2023-10-22 - Fixing a Roland FP-8 Electric Piano ------------------------------------------------ I was recently tasked with fixing an old Roland FP-8 electric piano. This phlog post is just meant to document what I did to fix it, in case someone else ever finds this information to be useful. Quiet Volume ------------ The main issue was that the internal speakers were very quiet. This is apparently a common problem caused by faulty electrolytic capacitors[0]. Since I've recently been trying to learn more about soldering, I decided to try replacing them by myself. To open the piano, some screws underneath need to first be removed (four on each side)[1]. Then, the top part of the piano can be flipped away from the keyboard part[2]. To make accessing the mainboard easier, the top part can be completely removed from the keyboard part by disconnecting two ribbon cables and unscrewing a lot of screws[3]. I was originally going to remove the mainboard to make soldering easier, but the cables connected to it were very difficult to remove, so I decided to just solder with the mainboard still inside the frame. This piano had been previously repaired by someone who didn't have the correct capacitors, so he replaced most of the SMD capacitors with THT ones[4]. I first only replaced the three capacitors that the previous repair person had left as SMD components (it seemed as if they might still have been original). That already fixed the problem, but I decided to just replace all capacitors while I was at it. An aside on replacing SMD electrolytic capacitors: The "correct" method for removing these is a hot air station, but I really haven't had much luck with that method. You have to use a lot of heat until the solder actually melts, potentially causing other things to be destroyed. Apparently, the capacitors can also explode sometimes when using hot air. Another way is to add some more solder on both sides, then alternately heat each side, always pulling the capacitor up a bit on that side. Whether that works depends on various factors, such as the board size/thickness and the type of solder used. On some boards, this method worked very well, on others, I needed to use a lot of extra flux and heat the connections for a long time until the solder underneath the capacitors properly melted. Then there are two physical methods that can be used. One is to twist the metal can of the capacitor while pressing it down. This method might be somewhat risky, but I haven't managed to rip any traces off so far in my experiments (it's important that the metal can isn't just ripped off, but instead twisted while pushing it down flat onto the board). This was still a bit risky for me, especially on such an old board, so I ended up just cutting the capacitors apart. If you cut the capacitor as near to the board as possible, you can pick the remaining pieces apart and remove the pins individually with a soldering iron. As far as I know, this should be relatively safe as long as you don't pull the capacitor while cutting it. Soldering the new capacitors in is the easy part. Just remove the old solder from the pads with solder wick, clean them with isopropyl alcohol, then add flux, place the capacitor on top, and touch the soldering iron tip to both sides with a blob of solder melted on it. The extra flux helps the solder flow underneath the capacitor. Make sure to place the capacitors the right way around since they're polarized (the side with the black stripe is the negative side). I bought capacitors from the Panasonic FC series, but I really don't know enough about electronics to be able to decide which capacitors would be best for this application. The needed capacitors for the mainboard are as follows: * 6x 10uF/16V/4x5.4mm (size B) * 4x 47uF/16V/6.3x5.8mm (size D) * 1x 100uF/6.3V/6.3x5.8mm (size D) Unfortunately, the previous repair person had also scratched the board in a few places. I took that as an excuse to finally get UV curable solder mask and cover the scratches. [5] shows the board after replacing all capacitors and covering the scratches with solder mask. [6] shows the same image, but with the capacitors labeled for reference. Note that I did not touch any of the ceramic capacitors or resistors, but the previous repair person seems to have messed with some of them. A few of the pads were also partially lifted or broken, but luckily, none of them were completely gone. Some of the damage may have been due to leaked electrolyte from the old capacitors (which was already removed by the previous repair person), but there really isn't any way to tell now. Broken Hammers -------------- Another common issue with this piano is broken hammers on the keys. Most of the hammers on this particular piano had already been replaced in the past, but they were already cracking again. Since these hammers are getting more and more difficult to obtain, I decided to glue them with two-component epoxy adhesive (specifically "UHU Endfest", but there are a lot of similar glues). To remove the keyboard assembly, some more screws need to be removed, including two on the underside of the piano (I'm not entirely sure how many screws are supposed to be removed because some were already missing anyways)[7]. To remove the keys, I found it works best to pull them apart by simultaneously pulling in one direction with a finger and pushing in the other direction with a flathead screwdriver[8]. The hammers can then be removed from underneath the keys[9]. None of the hammers were completely broken this time, but I decided to put some glue into the cracks to hopefully prolong their lives a bit. The bottom part of the piano is quite convenient to hold the hammers while the glue is hardening[10]. Whether this whole endeavor was sensible remains to be seen. One problem is that the hammers have to be pushed through fairly tight spaces to put them back in, so there can't be any extra glue on the sides. It probably only makes sense to do this when there are larger cracks or the hammers are completely broken so there's more space for the glue. Broken Volume Sliders --------------------- Finally, I fixed the broken volume sliders. I first wanted to try to clean them like I did with some volume controls in a previous article[11], but it turns out that these ones were completely scratched up inside[12], so no amount of cleaning would have fixed that. It might have been possible to find these as generic parts somewhere, but I decided to just spend the money on parts that were guaranteed to be compatible[13]. As mentioned in the description on the linked website, the levers on these are slightly thinner than the ones on the original parts, so I added some tape inside before sticking the knobs on top[14]. While I was at it, I also covered some scratches on the circuit board with solder mask (yes, I really was just looking for excuses to use my solder mask at this point)[15]. The solder bridges that can be seen between some pins don't matter because those pins are connected anyways. Note that some of the solder pads are partially broken. That must have been done by the previous repair person because they were already that way when I first saw them (it was quite obvious that the volume sliders had been desoldered previously - perhaps they had been swapped as mentioned in [16]). Luckily, none of them were completely broken, so I didn't need to additionally learn how to do pad repair. Technical note: These are 10K stereo potentiometers with a logarithmic scale. When looking at the pins with the orientation in [17], the pin at the very left is the logical second pin, the second-to-left pin is the logical first pin, and the two pins at the right are both the logical third pin (the two rows of pins each control one of the two channels since it's a stereo potentiometer). The slider is at its minimum position when it is at the left, seen from the angle in the linked picture. This means that the resistance between the logical first and logical second pin is about 0 Ohm, and the resistance between the logical third and logical second pin is about 10 kOhm. The middle pins are just for mechanical stability. If no potentiometers in the correct size/pin configuration are available, I suppose it would also be possible (although slightly hacky) to just get any 10K logarithmic stereo potentiometer, mount it on the outside of the piano, and connect wires from its pins to the correct solder pads inside. Remaining Problems ------------------ After fixing all the other problems, I noticed that some of the buttons on the front panel don't work very well anymore. It shouldn't be too difficult to replace these, but the person playing the piano doesn't need them very often, so I decided to leave them in that state for now. [0] https://forum.pianoworld.com/ubbthreads.php/topics/1629176/How_loud_Roland_FP-8_supposed_.html [1] gopher://lumidify.org/I/phlog/2023-10-22-piano/piano_bottom.jpg [2] gopher://lumidify.org/I/phlog/2023-10-22-piano/piano_open.jpg [3] gopher://lumidify.org/I/phlog/2023-10-22-piano/piano_top.jpg [4] gopher://lumidify.org/I/phlog/2023-10-22-piano/pcb_before.jpg [5] gopher://lumidify.org/I/phlog/2023-10-22-piano/pcb_after.jpg [6] gopher://lumidify.org/I/phlog/2023-10-22-piano/pcb_after_labeled.jpg [7] gopher://lumidify.org/I/phlog/2023-10-22-piano/keyboard_removed.jpg [8] gopher://lumidify.org/I/phlog/2023-10-22-piano/key_removal.jpg [9] gopher://lumidify.org/I/phlog/2023-10-22-piano/hammer_removal.jpg [10] gopher://lumidify.org/I/phlog/2023-10-22-piano/key_rack.jpg [11] gopher://lumidify.org/0/phlog/2023-10-15-speakers.txt [12] gopher://lumidify.org/I/phlog/2023-10-22-piano/scratched_slider.jpg [13] https://www.synth-parts.com/produkte/potentiometer/schiebe-potentiometer/5397/volume-schiebepotentiometer-fuer-roland-ewa-na0x10a14 [14] gopher://lumidify.org/I/phlog/2023-10-22-piano/lever_tape.jpg [15] gopher://lumidify.org/I/phlog/2023-10-22-piano/slider_pcb.jpg [16] http://www.shoutingelectronics.com/2015/10/14/roland-fp-8-keyboard-repair-capacitors-and-hammers/ [17] gopher://lumidify.org/I/phlog/2023-10-22-piano/slider_pins.jpg