 I have a new this. This is a cheap and cheerful voice recorder module and this is its storage which is a small flash chip but I thought it would be interesting to try and dump that flash chip and it immediately fails to detect the chip. So what I'm going to do is try and solder some actual pins onto that. But still nothing. Now I do have a plan C. Okay so oh wow that was easy. And now it's reading. Of course the cheap and nasty device is now non-functional. Well I can just solder this back on that's quite straightforward but it would be nice to be able to put the device on and off in order to play with it. But what I could do is to take some header sockets and solder these on. Here these. So I can do the same trick with bending the pins together and this will solder onto these terminals and then I can plug this thing into them. Of course I need eight. So grab the corner cutters. Okay so that's supposed to be about the same as this and the pins are roughly in the right place. The spacing between the yeah that's not going to work. I can just about get the spacing right by bending the pins like this but the problem is that these big chunky connectors butt up against the battery holder here. Oh yes and for fun let's try putting the battery back in and seeing what happens. So play back nothing play there's a record rather. Yep the beep means it's starting to record and normally you press it again and get a double beep but it's clearly failing to talk to the flash chip and just giving up is crashing. If there was any tracing it would be dumping error messages. Okay let's try some bits of wire then shall we. So now we take our chip connector. This one plugs in this side. This one plugs in this side like so. Let's put the battery in and we push the button and nothing happens. So chances are that we've bridged a couple of lines so let's just get the meter out and test this. No that was just me being clumsy. Okay that doesn't seem to be any obvious bridges. So is it wired up correctly? So pin one she's going to hopefully the right place. Pin two to hopefully the right place. Pin four is on the right pin. That is pin three whereas for the other side yes yes yes. So it should work so that's the same behavior we had before with the flash chip disconnected. It's failing to talk to the flash chip and is just giving up and discussed. So what could we have gone and got wrong? Well shorting pins for a start but we have just tested that. Could I have shorted them to something else? There are a few components which could be getting in the way. The other option is bad solder joints. Please think it's probably time for the oscilloscope again. Okay here we are and if I run it and go we get a capture but it looks kind of terrible and I'm wondering whether I actually have stray capacitance in these wires. That will cause everything to go horribly wrong. If this is trying to push eight megahertz signals across that. Let's just try record instead. Yeah. I don't know the order actually. I didn't plug them in the same way round as before. So channel one that's yellow should be chip select actually but that's not. There's a little bit of a glitch here but I don't believe it signifies anything. So here you get four pulses. This is probably the clock now and the data is distinctly poor quality which could be the long wires. The chip is definitely not responding. That's a little bit better. Yeah you can see the way that this is just a slewing up. I think that this is Mozi, a master out-slave in. This is the processor talking to the flash chip which this must be the response. Possibly this is just not connected properly. So that is channel two. That is this wire. Make sure that hooked on. Is it noticeably noisier than the others? I think it could be. So let's take a closer look and hook all this stuff. This is why I use test clips you see. Do we actually have a decent connection? The pad down here looks alright I have to say. So that's this one connecting to here looks alright to me. I think the wires might just be too long which is a shame as I'd have to resolder everything. Although now I think of it. Remember that pin I didn't hook up? This is hold. It appears to turn everything off on the chip. Now if that is floating low to make it asserted then that means the chip will just not be responding which is what we're seeing here. So let me just actually try bridging that just to tie it to power the way it was originally and let's see what happens. Losing stuff all over. This is clumsy as hell. I need to try and solder this on without losing the wire like that. Never mind. Now I can solder this on to here if I can make it stay put. It's just too springy. Also important safety tip don't try to stabilize the soldering iron by resting it on your finger. I've actually done that once. Didn't work well. Okay so oh well I actually forgot to turn that off so let's see how that behaves now. Nope no change. It's still just refusing to talk to it. I will admit to being kind of stumped. I'll do some investigation offline. This is a horrible tangle and see what I come up with. There has been a breakthrough. Watch this. So the secret was that by extending the wires to the flash chip I also needed to add in a decoupling capacitor that is a simple .1 microfarad electrolytic connected between the power and ground lines of the flash chip. And if I trigger another capture it failed to trigger at the right places. Do that again. There we go. So that you can see that we now get rather less glitchy signals. However the clock which is the blue one is not really what I would call square and there's some ringing there when the clock stops. However it appears both the processor and the SBI chip are capable of decoding the resulting signals so that's fine. Probably what this needs is remove some of these probes now. Probably what this needs is a capacitor with shorter leads because it's still quite a long way from the actual chip. But now the thing is actually working which is great and it records to testing. So that is good. Right well there is one more thing which I want to do which is I want to take the chip out of the ZIF socket which is actually straightforward. And now it won't work anymore. Who would have thought? It still thinks it's recording though. Hasn't gone anywhere. Yes. Since I have this kit containing a SOIC adapter I'm actually going to stick this thing on the adapter. That means I won't have to use this big bulky ZIF socket and it'll also give me some much needed practice in soldering these SOIC chips onto things. So let me get the last bit out of the bag. There we go. So I will clear this mess off my desk and get set up for doing that. All right. So what I'm going to do is to solder the chip onto this adapter board. In fact the pin spacing is incorrect. Good. It's got the right spacing in this side. So I'm going to solder the chip onto this adapter board by placing it approximately here. I haven't actually done this before. I'm going to use the hot air gun for that. So one option is to use solder paste and I don't have any. You basically it's soldered in a sort of goopy flux like matrix which you spread onto the pads and then you heat it up and it melts into actual solder. But I'm going to have to use actual solder. So I'm just going to tin these pads and there should be all the solder we need like so. Just check that for bridges. Now that looks fine. And then we're going to take the chip and line it up carefully on the pads which it will now will not go down because the solder on the pads is lumpy. That was almost correct. So we get out the hot air gun. Which you can hear start up. It gets up to temperature and then we just toast it and it should just drop into place except it hasn't because it's gone skew. So let's just take that off again. Come on let's melt the solder. Okay this is not working. I need it to be placed more accurately on the pads. So let's try it like that and apply some a little bit of pressure. Just make it go down. Okay that's skew. This is fiddly until I need the practice. Okay let's clean this solder off and try this again. Make it a bit less lumpy. That looks more like it. Okay so I have a little bit of pressure. This is just going to try and keep the chip in place. Although putting the tweezers onto the chip does seem to knock it off. Okay now toast it. Maybe I don't want any pressure. Let's try a little more distance because I just wanted to drop down. Okay I can see the flux melting. What's that done? Ouch. To nobody's great surprise it's quite hot but yeah that seems to have worked. So there is only one bit left to do which is to solder the legs on and I'm just going to get a little bit of help to do that. So the help consists of this same piece of breadboard that I was using before. All we do is we push the two bits of head up in through the board and then mount that on the breadboard and that will keep everything upright. So this is just a really straightforward piece of through hole pins. Ah that was the hot air gun finally powering down. I don't know how much of that came over on the camera. Okay so now we should have a nice dip adapter containing a two megabyte flash chip. So let's get this thing out. Pin one is there. So we remove the ZIF socket. Now it goes this way up like so. Does it play back? No it does not. That could be because it's backwards. Yeah so pin one should be here. It should be this way or this way around which I think was the way I had it. Right it's not recording so it's failing to talk to the flash chip. Let's double check the continuity because it may be that it's not the right way around. So this is pin one of the chip. It goes there. Yeah yeah yeah yeah yeah yes and yes. Okay so the adapter works. So why is it not talking to the chip? Well the answer is very likely going to be to do with that decoupling capacitor. And this will probably need to be changed to something bigger or smaller because this thing is going to have weird characteristics which this little adapter will not have. So let's take this off. Does that make a difference? No. Okay I'll have to experiment with a few different capacitors and see what comes out. That was annoying. So after some debugging I think I've got a handle on what's going on. This included getting all the relief probes out again. But if I do continuity test from this pin to the corresponding pin here you should be able to hear that beeping. However if I go up here there's no connection and that is clock. So nothing will work unless that's soldered up. So I think that that's just a dodgy solder joint. So let's just reflow that with the soldering iron. Let's try that again. So it's this one to here. Okay so let's stick the battery back in, push the button and nothing! Fantastic. So the adapter goes this way up. And in fact this thing is now unplugged but it will form a quite useful holder. So this should get continuity here, here and nowhere else. Okay so this is pin one so this should go here. Interesting. Okay let's try it the right way around this time. This is pin one. That should go here. And it's not, no it should go here. The one with the dots in it. Yes. So the solder gets contact, the leg gets contact. Solder, leg. Solder, leg. Solder, leg. Be here. Solder, leg. Solder, leg. Solder, leg. Solder, leg. Solder. Huh. Now I'm sure I tested this previously and that worked. Okay so we reflow it. Let's give that another go. Solder, leg. Okay we're going to plug it back into here. Now pin one is because this way around. Alright if you push the button and nothing, no beeps at all. So I think I may have plugged this in backwards. Yes I did because this way around it really doesn't help that the chip is fastened to the adapter at 90 degrees to the adapter pin out. Okay nothing. It's possible I managed to fry it somehow. Let's try that in the EEPROM programmer again. Okay so we drop the thing in here, run the test. Okay it has correctly detected the device but it hasn't read anything. Oh yes because I need to tell it it's a 25F, no. Now I need to tell it it's a PN25F16B at Zoic 8. Okay yes we want it to go ahead. Right it's reading. The chip is fine so it's something to do with the way that this board is talking to it that's making it go wrong. I'll just wait for this to finish so that I can take a quick look at the hex dump just to make it look see if it looks sane which it does. Problem solved. Yeah oh I took this off and started working again. So clearly the characteristics of this ZIF adapter are so different from this that this requires the decoupling capacitor but this mustn't have one. That is weird. I am very clearly going to need to learn more about decoupling capacitors. Yeah so that's all kind of strange. Anyway that has been quite successful. I have done a number of nasty bodges to this thing and it now still works more or less. Well as well as it ever did there's some rather useful bits I can pull off this such as I now have a handy two megabyte flash chip and a dip socket. There's a bad microphone. There's some button cells. So I think that when the real hardware that I actually want to experiment with shows up I will be in a reasonably good place to figure out how it works. Hello what's this? What's this wire important? Apparently it wasn't. It's the power supply for the chip. Oops that should go on here. Was this what was causing the problem? No apparently that was it. It was just a capacitor that wasn't wanted and it's getting enough parasitic power through its data lines to work fine. Well I thought that was entertainingly random. I hope you enjoyed this video. As always please let me know what you think in the comments.