People at Lynxmotion, I would like to recommend that you add some sort of reverse polarity protection to your SSC-32. It could be as simple as putting a diode in series with the input power lines. This would make people a lot more comfortable working with the board, and save people a lot of $$$ if they fry it. I’m sure there are many people out there (including me) who have fried an SSC-32 board by accidentally connecting reverse power. If you don’t add a polarity protection, could you please at least add something to your documentation on how to repair the board (which components are most likely damaged, how to replace them, etc.)?
I can think of a couple of reasons for NOT doing this:
The typical forward voltage drop of a diode is 0.6-0.7 volts. The 5 volt regulator on the SSC-32 and BBII is the LM2937. The delta between the input voltage and output voltage is 0.5 volts. Add to this the 0.6 volts of the diode and now you need at least 6.1 volts from your battery before the voltage regulator quits working. If you were to use the servo battery to supply VL with power, every time you servos pulled enough current to cause a dip in the battery voltage (below 6.1 volts), you’d reset the processor(s).
Since each side of the SSC-32 can deliver up to 15 amps of current, you’d need diodes capable of handling more than that. A 15 amp diode is pretty big to start with, and you want to add (2) of them, plus and additional diode for VL( a smaller diode, but still an additional part) - in case you used separate supplies, like I do. This will most likely add to the length of the board(s) to make room for them. Now what you did is make the new boards incompatible with ALL the existing stuff that Lynxmotion sells. That would mean redesigning everything ($$$$)!. The problems just start there.
The voltage regulator on the ssc-32 is not rated for 15a. Below is from the user’s manual. The problem appears to happen when polarity is reversed to the VL regulator and not to the VS busses, which are rated for ~15a.
The Low Dropout regulator will provide 5vdc out with as little as 5.5vdc coming in. This is important when operating your robot from a battery. It can accept a maximumof 9vdc in. The regulator is rated for 500mA, but we are de-rating it to 250mA to prevent the regulator from getting too hot.
You are somewhat correct with the inaccuracies I stated. I was merely trying to curtail the thread because, well, lets face it; its not going to happen. I know sometimes things happen and then we have to pay for it. I’ve had to buy things a second time because of mistakes I’ve made, but its all part of the learning process.
As for my erroneous info:
Yes, if I were to re-layout the board and wanted to add the protection diodes, I would probably use Schottky diodes.
Yes, you are correct. The the voltage regulator for the SSC-32 AND the BBII will NOT handle 15 amps. I wasn’t trying to infer that it was, but after rereading that section I see how you could come to that conclusion. It is only a 500mA part when properly heatsinked. But, like you stated, Lynxmotion limits the current to 250mA because it is not heatsinked.
You are misinformed yourself on the max input voltage. According to the datasheet here:national.com/mpf/LM/LM2937.html , the max input voltage is 26 volts. But, understandably, Lynxmotion says the max input voltage is 9 volts to keep the power dissipation low enough to run it without a heatsink.
9 - 5.5 = 3.5 volts. 3.5 volts times 250mA = 0.875 watts. Makes sense to me.
zoomkat, I’m NOT trying to get in an argument with you and will always welcome your comments, as I hope you’ll continue to do mine.
If you had read the ssc-32 owners manual (and as noted in my post) you would see I just copy/pasted from the ssc-32 owner’s manual and made no statement of my own as to any input voltage values. Generally speaking 250ma is 250ma no matter weather the supply voltage is 7v or 26v.
I was merely stating WHY those numbers were chosen.
And the voltage level DOES make a difference, if your going to fix the current at a specific level(vice-versa as well). P=EI, familiar with it?
Jim could of said 12 volts max, but he didn’t. I wonder why? It’s a pretty common voltage. Maybe because the voltage regulator would be dissipating 1.625 watts!
Per the data sheet below, the 5v regulator chip should generally be bullet proof (over current and over temperture protected, reverse polarity protected to 30v). If you look at the schematic of the ssc-32 board in the owners manual, there appears to be electrolytic capacitors on the VCC side of the regulator. C12 appears to be rated for 10v, which may be the real reason why 9v was chosen as a max VCC voltage. The caps may also be what is damaged during a reverse polarity hookup and not the regulator chip.
Yea, I read that the regulator is reverse polarity protected too. Makes you wonder why PrinterFace’s board was damaged? I wonder if it passed the (reverse polarity) voltage on thru? Not a very part if it does.
I looked at C12 on the SSC-32 schematic and noted that it is on the OUTPUT side of the regulator, which (for all practical purposes) has nothing to do with the input voltage. So if the regulator does its job correctly, C12 could be rated at 6V and be just fine. Looking at my SSC-32, C12 is rated at 20V.
some people reversing polarity shouldn’t end up costing all of us more money, however better attachment would be nice, i’ve had wire come out of the screw terminals just slightly and begin to touch which caused my battery to smoke and melt some of the plastic covering…no big deal, its pretty difficult to get the wires in to begin with since they cant be crimped and there is very little separating the 2 wires which made it kinda dangerous to people like me but that’s just my input
I hear ya brother. I think bigger/stronger screw terminals are much needed. A lot of today’s battery packs come with 14 gauge leads, and its darn near impossible to get all the wire strands in w/o one or two hanging out. Even the 18 gauge wires on the switches they supply with some of the kits, you have a hard time getting all the strands in. The other problem with the little screw terminals is that because they are so small, you have a hard time getting enough torque or grip to keep the bigger gauge wire in.
It just takes a little finesse, that’s all. The BBII and SSC-32 are GREAT products and I’m glad they are around.
Unless you are using two batteries for servo power, you can install the ‘VS1=VS2’ jumper and put the B+ on VS1(+) and B- on VS2(-), or vice-versa. Just something to think about.
I don’t know exactly what happened to the board. if you want more details on my situation (or have a solution), you can check out the thread at lynxmotion.net/phpbb/viewtop … highlight=
core2, you have made some good points against the idea. firstly:
that is true, and I agree. i have fried things too and it has helped me learn. but this is an expensive board, and i don’t think that learning about polarity is the best use for it. bur if you really want them to learn by giving 'em smoke, you could put the diode in parallel with the inputs, so that if they reverse the power, it will just short to ground thru the diode . that would also clear up the voltage loss issue.
that’s true too. one alternative solution i can think of is to have a jumper that would bypass the diode so if people really need to they can just jumper it. Also, on the issue of board space, I can’t imagine a diode taking up that much space.
I have a Stamp Stack II, which is pretty much a mini version of the Basic Stamp II (a great microcontroller, by the way). it has a wonderful polarity protection system that works really smoothly. it doesn’t have to be as nice as that, but this is the general idea (minus the LED, and plus a jumper to bypass it). it seems to use some sort of specialized 3 pin device instead of (or containing) a diode. check it out. hvwtech.com/products_view.asp?ProductID=42
and finally, core2, i agree that it is unlikely that anything will come of this thread in the way of something actually happening. but, you never know, it might.
It wasn’t my intent to step on any toes. I apologize if I did!
You would think adding a diode would be a trivial thing, and in this case I’m sure it would be. But, I have laid out enough boards to know otherwise.
I’d like to see bigger screw terminals. If the terminals were bigger, it would cut down(or eliminate) the number wire strands hanging out, and possibly shorting together, because they all didn’t fit in the little slot.
yes, core2. i agree with what you said about the screw terminals (they were partly at fault with my polarity mistake). so let’s see what ends up coming out of the poll and thanks for your logical points.
Here’s the deal… We already have too many support issues due to processor brownout. I will never add a “feature” that will inevitably cause this to get a lot worse. Yes it’s simple to attach a 9vdc batter to eliminate the brownout problem, but no one does it, even when it’s made very clear in the assembly guides and tutorials. The 9v input de-rating was chosen to prevent the regulator from heating up. Because we de-rated the regulator we were able to use a cheaper lower voltage cap, so in a way you were both right. The SSC-32 has a very tight layout. We do plan to try to incorporate larger terminals in the next run, but I don’t know if it can be done yet.
Check your wiring, then check it again, then have a friend check it… Then apply power.
but that’s just my input