So I had the stencils cut on laser and it didn't turn out so well. It is usable but there are some quite big problems. It was tried with 0.1 mm transparency (for laser printers) and 0.2 mm one for general office use. The main probles is with size – as I made the drawing from soldermask gerber I didn't resize it to small enough size. Also given how my gerber viewer exported the layer to SVG and DXF, most pads were polygons with circles in corners, when shrinking it created even more complex forms. Which sadly turned out to be problematic. Mainly on the power back board – stencil for this turned out to be pretty unusable. There are starting holes outside of rectangle (as seen in photo). But I also tried to lasercut stencil for another board I am planning to build – one with special gerber file for stencil with true rectangles and that turned out fine so I am hopeful that with more time customizing my drawing (mainly making polynoms into rectangles) in AutoCAD I will be able to get good results. Definitely will try again.
[attachment=0] [attachment=1]
I also ordered parts. Of cource some weren't in stock and some items weren't even in catalogue so will still probably need to order from Farnell (well, local distributor when in vicinity to save on postage). Not sure if I will get the parts currently not in stock as this is TME and they are quite arrogant about fulfilling orders, but I got the bulk today.
I made some observations about BOM:
The FZT605 darlington array is SOT223 but the footprint on board looks like it could also accomodate DPAK. (Also in the illustration it looks like DPAK.) Quite interesting (at least for me), any reason for it? I was planning to use NZT605 from Fairchild but maybe something in DPAK would be better.
What are you using as the J5 RJ11v header (in BOM as 8 pin, actually just 6 pin with holding holes)? I found header which could fit in TE Connectivity catalogue. Not sure if I will use it or do it some other way.
By measuring I concluded that I can solder diodes in miniMELF package onto pads for SOD123 (which use diodes specified in BOM). The illustration also shows miniMELF. And it is much sipler to buy diode in miniMelf that SOD123.
Still I will have to study datasheets of most of the diodes or buy them from Farnell. But I don't think that they are so specific as to be problematic to substitude.
In BOM there is pot BOURNS 3362. I think that 3362 is leaded version and the pads on PCB are for SMD. I ordered 3364X and it fits.
For anybody ordering diodes for rectification, I could't find SR580 anywhere alse than Comet. So it it just diode I ordered some other model. Only it is P600 package and there in not enough room for it to put flan on PCB. The TVS is also P600 so I will put 3 diodes on PCB and 2 on them creating pyramide. Will look interesting.
Don't know if I am just looking in wrong slaces but the SMD 0805 27kohm 0.1% resistor seems pretty hard to find. Only Farnell seems to have them. TME stocks only 27.4kohm.
Looking at L1 maybe the reason for the so much covered vias on back board (about which I was complaining in my previous post) is that some chinese worker actually thought about it. The pads on the coil are huge and expand beyond the pads on PCB, so if the vias on PCB were even little uncovered there would be chance for short. Very interenting.
[quote author="sparkybg"] The same is with color display - if it is cheap enough for a company like JBC, why not put it in there. There are MCU-s with internal TFT controller on the market. And 3.5inch TFT is definitely cheap enough. It starts around $10 on ebay, which means it is even cheaper for a company. If I ever make another revision of the controller, I will put a 3.5 inch TFT+ touch screen in it. You have $10 display, $10 MCU controlling the whole thing, and no mechanical switches whatsoever. When you add to this the panel of the box, which will be considerably simpler to manufacture and with less parts, you may end up with a cheaper solution using a color TFT instead of 7 segment or OLED + switches/encoders.[/quote]
You are probably right. The price difference is small. But still and maybe because of this I think that it is marketing move. JBC uses it on higher models. (Not the single handle version one.) Metcal (or Thermaltronics) is using simple 1602 display and it is enought. But then they have the 13.56MHz curie point heating technology as selling article. JBC is in the resistive heating pond which is so far lot more competitive with chinese manufacturers so it needs to separate itself better and the color LCD probably produces nice enge.
I am little leery of touch interface on soldering station. I can see myself trying to change temperature or other settings while holding soldering pen as cigarette with fingerpads dancing on touchscreen. And that is good way how to stab the touchscreen with hot iron and I am not sure cheap screen from China most people like me will be able to get theirs chubby finger on will be able to withstand it.
Yes, I was little worried about drilling accuracy. But that turned out fine. The soldermask is little off but that should not make a problem, because it does not actually cover the pads. It is little strange that vias on front PCB are actually better looking and partially exposed. On the back PCB they are covered with soldermask which creates little bubbles (which is not so pretty). But on the other side of back PCB there are somevias which are similar to vias on front PCB (more on photo.) My personal guess is that it is result of existence of drill drawing and guide in front PCB gerber files (as this is seen in all 10 copies of each design, not just one board.) The silkscreen is good I think but I have seen on internet photos with comparable print and comments ranging from not much to horrible. So...[attachment=0]
[quote author="sparkybg"]You can make them in the kitchen oven, if you have stencils. :) However, I made them purely with soldering iron. No hot air whatsoever. Just get yourself a decent flux. I am using BGA no-clean flux. [/quote]
I was thinking of somethink similar as I have old oven in kitchen I am going to renovate soon. I would still need to mock up some form of controller as to not heat it more that the 260°C limit electronics parts seem to have. I have this paste and if I am reading the profile correctly it approaches this limit point.
I am not sure my hand would be still enough not to scrape IC packages with hot iron. BTW I am using KINGBO RMA-218 from China (Aliexpress) and hard rosin dissolved in alcohol with glycerine for SMT work and so far it worked quite well combined with Stannol HF32 S-Sn60Pb39Cu1 0.32mm wire. But I have not yet made too much. Truth is I have been little leery about investing 30 EUROs into brand flux from reliable supplier.
[quote author="sparkybg"]You can use 0.003 and 0.004 ohms. On the schematics there's a table what the resistor values of the amplifier are for 0.003 and 0.004 shunt.[/quote]
Ah, of course. I was little confused as on the back schematics these is hard-coded value of 0.003.
[quote author="sparkybg"][quote author="Erythros"] The 60V in TVS is reverse standoff voltage? [/quote] Yes.[/quote] Ok, after this I have found that the note Tranzorb actually means name of product line from VISHAY.
[quote author="sparkybg"]If the pad layout and the voltages are the same, you can use another controller, but you will also have to modify the firmware to support it. The commands and the initialization of every OLED controller chip is different. So if you ask me - get a SSD1306 OLED.[/quote] That's given. Looking how it is solved in u8glib pretty much proved that it is just shifted. The SH1106 has matrix 132x64 so they just place the 128x64 in the middle so it is shifted by two points from left and right. And playing with it I can see two points width bar on right side when driving as SSD1306. But it will be surest to gen panel with SSD1306. I just seems so tiny, even the 1.3 inch version. But then the bigger ones (2.4) are expensive: about 20 USD per panel - that seems little excessive for soldering station (like the colour LCD on newer JBCs).
BTW, is there some hidden design factor for use of IPD053N08N3 or should similar N channel MOSFET with similarly low Rds(on) and values of Id, Vds, Vgs and Pd also fit the bill?
[quote author="Circuiteromalaguito"]Can a 7seg and OLED displays be used at same time? Is the firmware ready for that?
The idea is to use a big 7seg display for temperature and maybe other important stuff (change iron? Sleep mode? OFF?) because of greater visibility.[/quote]
I don't think I get the idea behind this. The OLED has very nice viewing angles. And it is just soldering station - there is not much information you would need displayed.
Well, I got ordered PCBs in my post. They are not bad, but not 100%, however will work nicely. [attachment=0] When seen in natura it looks really packet. Made me realize that hot air reworking will be probably needed. Which is kind of ironic, given that I wanted nice soldering iron in first place and now I will have to buy el cheapo hot air to build soldering station. Well that is life I guess. Now I will have to talk to friend with CO2 laser about making some stencils.
I still have some questions about BOM:
Not sure I get the Ja, Jb, Jc, Jd, Je, Jg resistors with 0 value. It looks like they are parallel to the ULN2003 (by the gerbers).
The ULN2003L specifically the L in the end. That is some specific model? Because I could not find it. Only LV which is low U and I model which is in opposition to the note in BOM about it being high U,I model.
Also you have there in BOM on lines 90 and 92 two zener diodes 3V and one 3.0V - I take guess that they are the same.
The WSL3637 is 0.003 ohms?
The 60V in TVS is reverse standoff voltage?
Will have to finalize order and choose some nice eshop, Farnell is on most items terribly expensive.
I have received OLED display modules from China. I ordered the 0.96 and 1.3 inch models. Let me tell you these are tiny. I first received the 1.3 inch model and my basic reaction was that I will use the bigger one when it arrive. It was quite shock when I realized that that was the bigger one. And they use SH1106 instead of SSD1306 which is specified in BOM which is similar IC but it has matrix of 132x64 so you have to offset the image.
Don't know what to do, they seem incredibly small. Probably will go to the 1.3 route, bigger ones uses different IC pinout and I already have the backplane PCB.
[quote author="neslekkim"] What do you mean with running low?, low temperature?, the T12 tips are almost cold on th eend, so that should work ok I guess, if you have weight in the bottom so it doesnt tip over.
The hakko piece is solid aluminium I think.[/quote]
Yes, I meant temperature. Aluminium is nice but you have to anodize it. (Which is incredibly simple, but you need the pigment.) That take me to: Do you know how hot are the JBC C245 tips? Can you take them out of the T245A just by hand?
[quote author="sparkybg"] If you are planning to drive them with this controller, you will have to find out yourself how they must be connected, and then create a profile for them.
JBC microtweezers are an alternative. And maybe a better one - they seems smaller on the photos, and are again 80W (2x40W). They work extremely well form 0201 upto "D" package (never tried something bigger). And it seems they are a bit cheaper.[/quote]
By "reasonable" price I meant about 1/4 of the current price of JBC microtweezers. (I take it that we are talking about PA120-A.)
If I get them then I will try to use this controller. But that's if.
[quote author="tmatthewj"]Hi sparkybg and Hi guys!
My first question is for those who use jbc c245 with the non genuine jbc station. How do you change the tip? On the original jbc desk stand there is a metal tip changer. But if I build a controller and buy a t245 handle with some cartridges then i won't have this little tool for tip change. Or you use a simple plier to change?[/quote]
Do you own original JBC or is this theoretical question? Because I have only seen the changing by stand on video but I always thought that I would rather use the silicone pad Metcal's stations come with. On one side it is nice that it cuts power but it didn't look very usable.
[quote author="tmatthewj"]Second question is also for those who has custom solution. :) How do you store tips? I saw this solution, but it's expensive. http://www.jbctools.com/sc-b-extractor- ... t-959.html I am thinking about building a similar stand for my cartridges. In this way I could change the tips easily. [/quote]
I store tips in plastic bag, but I am using wooden stand for tweezers, pliers, scalpels and syringes with flux which looks little bit like stand for pipes. I wasn't that hard to make but I guess that it depends on your skills.
[quote author="neslekkim"]I bougth an hakko holder: https://instagram.com/p/3FEfk1jOgA/ It's heavy and compact, wanted to get the JBC tool, but it's too expensive, and too roomy..[/quote]
From what it's made of? The T12 tip are running relatively low so it got me thinking about printing something similar.
BTW, do you anybody have some experience with Weller WMRT 12V/80W desoldering tweezers? I may be able to get them for reasonable price.
[quote author="lapas"] 2. Is your soldering station compatible with Weller WXP series irons? Can it read standby sensor and control status LED (see the attached image)?
[quote author="sparkybg"] The important thing here is being equal. When you put the PCB on higher temperature, all the resistors goes up or down together, so the gain stays the same (well, sort of...). So, yes, 25ppm/c is enough.[/quote]
Good to know, thanks. Given the fact that 10 ppm/°C resistors are about 6 times more expensive than 25 ppm/°C resistors. My worry was that there is some hidden absolute reference. But thinking about it, that would be probably stupid.
[quote author="sparkybg"] Slightly modified 5.0 front on 5.2 back. :) But the performance of 5.0 and 5.2 is the same anyway.[/quote]
I was little selfish in that regard. I was interested if there is working 5.2 prototype as I wouldn't want to end up with 20 leaky cup mats.
Nice. That is the 5.2 version or the 5.0 as shown in previous videos? (With changed OLED colour.) I have problem reading the silkscreen.
It is my first project where I will be using 0.1% resistors. So I am thinking about temperature coefficient. But given normal temperature range I imagine that ± 25ppm/°C, which gives |0.1%| at |40°C| will be fine?
Well, I was thinking about buying aluminium square tube (jäckl) which can be had in 120X120X4 or 100X100X4 custom length cut. Mill into it some grooves. Put the the front PCB on front panel, back on back panel and toroid between them. Maybe mill tipping display holder. Rubber legs on bottom.
I have sent some requests for quotes on toroidal front. Sometimes custom is on par with of the shelf. (In price department.)
[quote author="mrjoda"]encoder - you told me yet about enclosure I accept that of course. My idea for you was add some connector for encoder and modify SW for that, for example with one jumper - connect - use buttons , jumper disconnect - encoder is using - buttons are direct temp. memory[/quote]
I was thinking about it either. Encoder could be nice. But the controller is version 5.2. On videos there is model 5.0. So there probably was one iteration more. That is lot of experimentation. (Moreover that being mark 5.) As we do not have CAD files, only gerbers, to remade whole board with encoder connection would provide gigantic number of possible f**k ups you could do. So better take what somebody else was so nice to prove and be happy about that.
[quote author="TilmanBaumann"] was wondering the same. But it turns out that is described on the first page in the feature list.
- resistive instrument identification (upto 625 different instruments can be identified by 2 resistors on the connector)
I had already suspected ID chips and the like. But honestly, resistors are so much more easy. :) [/quote]
I remember that sparkybg had some fancy to use 1-wire connected MCU to do the identification. Which is little overkill. I can see the need for MCU when you want to connect somethinkg like accelometer or gyroscope as I have seen Kurtz Ersa doing for sleep function.
[quote author="Circuiteromalaguito"]Is it Open Hardware? What about putting it at GitHub?[/quote]
Not everything has to be on Github. But stating some license would clarify the issue. Something like "do make it, do not sell it" or "do as you please, I do not care".
[quote author="sparkybg"]As I mentioned earlier, I made the PCB-s of the project for specific enclosures I already had. That's why it uses buttons.[/quote]
That case is not so bad. It basically is 100x100xsomething box. The biggest problem is to find some toroidal transformer which which would have d=100mm. So far I have found that these smaller diameter and higher toroid cores are mostly for bigger output (about 250VA).
Thanks. That settle it that I will have to do my own.
To another note: I was doing some sourcing today and for 3 boards (front, back, oled) I was getting quotation about 80 USD delivered for 10 boards. (Actually from 80 USD for economical run to 200 USD for "best material" - same fabhouse, I do not know the difference, because specified parameters were the same.) So if you will be doing your own group buy with bigger number of pieces, you could get better prices. (I will have to decide tomorrow if I am going to place that order as I do not really need ten pieces.)
Well OK. I understand that it is probably more work than it is worth and nowadays everybody can have PCBs made cheap in China. I have just send some quotation questions to fab houses. (Just saying that people [better to say me] are interested anyway.)
BTW, the stand sensor PCB is designed for which specific holder?
I have gotten my new chinese copy of Hakko FX-9501 in post today. And it is quite nice device. The grasp is quite better than my current Solomon device. (As I have need to hold it like pen because of trembling hands I am constantly burning my fingertips.) I will have to order some adequate tips for it (chisel, well), but so far so good. Thanks for the advice.
By the way, what kind of transformer are you using? By the fuse I would say something above 120w. Unfortunately I do not have any 24V on my hands, only the 29V 150VA toroid and I am thinking if I could use that (I have never adjusted toroid one, it is quite new in warranty period yet) by not fully opening power transistors. (If I understand it correctly the 24V most irons use is because of safe voltage in dangerous environment area.)
I did not mean to critise your work. Maybe that is how it came out but that was not the intent. I just saw your video and that got me thinking. That is all.
I have been looking at Aliexpress to buy handle for this project. I have been using something similar to SL-30 but that handle is little too big, I would like more penlike handle. These chinese Hakkos seem like good fit with nice price/feature ratio.
