I made up the 3 ribbon cables today for the 3 stepper motor driver boards. As I did so, it occurred to me that the cables connected these boards to the motherboard, and that as the cables had the same connectors on either end, the motherboard must have an IDC header, which is the part I ordered singly, instead of a set of 3.
Sooooooooo, I nicked the header from the motherboard kit in order to make up the second board.
On positioning the parts for the second stepper motor driver board, I was careful to orient the chip correctly!
When I cooked the second board, I did so withOUT the large electrolytic capacitors in place.
I cooked on medium to heat the board up and then as soon as the solder paste started to melt I turned it up to medium high. All told, it took just over 3 minutes. Nothing moved at all. There was distinctly more smell this time - last time was almost smell-free.
On cooling, I checked all the contacts from the chip (which was perfect, with no bridges at all) along the traces as drawn on the board. I removed the solder balls with the back of the knife blade again, and checked the resistance across all of the solder mount resistors from solder to solder. All was well. I had just one capacitor that had made good electrical connection but was not good mechanically, having far too little solder. I put the tiniest bit of paste on that end and used the soldering iron with chisel tip to make it good. I also used paste to solder the electrolytic capacitors in place, carefully positioning C11 as far away from the end of the chip as possible whilst leaving exposed solder at the other end of the capacitor to facilitate soldering.
I then ordered 2 more headers for the next board and the motherboard, and a replacement driver chip from Farnell.
I soldered in the stand-in giant resistors, and the through-hole components (with the nicked IDC header), and inspected and tested everything (that I could follow and understand).
JOY! This time, no drama, and the power LED lit up. Yes!
Now I have to wait for my next Farnell delivery before I can do the third board.
Monday, 21 September 2009
Verdict: death by misadventure
"He died instantly, they say. No pain. He didn't stand a chance, the poor chip."
It seems the first stepper motor driver board's chip didn't stand a chance (and there may have been nothing wrong with the soldering after all). On inspecting the first board for damage, I noticed the semicircle drawing at one end of the chip's print on the board, and realised this was probably supposed to align with the dimple at one end of the chip. Unfortunately, being a complete electronics novice, and being unaware of the marking/importance, I had the chip the other way around. There had been nothing in the instructions to be careful about the chip's orientation. Guess we are supposed to know that. Well, I do now! (Funny how everything else was specified, though.)
I live and learn.
It seems the first stepper motor driver board's chip didn't stand a chance (and there may have been nothing wrong with the soldering after all). On inspecting the first board for damage, I noticed the semicircle drawing at one end of the chip's print on the board, and realised this was probably supposed to align with the dimple at one end of the chip. Unfortunately, being a complete electronics novice, and being unaware of the marking/importance, I had the chip the other way around. There had been nothing in the instructions to be careful about the chip's orientation. Guess we are supposed to know that. Well, I do now! (Funny how everything else was specified, though.)
I live and learn.
Sunday, 20 September 2009
Rice Krispies
Apparently, Churchill said, "Success is going from failure to failure without a loss of enthusiasm".
I must be heading for success then!
Firstly, I must thank Nophead for sending me those spare parts - I have practice material! More later.
The instructions for the stepper motor driver board say to plug it in to see if it works, but I didn't have a 12V power supply. Maplin, as usual, didn't have what I wanted, so I ordered from Farnell again and ordered spare surface-mount LEDS, too. I was underwhelmed to find the power supply didn't come with a lead! I dug around in the long-since-abandoned-shelving upstairs, and found the power lead for either my old PC or its CRT monitor. It had a 5 A fuse, so that seemed OK. Plugged it into the power supply - and nothing. Checked all the switches were on and the socket was live, all OK. But no power from the PSU. Wondered about that old lead? Then I remembered somebody telling me something about how ATX power supplies differed from AT power supplies, so some internet hunting later, and I found out how to short out pins in the power header for testing purposes.
Usually the power header is a 20 pin affair, but mine looks bigger. It has a separate 4 pin block slid* onto the end of the larger connector.
Here is a picture, with the power header shorted between pins14 and 17:
Thank goodness for the internet and the wealth of information available on various websites.
This done, I switched back on - and it was a good job I was watching the fan closely, because I watched it disappear as it spun at speed, but I couldn't hear a thing. Wow, I know it is called a "silent" PSU, but I didn't expect that they meant it!
Barney reported a blown chip, with the advice to check continuity on the board before plugging in. My meter has dead batteries....
Now I had a working PSU, I switched off, plugged one of the Molex-type leads into the board until it snapped into place and switched on with my finger over the off switch. Here was the crackle and pop. I switched off very quickly with a slight burning smell developing. What was that, you said, Barney? Something about checking continuity everywhere?
Ooops, I have given the chip a contour! but it doesn't look like the board is damaged.
This time following Barney's advice (!) I shall nip the legs with the tin-snips if I can to remove the dead chip. And I'm going to Screwfix now for some replacement meter batteries - open on a Sunday and ~£2 cheaper in total than Halfords.
Then I shall inspect the rest of the board, and CHECK CONTINUITY everywhere. Without the chip in place, I should get better access. Then, if all is well, I shall replace the chip and check the continuity from its legs, too.
Soldering/desoldering learning.
Following on from a link off the forum for Kapton tape supplied by DealExtreme in the States, and looking to see what else they had, I noted that they had desolder braid with lots of reviews, including one that said this particular desolder braid had flux in it, so you don't need to add any. Ah-ha, maybe that's why my desolder braid isn't very successful for me, as perhaps it doesn't contain flux, and I wasn't adding any.
Order from CPC (another part of the Farnell empire) a flux pen. Also a conductive pen as a temporary fix in case I blow any of my board tracks.
I shall try the fluxing/desolder braid to remove the remnants of the old chip legs. If I'm not getting anywhere I shall leave them in place rather than risk damaging the board. Then I'll clean the area with surgical spirit, in case there is contamination after the smokey smell.
Also, I have seen a video of soldering 1206 parts by hand with solder wire, and this requires flux to be put down first and touched with the soldering iron to "tack" the part in place, before laying solder wire partly along the end contact of a component, and just bringing the soldering iron (using a chisel tip, with the flat face downwards) down onto the edge of the pad, sliding it in to touch the part and sliding straight back away. This gives minimum iron-exposure to the 1206 component. I'm going to practise this technique when I have my flux pen. I suspect my 0.7 mm solder wire may be too thick, though; it needs more like 0.4 mm.
I have already practised using the chisel tip and soldering a 1206 component using the solder paste, which itself tacks the part down. This was pretty successful, too. Knowing I can use the soldering iron with the solder paste, when I replace my chip, I'm going to use the solder paste in a thin line across all the pads.
Farnell sub-total = £19.64 (agh, I forgot to order the headers!)
CPC sub-total = £25.14
ScrewFix sub-total = £3.82
Running total (excluding set of washers and screws from B&Q) = £432.63
So I STILL need 2 headers and another A3982 chip.
* On a complete and utter off-topic aside, I was thinking about how the past participle of the verb "to slide" is slid. That's (about?) the only one. For glide it is glided, pride it is prided, side it is sided, and for ride it is ridden.
Isn't English a wonderful and contrary language?!
I must be heading for success then!
Firstly, I must thank Nophead for sending me those spare parts - I have practice material! More later.
The instructions for the stepper motor driver board say to plug it in to see if it works, but I didn't have a 12V power supply. Maplin, as usual, didn't have what I wanted, so I ordered from Farnell again and ordered spare surface-mount LEDS, too. I was underwhelmed to find the power supply didn't come with a lead! I dug around in the long-since-abandoned-shelving upstairs, and found the power lead for either my old PC or its CRT monitor. It had a 5 A fuse, so that seemed OK. Plugged it into the power supply - and nothing. Checked all the switches were on and the socket was live, all OK. But no power from the PSU. Wondered about that old lead? Then I remembered somebody telling me something about how ATX power supplies differed from AT power supplies, so some internet hunting later, and I found out how to short out pins in the power header for testing purposes.
Usually the power header is a 20 pin affair, but mine looks bigger. It has a separate 4 pin block slid* onto the end of the larger connector.
Here is a picture, with the power header shorted between pins14 and 17:
Thank goodness for the internet and the wealth of information available on various websites.
This done, I switched back on - and it was a good job I was watching the fan closely, because I watched it disappear as it spun at speed, but I couldn't hear a thing. Wow, I know it is called a "silent" PSU, but I didn't expect that they meant it!
Barney reported a blown chip, with the advice to check continuity on the board before plugging in. My meter has dead batteries....
Now I had a working PSU, I switched off, plugged one of the Molex-type leads into the board until it snapped into place and switched on with my finger over the off switch. Here was the crackle and pop. I switched off very quickly with a slight burning smell developing. What was that, you said, Barney? Something about checking continuity everywhere?
Ooops, I have given the chip a contour! but it doesn't look like the board is damaged.
This time following Barney's advice (!) I shall nip the legs with the tin-snips if I can to remove the dead chip. And I'm going to Screwfix now for some replacement meter batteries - open on a Sunday and ~£2 cheaper in total than Halfords.
Then I shall inspect the rest of the board, and CHECK CONTINUITY everywhere. Without the chip in place, I should get better access. Then, if all is well, I shall replace the chip and check the continuity from its legs, too.
Soldering/desoldering learning.
Following on from a link off the forum for Kapton tape supplied by DealExtreme in the States, and looking to see what else they had, I noted that they had desolder braid with lots of reviews, including one that said this particular desolder braid had flux in it, so you don't need to add any. Ah-ha, maybe that's why my desolder braid isn't very successful for me, as perhaps it doesn't contain flux, and I wasn't adding any.
Order from CPC (another part of the Farnell empire) a flux pen. Also a conductive pen as a temporary fix in case I blow any of my board tracks.
I shall try the fluxing/desolder braid to remove the remnants of the old chip legs. If I'm not getting anywhere I shall leave them in place rather than risk damaging the board. Then I'll clean the area with surgical spirit, in case there is contamination after the smokey smell.
Also, I have seen a video of soldering 1206 parts by hand with solder wire, and this requires flux to be put down first and touched with the soldering iron to "tack" the part in place, before laying solder wire partly along the end contact of a component, and just bringing the soldering iron (using a chisel tip, with the flat face downwards) down onto the edge of the pad, sliding it in to touch the part and sliding straight back away. This gives minimum iron-exposure to the 1206 component. I'm going to practise this technique when I have my flux pen. I suspect my 0.7 mm solder wire may be too thick, though; it needs more like 0.4 mm.
I have already practised using the chisel tip and soldering a 1206 component using the solder paste, which itself tacks the part down. This was pretty successful, too. Knowing I can use the soldering iron with the solder paste, when I replace my chip, I'm going to use the solder paste in a thin line across all the pads.
Farnell sub-total = £19.64 (agh, I forgot to order the headers!)
- 350 W PSU, part number 1277264, £15.68
- green LEDs, 1206, part number 122672, £0.061 each (x10)
- red LEDs, 1206, part number 1318261, £0.079 each (x10)
CPC sub-total = £25.14
- no-clean flux pen, part number SA00859, £5.79
- conductive pen, micro-tip, part number SA00462, £19.35
ScrewFix sub-total = £3.82
- LR44 cells, pack of 2, part number 44122, £1.91 (x2)
Running total (excluding set of washers and screws from B&Q) = £432.63
So I STILL need 2 headers and another A3982 chip.
* On a complete and utter off-topic aside, I was thinking about how the past participle of the verb "to slide" is slid. That's (about?) the only one. For glide it is glided, pride it is prided, side it is sided, and for ride it is ridden.
Isn't English a wonderful and contrary language?!
Tuesday, 15 September 2009
Step(per) 1 finished!
I finished the first stepper motor driver board today. My second mega-resistor went better than the first. I am amazed how often I end up needing to use Blu-Tack to hold things in place; it should be named on the parts list under tools!
So here is the front of the board:
and here is the back, with its near-perfect soldering:
I just wish I could test it, but I don't yet have a 12V power supply, nor do I think I can mock one up, as I did for the 5V previously.
I am very grateful to Nophead who has promised me a replacement 100 nF ceramic capacitor and surface-mount green LED, so I should be able to do the reflow work on the other 2 of these boards towards the end of this week.
I'm hoping to read Barney's blogging of his reflow experiences before I do more, to see if I can pick up any tips.
I really appreciate the help and encouragement and practical assistance given freely on the forum and in people's blogs and comments. It's something special.
I have learnt that these boards that have been giving me quite a challenge are now available ready-assembled! Oh no, they're not are they? I could have waited for that had I known it was in the offing.
Oh well, at least at the end, I can say "I made all of that", with some pride (especially if I have a working 3D printer!) because I knew how to do NONE of this 6 months ago! And I couldn't have done it without the RepRap "community" being what it is.
Meanwhile, I can tackle the cabling, or I've still to try out those penny washers on my mechanical build.
Monday, 14 September 2009
So far so good
I was soldering the remaining components on the dreaded stepper motor driver board last night.
I started with the small components, my fine soldering iron tip and solder wire. Having clipped one of the RJ45 sockets into the board, I discovered that it was impossible to get the other one in due to the ridges on the sides of the components. I carefully unclipped the first, and filed off the ridges on the sides of both RJ45 sockets that will be next to each other on the board.
I did some excellent soldering of the small-leaded components, and I was so impressed I took a photo'....
This is only the fifth time I've soldered. Can I just say woo-hoo!!!! Yeah!
I found that if the solder wire wouldn't melt after more than 4 seconds, cleaning the soldering iron and/or re-tinning and cleaning it improved the heat conduction. I also noted that just 4 dabs of the 0.7 mm solder wire per lead was perfect for these fine leads.
For the break-off 4 way header, I "Blu-Tak"ed the front onto the edge of the board to make a firm fit ready for soldering. I was glad I'd re-mounted the edge electrolytic capacitor because I wouldn't have been able to fit the Molex-style connector otherwise.
For this and the other large-lead components, I swapped to a larger soldering iron tip.
I used 8 dabs of solder wire for each of these larger leads but on inspection I found this left the solder only level with the board rather than beaded, so I went round and did the same again to get a bead! You really need thicker solder wire for such components.
Here is the finished through-hole soldering. You can see the Blu-Tak still attached on the left of the picture.
For the first mega-resistor, I tinned the leads, which had been bent round very carefully to align with the pads without touching any neighbouring components, nor to have the resistor body touching anything. I also tinned the pads, and then used the soldering iron at an angle that prevented contact with the chip/neighbouring capacitors etc to melt it all together using a little more solder to get a firm footing.
This was extremely awkward and took me a long time. I stopped after the first one, so I still have the other to go. I am not looking forward to having to do this on the other 2 boards. Don't buy these resistors!
It was too late to blog afterwards, but I'm afraid that's when I took the photo's, so they're not so good - but my soldering was fantastic!
.
I started with the small components, my fine soldering iron tip and solder wire. Having clipped one of the RJ45 sockets into the board, I discovered that it was impossible to get the other one in due to the ridges on the sides of the components. I carefully unclipped the first, and filed off the ridges on the sides of both RJ45 sockets that will be next to each other on the board.
I did some excellent soldering of the small-leaded components, and I was so impressed I took a photo'....
This is only the fifth time I've soldered. Can I just say woo-hoo!!!! Yeah!I found that if the solder wire wouldn't melt after more than 4 seconds, cleaning the soldering iron and/or re-tinning and cleaning it improved the heat conduction. I also noted that just 4 dabs of the 0.7 mm solder wire per lead was perfect for these fine leads.
For the break-off 4 way header, I "Blu-Tak"ed the front onto the edge of the board to make a firm fit ready for soldering. I was glad I'd re-mounted the edge electrolytic capacitor because I wouldn't have been able to fit the Molex-style connector otherwise.
For this and the other large-lead components, I swapped to a larger soldering iron tip.
I used 8 dabs of solder wire for each of these larger leads but on inspection I found this left the solder only level with the board rather than beaded, so I went round and did the same again to get a bead! You really need thicker solder wire for such components.
Here is the finished through-hole soldering. You can see the Blu-Tak still attached on the left of the picture.
For the first mega-resistor, I tinned the leads, which had been bent round very carefully to align with the pads without touching any neighbouring components, nor to have the resistor body touching anything. I also tinned the pads, and then used the soldering iron at an angle that prevented contact with the chip/neighbouring capacitors etc to melt it all together using a little more solder to get a firm footing.
This was extremely awkward and took me a long time. I stopped after the first one, so I still have the other to go. I am not looking forward to having to do this on the other 2 boards. Don't buy these resistors!
It was too late to blog afterwards, but I'm afraid that's when I took the photo's, so they're not so good - but my soldering was fantastic!
.
Saturday, 12 September 2009
And one was where 3 should be...
I have set out the remaining components for the first stepper motor driver board. You can see the re-positioned edge electrolytic capacitor.
The two enormous (in comparison) black components on the right are the stand-in resistors, with their leads bent to fit the board's contact pads!
The component in the middle of these two is the high power header connector - you may remember that's the part I ordered in numbers way over the top, because actually you only needed one, which is broken into 4-way parts. The first go caused a 5-way part (!), so after that I tried needle-nose pliers (too wide), and then cutting the plastic with tin-snips. This worked a treat, but I was glad I was wearing eye protection because each cut part fired off, and bounced off 2 or 3 obstacles before disappearing! I would see where it went first and second and then lose it! Took me a while to find it, based on the landing noise it had made!
See the blue header at the bottom? That's the only one I have. Remember there are 3 stepper motors, and therefore 3 driver boards? Somehow I failed to order 3 of these headers!
So for future orders, so far, the list goes:
So it will be a while before I finish these three stepper motor driver boards.
The two enormous (in comparison) black components on the right are the stand-in resistors, with their leads bent to fit the board's contact pads!
The component in the middle of these two is the high power header connector - you may remember that's the part I ordered in numbers way over the top, because actually you only needed one, which is broken into 4-way parts. The first go caused a 5-way part (!), so after that I tried needle-nose pliers (too wide), and then cutting the plastic with tin-snips. This worked a treat, but I was glad I was wearing eye protection because each cut part fired off, and bounced off 2 or 3 obstacles before disappearing! I would see where it went first and second and then lose it! Took me a while to find it, based on the landing noise it had made!
See the blue header at the bottom? That's the only one I have. Remember there are 3 stepper motors, and therefore 3 driver boards? Somehow I failed to order 3 of these headers!
So for future orders, so far, the list goes:
- 2 more headers, part 110-3949 from Farnell
- 1 more ceramic capacitor, 100 nF, part 165-0887 from Farnell
- and 1 green 1206 2 mm LED, which I might get from Farnell as well, say part number 1226372 at 6p.
So it will be a while before I finish these three stepper motor driver boards.
Fix part 2
I cleaned off what I could of the solder on the removed electrolytic capacitor, and manually soldered it back on to the board, using Blu-Tak to steady the side as I did so.
I de-soldered the dud LED, and checked it afterwards - it really is dud. I checked a new one, which lit, and soldered it into place, again using a tiny bit of Blu-Tak against its side. Now it doesn't light. Hmmm - methinks my logic is off, possibly because of the rest of the circuit. I just can't believe this LED is as dead as the other after simply soldering it on. Unfortunately I don't have a 12V supply with which I can check this part of the board, which is the power circuit.
I de-soldered the dud LED, and checked it afterwards - it really is dud. I checked a new one, which lit, and soldered it into place, again using a tiny bit of Blu-Tak against its side. Now it doesn't light. Hmmm - methinks my logic is off, possibly because of the rest of the circuit. I just can't believe this LED is as dead as the other after simply soldering it on. Unfortunately I don't have a 12V supply with which I can check this part of the board, which is the power circuit.
Subscribe to:
Posts (Atom)