latigid on

Frequent Writer
  • Content count

    1,911
  • Joined

  • Last visited

Everything posted by latigid on

  1. Troubleshooting midiphy SEQ v4+

    There's an address for FOOTSWITCH in the hardware config file, but nothing for gate. Also, I'm pretty sure there is no routing for this in software, but it would be useful for a future update.
  2. Troubleshooting midiphy SEQ v4+

    Please post your troubleshooting issues here and we will do our best to get everything up and running! The more info you provide, the better the answer can be formulated. What did you try so far? What wasn't working? Pictures, pictures, pictures!  
  3. Hi Henry, welcome! We developed the SEQ v4+ with those concerns in mind. It is different to the Wilba project. https://www.midiphy.com/en/shop-details/137/47/midibox-seq-v4-lh-full-essential-kit- On that page you have the full kit available, including the case and hard-to-source parts. There are left- and right-handed cases defined by the position of the JA subassembly. The combined BOM can be copied directly to Mouser's import tool, so it's super easy to get parts. Yes, sometimes Mouser goes out of stock, but many of the parts can be substituted by others (e.g. standard resistors, caps, ICs). Peter also has produced full video tutorials:       Best, Andy
  4. BLM 16x16+X build guide

    Nice!    
  5. BLM 16x16+X build guide

    Here I will document the parts list and build guide for all BLM users. Once there's a bit more knowledge a wiki page could also be written.   BOM: http://www.mouser.com/ProjectManager/ProjectDetail.aspx?AccessID=e31c510729  Quantity Recommended Description Resistors 660-MF1/4DC1002F 4 4 10k through hole 652-CR1206FX-1002ELF 41 100 10k 1206 652-PTL45-15R0-103B2 4 4 10k slider, 15mm shaft, red LED 652-PTL45-10G0-102B2 - - 1k slider, 10mm shaft Resistor networks 652-4816P-T2LF-10K 5 6 10k bussed (165 pullups) 652-4816P-1LF-1K 5 6 1k isolated (current sinks) 652-4816P-1LF-220 5 6 220R isolated (green LEDs, also red LEDs) 652-4816P-1LF-56 5 6 56R isolated (blue LEDs) Capacitors 647-TVX1C471MAD 1 1 470uF electrolytic axial 80-C1206C104K5R 23 50 100n ceramic 1206 647-F931C106MAA 5 10 10uF tantalum 1206 Diodes 512-BAT54 36 100 Schottky diode SO-23 512-1N4148TR 289 300 1N4148 diode 78-LS4148-GS18 5 10 4148 quadroMELF LEDs 604-APTR3216ZGC 289 300 LED green 1206 reverse mount 604-APTR3216QBC/D 289 300 LED blue 1206 reverse mount 604-APTR3216SURCK - - LED red 1206 reverse mount Transistors 512-BC81840MTF 36 100 NPN BJT SO-23 ICs 595-SN74LVC1G17DBVR 3 10 Schmitt trigger, overvoltage tolerant SO-23-5 595-SN74HC165DR 5 10 74HC165D SOIC 595-SN74HC595DR 15 20 74HC595D SOIC Connectors 523-T3507-000 1 1 DIN 8 panel connector 523-T3504-001 2 2 DIN 8 plug --- 2 - 2x5 DIL header male, snap from a larger piece --- 1 - 2x5 IDC plug female (miniCore) Hardware --- 40cm - 10-way ribbon cable --- 20 - M3 nut (stainless) --- 20 - M3 washer (stainless) --- 1(+4) - M2 grub screw, 14 mm (stainless) --- 1(+4) - M2 nut (stainless) --- 1(+4) - M2 washer (stainless) --- 2 - M3 standoff, 12mm male-female --- 8 - M3 standoff, 25mm (for spacing PCB) Buttons --- 18(19) - adafruit 4x4 silicone Spacer --- 1 - spacer cut from 3mm opaque acrylic Parts notes: You can choose whether you go for Blue+Green LEDs (recommended) or Blue+Red. You could interchange some Green for Red. Green+Red is not recommended as the colour mixing is poor. Slider height: I specified 15mm shaft sliders, you could change these for 10mm. 15mm will protrude 13mm out of the case, which might be a bit long. 10mm will only protrude 8mm, so they would be less likely to break off and will probably be a bit more sturdy. TK. suggests 1k should work, the others are too high in resistance to get good jitter-free values. You should use the indicated panel mount DIN, but a cheaper pair of DIN8 plugs might work. M2 hardware is a bit difficult to find in normal stores, but Boltbase on eBay has good deals.   Tools needed: Soldering iron, I managed with a normal tip but it’s a bit tricky with some components Solder, probably leaded. 0.5mm will make the finer pitch stuff easier but again I managed with 1mm Desoldering wick and/or solder pump. Useful when the pads get too much solder Tweezers to pick and place components Cutters for diodes and a few through hole components Good eyes or light and a loupe 2-3 small clamps Flat work surface Multimeter, test continuity between pads     Build notes: The BLM is not particularly difficult to build but there is quite a bit of repetition. I advise to complete one section at a time before moving on.   Soldering order: First up, there are 289 through-hole diodes on this board. It might be smart to get all diodes pre-bent and trimmed before doing anything else. The legs should be cut as short as possible before soldering otherwise the button pads won’t sit flush. If you cut the diode leads they tend to fall out; this can be a bit frustrating. Some diodes also overlap with the transistors, make sure they don’t electrically short. So you may wish to solder all of the transistors first, followed by the diodes. Read on for suggestions on mounting these. The down side with this way would be that it’s easiest to put the LEDs in when there’s nothing else on the board… please have a go and share your experiences. I suggest that the PCB is initially clamped (using a scrap piece of material to protect it from the clamp) to the work space, otherwise your parts may jump around if the board gets bumped. If it is truly flush with your bench it will also help to keep the diodes/solder from going through the plated holes. Try to protect the exposed button pads on the PCB top by using a piece of clean paper or other material; don’t damage the surface by scratching it. LEDs are a little bit sensitive; don’t heat them for more than about 2 seconds at a time and apparently they don’t like static discharges... As far as I know, the humidity warnings you get with them are only really applicable to oven reflow soldering.   So: let’s go! First up you should tin the pads on the rear side only. I recommend clamping the board at a right angle so you can easily go down a “row.” Tin at least the LED anodes (just the top pad!), but you could do one pad of every SMD component (over a few days it will get harder to “wet” the solder as it oxidises). It will be easier to “tack solder” if you choose a pad which isn’t part of the ground plane. Avoid pre-tinning pads with thermally isolated connections to ground. (here the top side is actually at the bottom)   289*LED 1 Actually left on the PCB rear. Ensure the “C” symbol is oriented with the PCB silkscreen. Or you can think of it as making a “T” when the board is viewed the correct way up. (N.B using the default firmware LED1 is the main step indicator and LED2 is used as the cursor for most modes.) Take a rest.   289*LED 2 Actually right on the PCB rear. Take another rest. Once these are placed you can go back and finish all of the cathodes, also touching up the anodes if needed.   3*SOT23-5 Schmitt buffers (IC21-23) Located near header J1. These are pretty tricky, so go slowly and use as little solder as possible. Clean up with braid if needed.   36*BC818 transistors (remaining SOT-23 footprints) Orientation is fixed, again don’t overheat them! Note the line of transistors down the centre of the board (when viewed from the back), another at the left plus 4 more along the bottom. 36*BAT54 Schottky diodes This BLM has a small issue with ghosting (some LEDs are dimly lit when they aren't supposed to be). A fix involves soldering a BAT54 diode 1:1 on top of the transistor, this is hard to photograph but if the solder blobs are big enough they will span from the SMD pads all the way up to the BAT54. Not so big that they short out the two adjacent transistor pins! You may want to see if you are happy with the ghosting before applying this fix; it's only noticeable in complete darkness. 36*1206 10k resistors  Some silkscreen values are missing, sorry about that! They are located next to each BC818 transistor (+BAT54). Don’t worry if solder bridges form from the resistor to the transistors, the pads are common. But don't short out both ends of the resistor! +5*1206 10k resistors (R8-R12) These form the 4 power on delay circuits located at the left, plus one on the bottom. 5*MELF diodes (D1-D5) They are polarised with the black band oriented towards the silkscreen corner line. 5*10uF tantalum caps (C6, C11, C16, C25, C28) Orientation is very important with the grey band towards the silkscreen corner line and + symbol. TANTS MAY CATCH FIRE/EXPLODE IF MOUNTED INCORRECTLY! 23*1206 100nF ceramic caps (remaining 1206 pads with rounded corners) Located next to each IC. Note C26/C27 and R6/R7 don’t need to be installed unless you want to terminate the SC/RC lines, it works fine for me without these.   289*1N4148 diodes These MUST be soldered on the SAME SIDE AS THE COMPONENT with the leads clipped as short as possible BEFORE soldering. Solder just enough to wet the “bottom” pad. If too much solder is applied then the button surface won’t be flat and the pads will be uneven. A flat work surface will help to keep the solder from reaching the PCB “front.” The other way is to place all the diodes without trimming (making use of spacers attached to the PCB corners), then chop the leads. If you do this they have a tendency to fall out when you flip the board, but it works. Similarly they MUST be soldered from the rear i.e. the same side as the component is placed. Be careful with diodes that are placed over a transistor or one of the few that have their legs bent out of the way to avoid a mounting hole, they need slightly different lead lengths. Take another rest.   5*74HC165 (ICs 1, 5, 9, 13 and 17) 15*74HC595 (remaining ICs)   Now attach standoffs to the corners and flip the board to mount the resistor networks: 5*10k bussed (RN1, 5, 9, 13, 17) Note that these are POLARISED according to the silk screen. 5*1k isolated(RN2, 6, 10, 14, 18) You can follow the silkscreen but it doesn't matter here. 5*56R isolated (RN3, 7, 11, 15, 19) if using blue LEDs for colour 1. 5*220R isolated (RN4, 8, 12, 16, 20) if using green LEDs for colour 2, also for red LEDs. RN19 and 20 are on the rear of the board   Through-hole components: 470uF electrolytic (C1) The “minus” end is marked with an arrow, the “plus” end has an indent aligned with the silkscreen.   Sliders (on the front of the board!) Sliders are 60mm long and spaced 60mm apart, so they fit together exactly. Mount all 4, then start by soldering one leg. Check the slider is absolutely flush against the PCB before continuing with the remaining legs. Current limiting resistors for the sliders' LEDs (can wait to check the brightness) You can swap the LEDs with common “2x3mm” sizes. Note the different pinouts:   Please bridge R1 with a piece of wire. Then solder the 2x5 DIL headers. Now you can apply power to the board (either through J1 or J3) to check the LEDs and choose the through hole resistors that set the LED sliders’ brightness. 10k was good for me. Note: please don’t send clocks/serial data to the BLM with the board’s power disconnected, it might damage the CMOS.  If a Core/miniCore module is connected then without MIDI input the LEDs should light with each button press (just use a loose button pad) in sequence: LED1, LED2, LED1+2, and off again. Cable wiring: Crimp a 10 pin IDC connector on about 40 cm of ribbon. Then solder the following pins on the DIN8 socket in order. It should be possible to do this outside the case and thread the IDC through the mounting hole. Thereafter a DIN8 cable can connect to the corresponding SEQ BLM connector, but you may want to wait to install everything together in the case.   Button pad preparation: Carefully cut 16 pieces of 1x4 buttons ensuring each has the two guide nubs at the bottom. I recommend scissors instead of a knife to keep it tidier as the silicone tends to unhelpfully flex under a knife. Also cut the corners off pads where M2 grub screws will enter from the front. Now the button pads may be placed (oriented so the nubs fit in the correct PCB holes). It’s best to go from one side to the other to keep everything flat.   Spacer Then place the acrylic spacer on top (mine is a bit different). There are extra rings of acrylic which can be used as “washers” to make up for the top and bottom right corners. You might want to add thinner metal washers as well to get the distance right. When ordering the spacer (e.g. from Formulor or Ponoko) make sure to choose opaque acrylic in 3mm thickness, probably white or black is best. If you buy milky or clear plastic some LED bleed might occur. BLM_spacer_3mm_acrylic.eps Apparently this one is better optimised into vector format, many thanks Phatline: BLM_spacer_3mm_Formulor.eps     Case assembly Lift the PCB into the case from the bottom (so your 289 buttons don't fall off) and loosely attach at least a few corners with M3 nuts. You need at least the centre M2 grub screw and can add the other four M2 screws if you feel more stability is needed. It’s probably best to thread them through the PCB from the rear; be careful as the thread pitch is quite fine (0.4 mm) and is cut into soft aluminium. I haven’t noticed any unscrewing, but you could add a drop of Loctite to be sure. Two M3x12 MF spacers attach on the top left corner (viewed from the back), this is where the miniCore will go. Next tighten the remaining nuts evenly; don’t overtighten as the PCB will flex! As long as the button pads were placed more or less in the right place they should sit evenly above the front panel. Or they may require reseating to get things even. Attach the miniCore module and a ribbon connector to J1 and panel mount the DIN socket. You need to jumper 4 unused AINs on the miniCore's J5B to ground which will prevent noise on these inputs.   Then close up the case, attach the bottom feet and you’re done! Congrats on finishing your BLM and I hope you enjoy using it!   Extras: I (or somebody else) will likely design a protection circuit so the extra AINs (4 available) can be used to control parameters on the SEQ. Modular voltage levels are NOT RECOMMENDED without this protection as outside the 0-5V range they will damage the PIC.   miniCore: Nothing too special here, just remember to mount the resistors first (i.e. before the DIP40 socket). The BLM data lines are buffered on the main PCB so potentially the 74HC125 chip could be left out. Also keep in mind that the female DIL header sockets must mount on the rear of the board. You can see that I made do with SIL versions that I cut into rows of five. Note: you can also use a PIC18F4685, currently stocked by SmashTV.   BOM: http://www.mouser.com/ProjectManager/ProjectDetail.aspx?AccessID=28e54a7e6f  Quantity Description Resistors 660-MF1/4DC1000F 1 100R 660-MF1/4DC2200F 3 220R 660-MF1/4DC1001F 3 1k 660-MF1/4DC1211F 0 1k21 (replaced by 1k and included above) 660-MF1/4DC5601F 0 5k6 (replaced by 4k7) 660-MF1/4DC4701F 1 4k7, solder in the 5k6 position 660-MF1/4DC1002F 2 10k, also usable for slider LED brightness in main PCB Capacitors 810-FK18C0G1H330J 2 C1,C2 33pF ceramic, 2.5mm spacing 594-K104M15X7RF53H5 3 C4,C7,C8 100n ceramic through hole (pull legs tightly through PCB bottom) 667-EEU-HD1H100 1 C3 10uF electrolytic 2.5mm spacing Diode 512-1N4148TR 1 1N4148 diode (can be taken from the 300 ordered with the BLM) Crystal 815-ABL-10-B2 1 10 MHz Crystal ICs 595-SN74HCT125N 1 74HCT125 quad buffer/level shifter 512-6N138M 1 6N138 optocoupler --- 1 PIC18F4620 from smashTV or other suppliers. Needs MBHP bootloader flashed. Hardware 517-929852-01-05-RA 2 2x5 DIL female header --- 2 2x5 DIL male header (snap from larger piece) 517-4840-6000-CP 1 DIP40 socket, quite expensive, try a local supplier --- 1 DIP14 socket   Parts substitutions: Change 1k2 resistor to 1k Change 5k6 resistor to 4k7   miniCore firmware: 1) The blm_scalar_v1_1 firmware can be downloaded from http://www.ucapps.de/mios_download.html It contains prebuilt hex files for various hardware configurations: project_without_ain.hex without AIN enabled (use this if J5 pins not connected to ground) project_with_4_mapped_ains.hex for Latigid On's BLM PCB layout, only 4 faders used project_with_4_unmapped_ains.hex AIN pins are mapped 1:1 (n.b. when using the BLM PCB/miniCore the sliders 1-4 are reversed in this configuration) project_with_8_mapped_ains.hex for Latigid On's BLM PCB layout if 4 faders + 4 extension AINs are used. It's recommended to jumper any unused AINs to ground project_with_8_unmapped_ains.hex AIN pins are mapped 1:1 (n.b. when using the BLM PCB/miniCore the sliders 1-4 are reversed in this configuration) 2) if you are using a PIC18F452, it's required to change the PIC device configuration while flashing the bootloader: the brown out reset level has to be changed from 4.5V to 2.7V as shown in this picture: Otherwise PIC could be reset sporadically (depending on the number of enabled LEDs) because the voltage level could fall below 4.5V! This change is only required for PIC18F452, other pics (such as PIC18F4620 and PIC18F4685 will work w/o this change). Unfortunately this change can only be done with a PIC programmer. Please contact TK. if you don't own a PIC programmer. He could send you a replacement PIC18F452. Alternatively use a PIC18F4620 or PIC18F4685 if you own one.       Acknowledgements Many thanks to TK. for a great concept to begin with and for providing firmware support to get my kludged prototype working. And to all builders: thanks for trusting me, it's not too cheap for DIY but I hope you agree the end product is worth it.     Addendum   It seems like some form of local power regulation is needed to ensure stable operation. As noted above, some PICs have brown-out detection built into the bootloader, so when many LEDs are lit, the PIC resets and won't work until the voltage rail stabilises over +4.5V. The simplest way I found was to interrupt the 5V line after the DIN socket (inside the BLM case) and add an adafruit Verter Buck Boost regulator. Mounting is stable using a SIL header on one row of the blank muck space, this can be connected to the unused functions on the small PCB.   Power consumption Using the Verter, standby voltage and current is 5.2V and about 30-60mA. With all LEDs lit (both colours) the draw goes up to 647mA, although this would never be done in normal operation. In track mode with all 16x16 blue LEDs lit the draw is 467mA and the voltage drops to 4.72V. I don't really perceive a difference in brightness, but it could only be a good thing! As long as the ADC inputs are scanned relative to the supply voltage you shouldn't lose control at the top end.      
  6. Chiphead (another stereo SID bassline demo :))

    Don't apologise, it's more than I've done in years! My comment was more of a compliment on how you can get very creative patterns/variations just with a few scene changes using the LoopA..
  7. Chiphead (another stereo SID bassline demo :))

    Who needs automation when you've got two hands and a LoopA?
  8. Troubleshooting midiphy SEQ v4+

    It's normal as we need to shift the serial I/O on pins 5/6 through the registers. It works somewhat like a bucket brigade delay if that makes sense.  You're doing great! Keep at it.  
  9. Troubleshooting midiphy SEQ v4+

    Replacing the ICs also includes RN1? Are digital inputs now working properly (no random messages in MIOS Studio)? Quite a few of the joints look "cold." Did you solder the THT parts from the top or bottom? When you say  this results in the picture above? It's not too surprising if so, as the static from your hands can toggle the CMOS gates. That's typically not the best way to control 74HC logic though -- the static can also damage ICs. A picture of the rear would also be helpful. You can try to control the pins directly from the MIOS Studio command line with: set dout z 0|1 where z is from 0-15 on the JA board, 0 is low, 1 is high. Reset the Core, reopen MIOS Studio and don't load an _NGC file. http://wiki.midibox.org/doku.php?id=seqv4plus_jog has the matrix schematic. LEDs will light up if the the Cx column (dout 1 to 7; 0 not used for LEDs) is LOW and the Ay row (dout 10-15) is HIGH. Try not to light up too many columns at once. Note that it's not a proper scan matrix, so can be that multiple LEDs light up at once.    
  10. Troubleshooting midiphy SEQ v4+

    Good to hear some issues are fixed! Please check the resistor network on the JA board is the bussed type (15 resistors), not the isolated type (8 resistors). The ICs may have been damaged with backwards cables. Pretty please a picture?
  11. midiphy SEQ v4+

            We're just about ready for launch of the new SEQ v4+. This thread serves as a proper introduction, you can also ask questions on hardware etc.  I first became aware of the SEQ around 2008. This was about the time of the migration point of V3 to V4 and the Wilba design. I finally got my hands on a Wilba front panel in 2012 and I was rocking the hardware soon after. Around the end of 2016 I felt there was room for improvement.  a number of my TL1100 buttons were failing on my unit there was still no "standard" case available, also the rear panel was not well defined the PCB supply situation was (IMO) non-transparent and confusing for new users I had ideas about hardware changes What started as a simple question morphed into a multi-page idea thread. It was clear that the idea of incorporating a 16x4 matrix was not feasible from a software perspective, but it at least catalysed the idea towards a flexible "selection row" and other handy features. i liked the idea of mechanical keyswitches for their quality and very reasonable price. I then worked on a concept of illuminating these, finally settling on a misused superflux LED through a Matias switch. It became clear that a "complete solution" would require redesign of all of the boards, so I spent the winter of '16/'17 doing that. This entailed sourcing hardware items such as high-quality OLED displays, Matias switches and keycaps, a new Core breakout board etc. I periodically checked in with @TK. to confirm the right direction for the hardware (matrix routing and button features). Without saying, the SEQ platform is an incredible piece of work and TK. deserves a lot of credit and praise. From the start, I was in contact with @AdrianH about potential case designs. He has taken all board outlines and done the CAD work himself. I think you will find the case an excellent solution to cramming a lot of hardware inside while keeping the panels clean.  Peter @Hawkeye is the co-partner of MIDIphy. He handles the business side and also webdev. Perhaps more importantly, he is a very enthusiastic builder and documenter of all the designs. Normally I will do a test version, but the product photos and demos are usually his. I have spent endless hours discussing and debating/reasoning with him and there is always a positive and fruitful outcome. Hats off to you!   So: thanks for all of your interest so far. The aim is to provide you the best opportunity to build your very own SEQ. Andy --- Video Build Tutorials ---   --- MIDIbox NG Downloads (for testing) --- SEQ_L.NGC fluxtest.ngr
  12. Troubleshooting midiphy SEQ v4+

    Okay. It's normally not an issue here but sometimes the boards benefit from a clean. It may depend on the solder/flux that you use. Do you use lead or lead-free solder?
  13. Display stopping Aoutng working

    Power issue pulling down the rail?
  14. Troubleshooting midiphy SEQ v4+

    What was the issue with SW-9-12? To test the MECs on SW17-20, like I said, there is no connection unless that switch is inserted (just the marked one). It's the mostly likely explanation. Inserting without soldering might work if the leads scratch through enough of the oxidation that naturally builds up on the through holes and component leads. Or it might not.
  15. Troubleshooting midiphy SEQ v4+

    For the row of MECs SW17-20, you do actually have to solder in the switch that has the "sink bridged by switch" label on the silkscreen. The metal part of the switch provides the connection. For SW9-12 I don't know. Wrong transistor type maybe?
  16. Troubleshooting midiphy SEQ v4+

    If SW1-4 work, then the problem is on the sink side. Check IC2 and associated resistors T3, T4 and the resistors/diodes adjacent. Are all diodes soldered with correct polarity?   Please post a picture with your question. It avoids guessing games and is more informative in most cases.  
  17. Troubleshooting midiphy SEQ v4+

    To be clear, I mean one similar to the middle tip of this picture, not the one on the right. It does mean that you have to solder with the iron the same way each time. Maybe my old tip was degraded or my technique wasn't right, but I never could get things going using the very end of the conical tip, rather I had to use the side.  
  18. midiphy SEQ v4+

    The power options are: 5V vreg, could be left out +5V rail from the case USB (power only) DB-25 cable (can suffer from voltage drop, thin wire)
  19. midiphy SEQ v4+

    Not measured yet, but the only hungry thing might be the Vreg. All LEDs are driven from the +/-12V rails.
  20. Using Midiphy eurorack modules with MIDIbox SID

    Seems like it should work: http://ucapps.de/midibox_sid_manual_cv.html I thought the 8 gates would be on J5s, but apparently you can also assign them to a shift register. This means that you can use the J89 on the LineTX, I guess connected to the end of the SRIO chain. You would probably need to build a new firmware .hex file. J6 on the MB-6582 would connect to the J19 header on Line TX. The pinout is different but you can use flying wires or DuPont cable/hot glue. Seems like the Line TX should fit in the Pactec case. If it hits the CS PCB, then you might find another location for it and use the alternate 26-pin IDC header with a corresponding DB-25 ribbon connector. 
  21. midiphy SEQ v4+

      Can't match Peter's video skills, but I hope this gives you an idea of the Expander modules.  
  22. midiphy SEQ v4+

    Okay, I changed the IDC8 connector for an Amphenol part in the BOM; at least in the datasheet this is 0.1mm wider. 649-71600-008LF It should be black and the stock availability is a bit better. If anyone orders and it's not okay, please post your experience.  
  23. midiphy SEQ v4+

    One idea is to try another brand of shrouded header, e.g. 710-61200821621 from Würth 30308-6002HB (the corresponding 3M pin header) is has no stock available, but it might be worth ordering a piece and letting Mouser know :).  
  24. midiphy SEQ v4+

    Congratulations on a great build and thanks for the feedback!   Nearly there!    
  25. Troubleshooting midiphy SEQ v4+

    Good work! After ~10 years of soldering with a conical 0.5mm tip, I have recently discovered the joy of a 1.6mm chisel tip. It really is night and day with SMT as you get great heat transfer to the flat pads (always heat the pad if possible), then apply 0.5mm solder to the pin et voilà!  With a conical tip I always had to use the side of the iron for SMT and the heating was much more tedious. Still the conical tip is better for larger THT components, pots and jacks etc. I have both irons at hand and just plug what's needed into the soldering station :)