Supertronic answered most of your questions - I'll add in just a little more detail.
[quote author="teletypeguy"] can a 24mm reel be used (with offsets to accommodate it)? [/quote] Nope - the feed mechanism suits only the sizes that it has been designed for. I have the 240A machine - it has 21- 8mm, 4- 12mm, 2- 16mm feeders in addition to the large chip tray at the front. I think that the only way that you can use those large chips is to remove them from the carrier tape and place the chips into one corner of each pocket of the front tray.
[quote author="teletypeguy"] how are the wells at the front used for loading chips? Can you program different parts in different wells (eg: can I have a TQFP-64 in wells 1-5, and an SSOP-28 in wells 6-x?)[/quote] Nope to using different chips with different offsets in the front tray. There is only a single offset entry for the entire tray. That said - NeoDen might consider allowing different offsets for each individual pocket in the front tray *IF* enough people ask for that feature.
[quote author="teletypeguy"] Do you keep adding chips to the wells as the machine consumes them? Do you just push them into one corner and hope for the best alignment? [/quote] You choose one corner of each pocket (I use push the chip towards the rear-left of the machine) and calibrate with your chosen position. Accuracy is very good. I do not yet know if the machine will go back to the first pocket after it has grabbed chips from all of the pockets - I haven't needed to find out yet.
[quote author="teletypeguy"] has anyone run into issues with part height? Yeah, most res/cap/ic stuff is a similar height, but what about tall electrolytic caps (of varying heights)? Anyone try smt headers?[/quote] As Supertronic mentioned, max part height is 3mm. In addition, components that do not have a smooth, flat top surface won't work because the suction won't hold. Most electrolytic caps have ridges in the top; even if the machine could handle the height, the suction tips wouldn't hold onto the cap. Same goes for SMT headers - the machine simply isn't set up to handle such parts.
For me - headers simply aren't a problem because I use only through-hole headers. Most of the work I do is for Industrial applications and the people are too rough for surface-mounted connectors to survive.
My workflow is to mount all of the SMD parts onto the board and reflow. Then populate all of the through-hole parts (large connectors, headers, power relays, large capacitors, power resistors, etc) and run through the wave-solder machine. Then wash, dry, test.
[quote author="supertronic"] It's not a feeder offset problem the first board is ok the offsets for the next boards are right because head 2 puts the component at the right place. The first head using XS nozzle has the problem. If the component is traveling over long distances the accuracy is poor.[/quote]
Yeah - the loss in accuracy over large distance is why I suggested looking at how square the gantry is with respect to the front stops. If there is a gap on one side or the other, the gantry is not square and you will get an error that gets larger when the distances become large.
Other than that, I would suggest contacting NeoDen with an extremely detailed message about what you are seeing and what you have tried to do so far. I have been dealing with someone named "Mia" who has been very helpful to me.
[quote author="supertronic"]I have some problems in placement accuracy using XS nozzle if the distance between pick and place is longer. On multi board placement configuration.[/quote]
I'd be checking several things. First: power-down the machine and check to see that the gantry hits both front stops at the same time. In other words, no gap on either the left or right sides of the gantry and the front supports.
Check to see if the nozzle is exactly centered in the component tape. To do this, go to the main screen that shows the files on the SD card. But instead of hitting 'Load", hit "Edit" instead. Cursor down to the line that shows the nozzle offset for the stack that you are checking out. Note that the X and Y parameters will show the settings for that stack.
Then tap the top-right parameter box (marked 'Stack' and highlight the right-hand digit only. Use the "UP" button on the keypad to select the proper stack number. Then press the "Joystick" icon in the middle of the touch-screen. The head will move to that stack and drop the nozzle. Ensure that it is exactly centered in the component hole in the tape.
To change the offset, decide whether you are adjusting either the X or Y setting. Highlight the extreme right-hand digit in whichever parameter you are going to change. Then use the Up and Down buttons on the keypad to change the entry. After making a change to one of the parameters, hit the Joystick icon in the middle of the touch-screen again. That will move the head to the new location. Note that the head will move only when the Joystick icon is tapped.
Keep adjusting the offset until it is correct. Then press the "Enter" button to save your changes to the SD card.
Other things to look out for:
You mention "multi-board" placement. Is the accuracy good on the first board? If so, I would carefully measure the exact distance between the individual boards in the panel and enter those values into the multi-board header in the file (header ID #3).
More information from you might allow us to make more suggestions.
I wanted to mention something pretty cool that I noticed while I was running the SMTDemo file: there are a couple of places within the file where the pickup head is grabbing two components from the same stack. That is: both nozzles are in use and grabbing from the same tape.
The tape feed moves the tape TWO positions in this particular case! Because the machine knows that both nozzles are grabbing from the same tape, it moves the tape by 2 positions, then loads both nozzles sequentially. This saves time - the normal method would be to move the tape by one position, grab the component, move the same tape another position, then grab the second component.
I had previously mentioned that my TM240A had a defective vacuum pump when I received the unit. Although I got the pump working, I suspected that it would not be reliable and asked NeoDen for a replacement. I also mentioned that I had not received the demo PCB and Mia said that they would send one along with the pump.
The package with the replacement pump and demo PCB arrived Thursday. I'm not sure what ship method they used but the package arrived at least 6 business days earlier than I expected.
I didn't even bother installing the new pump (the original is still working), but instead loaded up the first 3 reel positions with tapes containing 0402, 0603, 0805 components. As it turns out, I should have also loaded up a reel of 1206 components - but I'll get to that later.
Anyway, inserted the SD card and loaded the SMTDemo.csv file from the card. Hit the 'Step' button - and immediately noticed major problems. The tape would not feed because the pin was centered at the edge of the tape instead of over the tow of feed holes and the nozzle was attempting to pick the part up from the area between the actual component and the feed holes. Oops, sez me - I must have somehow messed up the calibration. So: I fired off an email message to Mia at NeoDen with an extremely detailed description of what I was seeing. Because I knew that I wouldn't receive a response from NeoDen until the next day, I proceeded to mess with the various manual setup parameters as well as the calibration settings for the machine.
I was able to get the nozzle to be centered over the component but nothing that I tried was able to move the needle to be centered over the feed holes. So I gave up, restored the machine to its original settings, and waited for NeoDen to get back to me.
I received a message from Mia's private email address - he said that their engineer thought that I might have messed up something mechanical while I had the machine apart. The engineer recorded a video (in Chinese) that showed very clearly what I needed to look out for and showed how to fix the problem if in fact the problem was what they thought.
Turns out they nailed the problem exactly. If you recall, I mentioned that if you power the machine up while the main baseplate is standing vertical and sitting on your bench surface, it was necessary to lift the bottom edge of the baseplate off the work bench with a piece of 2 x 4 lumber or something similar. I also mentioned that the stepper motor had enough torque to cause the belt to stretch and skip teeth if you didn't do this.
I know this, of course, because that is exactly what happened to me the first time that I powered the machine up while it was standing vertically on my bench. But when the belt skipped teeth, the pickup head was at the homing position of bottom-left on the baseplate. As a result, the entire gantry became 'racked' - the right-hand belt skipped more teeth than the left hand belt. Of course, I didn't realize this had happened.
The fix was simple: removed power from the machine, then manually moved the gantry to the front of the machine and observed that there was a gap of about 2mm on the left side between the gantry and the front stop while the right side of the gantry was right tight against its stop. Backed the gantry up a little bit, slid a screwdriver shaft between the right-hand stop and the gantry, then forced the left belt to skip first one tooth, then one more. Now the left side of the gantry was right tight up against the front stop while the right side had a tiny gap. The video that Mia sent was quite clear that there must be NO gap on the left side, so that is how I left it.
Powered the machine back up, loaded the file, then used the 'Step' button to grab the first component. Now the tape feed worked correctly and the nozzle is EXACTLY centered on the 0402 component pocket on the tape. Note that I wanted to use the 0402 tape for checking the centering - the pocket is very small and its easy to verify how well centered the nozzle is.
Ran the SMTDemo file - the machine proceeded to pick up parts from all 3 tapes and place them on the board. I can't yet tell how accurate the placement is - the parts bounce all over the place because there is nothing to keep in place (no paste or anything). I'll try again on Monday after spraying the PCB with something sticky to hold the parts.
I mentioned that I should have also loaded a reel of 1206 components. The SMTDemo.csv file has a '1' in the 'Skip' field for all of the larger components (diodes, QFN32 parts, etc) but does allow placing of all of the small components. One thing to note: the CSV file mentions that the larger parts are 1210 instead of 1206 but the offset values used in the stack setup are in fact for 1206 components. So: I think that the '1210' in the components list is a mistake and that you should load a reel of 1206 components in stack #4, not 1210.
Anyway, my machine appears to be usable at this point. I will know better on Monday - I have up to 200 panels each of two different panels of SMD boards (6 boards in each panel) ready to run - including having paste stencils. I just need to wait until the lady who knows where all the parts are to dig them out for me - then I'll start playing <grin>.
Finally, I wanted to mention that I saved a copy of the video that Mia sent to me in the Cubby cloud storage site that I mentioned earlier. The file name is FixRackingProblem.avi (32MB) and you can get it from http://https://www.cubby.com/pl/Cubby/_84be51e9ba934374b6f13af5e66cb880. I am also uploading more pictures into the 'Canon' folder on that same site - these pictures are the under-side of the main control PCB and more pictures of the under-side of the Stepper Driver PCB (showing the LM2576 Simple Switcher +5V & +10V rails).
[quote author="TAUTIC"]Is anyone actively purchasing these? I'm looking at a few options but would like to go with a trusted source. Would appreciate any guidance.[/quote]
I believe that Ian is now set up as a dealer but he is currently in China. Regardless, I'm pretty sure that he can take your order, arrange for you to send payment directly to NeoDen via wire transfer, then have the unit drop-shipped directly to you.
I'm one of the 6 or so people who jumped in on the group-buy that Ian arranged earlier and I am *very* happy with my purchase so far.
For me, Ian *is* a trusted source and I have absolutely NO problem sending money where he tells me to send it because I am convinced that he is NOT a scam artist. In other words, I feel that he has a massively-positive reputation, based on his past history.
Here are my rough notes from taking apart my TM240A. These will change later when I get around to editing them.
TM240A internal details
First impressions: build quality is excellent. This machine appears to have been designed by someone who cares more about quality and long life rather than lowest possible cost.
There is actually not all that much stuff under the base-plate of this unit. With the main base-plate standing vertical, one can see the blow pump (black) and the two vacuum pumps (blue) near the top of the unit. Below that is the power supply (standard, off-the-shelf Meanwell 24V 8.8A 200W) and the two stepper driver boards. The brains in this unit is the control panel - this contains the entire user-interface, full-size SD card socket, processor. The touch-screen display is connected to this board by a soldered flat film flex.
Power: the Meanwell power supply has a full-range input of 100 - 240 Vac 50/60 Hz. The output of the 24V supply feeds the two stepper driver cards. The only other AC-powered component is the blow pump - this has a 220V sticker on the side even though my TM240A was delivered set up for operation at 120 Vac. Regardless, the blow pump does blow, so I'm not concerned.
There are two stepper driver cards. However, only one is fully stuffed, the other has only one of six stepper-driver stages populated. The stepper motors are bipolar and are connected to the driver board with plugable 4-pin screw connectors. A single 20-conductor ribbon cable connects both stepper driver cards to the main control panel. Because both driver cards are identical, that means that the single stepper channel on the second card gets the exact same step and direction commands as the matching channel on the first (fully-populated) card. NeoDen does this because they have two identical steppers driving a single shaft for the SMD tape-cover film take-up reels. More about that later.
Looking at the back-side of the fully populated stepper-driver card, one can see six identical Alegro A???? PLCC dual full-bridge driver chips. These chips are heat-sunk to that large, thick aluminum base plate with a layer of thermally-conductive tape. There is also a 5V regulator for logic stuff and another 10V regulator for one vacuum pump. In other words, each of the two vacuum pumps connects to one of the stepper driver cards.
Each of the stepper driver cards has a 20-pin ribbon connector. Two pairs of pins are ground, one pair of pins are +5V, one pair of pins are +24V. The remaining six pairs of pins are step and direction command inputs for the driver chips. 24V & 5V rails go to the main control board - the 24V rail feeds the placement head.
Placement head
The TM240A placement head is a marvel of engineering. There is some real magic happening here.
Each of the two nozzle assemblies has a tiny stepper used to rotate the held component, with a flexible hose that goes back to the valve block. Each of the heads moves independantly and are driven by a single stepper situated between them. As seen in the pictures, this stepper shaft is connected to a bar with rollers on each end. Quiescent state is with the bar sitting horizontal and both nozzles at the top of their travel. If the bar rotates one way, one nozzle goes down but the other remains at the top. If the bar rotates the other way, the other nozzle goes down.
The valve block is at the rear of the placement head. This block contains 4 indepndent valves and 5 ports. The block also contains the two vacuum sensors that let the machine know that picking up a part was successful.
Two valves are for each nozzle: suck and blow. The two blow valves are combined into a single input port since only one nozzle can place its component at any one time. The vacuum ports are independent because both nozzles can hold components at any one time.
I've got some pictures of the machine and the placement head. I will take some better (more detailed) pictures next week but these should give you an idea of what's inside.
I've been up to my A** in alligators, so I haven't had a chance to down-sample the pictures yet. But you can go to one of my cloud-storage sites and view them at your leisure. These are hosted on cubby.com - I have many cloud storage solutions to choose from but I have the most space on cubby.
My next message has some notes on what I saw inside the machine - these notes are rough and I will be editing them next week. But it gives you a bit in insight as to what the pictures show.
Multicore used to make a system that was used to repair plated-through-holes that had been damaged or ripped out by someone's carelessness. It came with 3 different sizes - one of those was 0.8mm. The bails (as they called them) were a solid rod of 63/37 solder that then had copper electroplated on top. The rod was then scored at about 2.0 mm intervals to allow each segment to be snapped off inside the hole.
In use, you would drill out the hole with the supplied drill bit, set the PCB on top of a flat-surfaced metal rod, insert the bail and bend it sideways to snap a segment off inside the hole. Then you would use a special impact tool to flatten the bail segment.
This causes the solder core to expand sideways and force the copper covering tightly against the inside of the hole. Then - all you did was solder both the top and bottom of the trace, then use a vacuum desoldering tool to empty the hole. Voila! A nice, reliable plated-through-hole for you to put your component lead back into and solder.
I used to use the system back in the days when we were making our own double-sided circuit boards in-house. Expensive and time-consuming - but it resulted in extremely reliable connections.
I have a link somewhere for a modern-day equivalent - but I'd have to go looking for it.
I still have the original Multicore system under my work bench - it gets used once or twice every few years. I am getting low on bails, though.
Well, my TM240A arrived on March 9th - but I had left for vacation the previous day. So the machine sat until I returned back to work today.
Again - I can only reiterate how well packed the machine was. Much better than I am used to seeing from Chinese suppliers.
But, to my sorrow, the machine did not work upon powering it up. Upon powering-up, there was a delay, then I could feel some stepper steps from somewhere, then a continuous beep until power was turned back off. A careful examination did not reveal any obvious problems, so I gritted my teeth and began to disassemble the machine.
This was harder than I had expected. The machine is a large plate of aluminum with everything bolted to it. That plate is then bolted into the chassis box - this box does not have an access panel on the bottom. So the only way in is to unbolt the main aluminum plate from that chassis box.
You start by removing the plastic IC holder from the front of the machine - this is where one would place large chips like TQFP and such. Then remove the many hex-head screws from around the perimeter of the base plate. You also have to unbolt the holder for the "Y" flexible cable raceway to get access to the screw in that corner of the machine. That's when you discover that the cables inside the box are quite short and you can't lift the plate up very far.
So: disconnect the AC power from the IEC power inlet. These are fast-on connectors and easy to remove. Be sure to note where the neutral connections are - if you get that wrong, you will destroy the switch upon applying power.
Now you can mostly lift the plate up from the rear - but not all the way vertical. Not until you reach in and unplug the 3-conductor plug that comes from the "Y" axis home sensor onto the main circuit board (where the switches and display is).
Even then, the ribbon cable that goes to the stepper driver boards is short. Best to disconnect that as well.
A careful examination of what was visible didn't show any obvious problems. So: now to remove that front panel circuit board.
It turns out that there are 8 screws holding that front panel onto the chassis box. Of course, the screws are underneath the lexan label that covers the entire control panel. Lifted that up and removed the 8 screws. Then I set the control panel inside the chassis box (so that I could get the last remaining cables free), and finally lifted the whole machine up and away from the chassis box. Disconnected the serial port cable from the PCB. Got the chassis box out of the way, set up a bar clamp to hold the machine vertical, and took a close look at the control panel PCB.
Ah-ha! There is a round cable with many conductors that also connects to that control panel PCB - and the connector didn't look right. It was nicely sealed in place with red goop - but it didn't look right. So: I peeled the red goop off of the connector and pulled the cable connector off of the board. Everything looked OK - so I carefully re-inserted the cable end back into the PCB connector. There is a lot of pins there and it took a bit of force - but it wasn't excessive. But now it looked proper.
So: I reconnected the other 2 cables that I had previously disconnected and applied power to the machine.
Success! The machine went through its homing sequence exactly like shown in one of Ian's videos, then parked the head at the top-right and rear spot.
Those short cables really annoyed me, so I set out to do something about it. That round multi-conductor cable that hadn't been seated properly goes to the placement head. There was a lot of slack in that cable that had been consumed by routing it all over the place on the bottom of that big aluminum plate. I cut its cable ties and removed it from the nylon cable clamps, then did a straight-line run to the control panel PCB. I secured it in place in the two nylon cable clamps that were close-by. That gave an extra foot of length.
Made a new ribbon cable with an extra foot of length between the control PCB and the lower stepper-motor driver PCB. That left only the homing sensor for the "Y" axis.
Turns out that this is a standard JST 3-pin connector with 0.1" (2.54mm) lead pitch. I don't have any of those, but I do have lots of 0.1" header pins. I marked the PCB with the wire colors, then pulled the plastic housing off of the pins sticking up from the PCB. Stuck my 3-pin header onto that plastic housing and checked - it fits the sensor cable perfectly. Soldered 3 colored wires (red, blue, black) onto the free end of the header and insulated the connections with heat-shrink. Removed the original connector pins from the PCB and soldered the wires into where the connector used to be. This gave me an extra foot or so of slack on that cable. And - its easily dis-connectible when needed.
So: now I was ready to properly test the machine's manual functions. Set the machine down on the bench with the control panel in front. Powered the machine up - everything looks good. Got into the 'Manual" screen and tried the various functions. Oops - no vacuum on nozzle #1. Great vacuum on nozzle #2 (and the #2 LED on the head goes OUT when the vacuum is good) and the blow function for both nozzles was also good. But no vacuum on nozzle #1.
Stood the machine vertical again and powered it up. Note - you have to have something under the lower edge of the machine so that the placement head can reach the "Y" home sensor. The stepper has enough torque to stretch the belt and start skipping teeth if the placement head is blocked from moving. A chunk of 2 x 4 worked perfectly.
Anyway, both vacuum motors were warm to the touch - but the #1 motor was somewhat hotter than #2. Also - I could feel vibration from motor #2 but not motor #1. Grabbed a Fluke meter - 9.8 Vdc present at motor #2, 9.4 Vdc present at motor #1. There are 3 in-line tubing connectors a few inches away from both vacuum motors and the blow pump - these make the transition from the larger tubing that goes to the 3 pumps to the small tubing that goes to the placement head. Disconnected the hose for vacuum #1 - no suction from the pump.
All of this made me think that the pump motor was jammed. Dismounted vacuum pump #1 - there are two heads (connected in series) and an access cover over the crankshaft. Pulled that cover and rotated the crankshaft. It moved - but there were 2 distinct spots that felt very 'gritty' - something was rubbing on something else. Rotated it a bunch of times over those 'gritty' spots and it got less and less. It finally got to the point where I thought it might work.
Put power back on the machine - Yep - pump is now working. But I don't expect it to be reliable, so I've sent a message to NeoDen asking for a replacement vacuum pump.
Laid the machine down flat again - all manual functions appear to be good.
The machine is still apart. I took some pictures today and plan to take more tomorrow. I'm happy to post those but: what resolution should I down-scale them to? The pictures right out of the camera are WAY too huge and IrfanView makes rescaling easy. But I'm not sure how far down I should take them. Opinions greatly appreciated.
I'll start another message talking about what I see inside the machine.
[quote author="digistump"]Did anyone else in the group by get theirs? [/quote]
Not yet, darn it all. I leave for vacation tomorrow evening and was hoping that I would get to play with it before I left. Now I have to wait until I return before I get to play.
[quote author="digistump"] Missing example PCB that was suppose to be included. [/quote]
I asked Mia from NeoDen if it was possible for them to include a PCB with placement file with my machine. She said "Yes". However, that doesn't mean that they were going to include a PCB and file with everybody's machine. I suspect that they simply didn't think of that.
However, I will make that suggestion the next time I talk to her.
[quote author="ammsolutions"] The Factory are very helpful. They have even agreed to sell me the additional spare parts kit that was not offered to me by the dealer I got mine from. [/quote]
Just curious - what additional spare parts are you looking to purchase? I'm looking ahead for what I might need for my machine.
[quote author="bearmos"] it sounds like, with the proper configuration it'll do modest boards just fine. [/quote]
I guess that it depends upon your definition of 'modest boards'. I strongly suspect that it will handle my boards just fine - but my boards are relatively low resolution: 0603 size and larger components, chips with lead pitch 0.65mm and up, currently at 13 mil traces (and wider) and 11 mils minimum clearance.
Accuracy may indeed become a problem if / when I start dealing with smaller components and smaller / tighter traces. But then I will have the experience that I need to choose a better machine.
Based on all that I have seen and read, I honestly think that is is a great machine for a small manufacturing operation like ours. Do note, though, that I still plan to sub out our volume boards (500 panels per month and up) - I just think that the big people can do it better and cheaper than we can.
But for our low-volume stuff, I think that this machine (TM240A) is going to be perfect.
I'm anxiously waiting for my machine to arrive - next few days, i hope.
We are using an old Electrovert machine - but its a relatively big brute - about 46 inches long. It needs 208 Vac 30A 3-phase power and a good exhaust system. The pot holds 300 pounds of solder and the largest PCB it will handle is 12" (its called a 12 inch machine).
We purchased it used (we were the 2nd owner) 20 or so years ago and its still going strong. I have to disassemble the solder pump every 6 or 7 years to clean out the accumulated dross but that's a fairly straight-ahead job.
We run anywhere from 400 to 1000 boards through per month.
We paid something like $7,000 for it (20 plus years ago) but its paid for itself many, many times over.
