Lead free? never heard
Come'on guys... who plugs machine/digitizing probes into their laptop or phones. You mean that USB cable that should be sitting near the spindle that comes out of the bob right? That'll be VBUS_5V and GND_0V only, no logic passing on the data pins so it should be safe (should lol).
Subscribed, for the day the inevitable post comes through "I done blowed up my mate's phone / controller / something".
I think the banana plug idea was Roger's, and more for just the tip (and only for a minute) as an alternative to the balls.
Don't get me wrong - the form factor is great. And now some worry-warts on a forum have highlighted the potential for problem you'll never have an issue because you're aware of it. But I'd still be a little nervous.
Nah, no problem. Folks that bloody dumb not allowed access to the shop floor anyway, so we're safe here!
'Banana', re-read thread... thx for the correction.
A standard USB uses pin 1 for power (+5v) and pin 5 for ground. If you use pins 2-5 as the programming pins and the shield for ground, then you can get by with just this 5 pin connector to program and during operation. If you look at the schematic, you can see that I put a diode and a cap inline with pin 5 so that it can be both the reset pin for programming, and the ground pin if you have a USB cable plugged in. So, no need to mark the connector because it will be fine if you plug in a USB cable. It should work if you use a shottky diode. I have not tested the circuit yet, however.
BTW, I am proud of myself for successfully not talking about gold balls and banana plugs for most of this thread. DOAHHH. There it is.
Some forum members might already have modified their android mobby pinouts? got to be careful
Could swing the VBUS/5V to pin 3 and GND to pin 1, if this can entertain cnczone a bit
ps: had my vendor to send me a baggy of S4. I can only breadboard this weekend... I swear I looked everywhere for my shotties and couldn't find them... just did a pwm solar panel/battery trickle for my car only a month ago, had my diode storage stolen, Max.
Yes, it does. However, he was talking about cooking lunch along with his latest batch of PCBs, so...Meh, lead free solder sucks.
I'm a software guy, really... and I'm afraid that's something of an occupational hazard... if I don't think of every possible result and write code to guard against it, sooner or later it will come back to bite me.problem you'll never have an issue because you're aware of it.
Somebody who shouldn't have been messing with it... but was anyway.Come'on guys... who plugs machine/digitizing probes into their laptop or phones.
I don't know who goes thru your shop, but if you ever get a new hire, and for whatever reason the probe isn't working, they might think "well, perhaps it needs to be charged". Or perhaps they are just curious to see what it enumerates as when plugged into a laptop. Personally, I've never actually seen a USB port "in the wild" (so to speak) that wasn't either a charging port or a data port (and often, both)... so having one on there might well set up an unfortunate expectation in someone who doesn't know any better.
Eh... I figure it's like carrying an umbrella... every time I take one, it never rains... but every time say "forget it" and leave it behind, I get soaked...
FWIW, Mini & Micro USB - yes. Old fashioned A & B are only 4 pins. (yes, I know that you know that, but I'm just making it clear to anybody else who reads this).A standard USB uses pin 1 for power (+5v) and pin 5 for ground.
Crimeny... I'm getting old. I just called USB "old fashioned"... and I clearly remember when it first came out.