 Hello, and welcome to another exciting episode of Rob Tries to Fix a Thing. Rob Tries to Fix a Thing! Today's thing is this HP3325A Synthesizer Function Generator. It synthesizes and generates functions. Sine waves and square waves, triangle waves and sawtooth waves, not at very high frequencies, only up to 20 megahertz for the sine wave and 10 megahertz for the square wave and even less for the triangular waves. But it's got buttons and it's got a nice little display and, well, there's a problem with it. The problem is that the power switch is missing. So how are we going to turn this thing on? By opening it up. Now you might have noticed that I just opened up the bottom side. That's because the switch is on the bottom. So the switch should be over here and usually there's a plastic rod that runs all the way back to a switch in the back. I don't actually see any switch in the back. In fact, if we look over here in the corner where the switch is supposed to be, we actually see two dangling wires. Maybe the switch was actually over here, in which case I'm going to have to replace that or hokey something up. So here we are looking at the back of the keyboard and display. This is called Service Group A. And according to the manual, the first thing you have to do in order to get this out is to remove this ribbon cable and also the signal and sync cables. Now it doesn't look like they're labeled. Oh, actually they are. This one says to sync and this one says one signal. So that's great. So we can just remove those and remove the ribbon cable and then proceed. We're now looking at the top of the unit and I've taken the top off and this is the ribbon cable. So, ow! Just look at that old school cable. It hurts. Next, the manual says we have to remove the plastic trim strip using a screwdriver or a similar tool or perhaps a rock or something. What it's talking about is the thing that used to be over here. Well, it's not there. So I guess we can move on. Remove the two screws from the top of the front frame beneath the trim strip. There's one here and one here. And two corresponding screws from the bottom side of the front frame. So I can do that now. Okay, and then it says to push the printed circuit board and front panel assembly forward to remove the front frame. Oh, that was easy. Now the ribbon cable appears to be stuck on that funny piece of plastic trim. So I'm going to have to loosen that up. Finally, remove the ten screws that hold the printed circuit board to the front panel assembly. So there are small screws all throughout here, which I will have to remove in order to examine the circuit. And that's the ten screws. And now we should be able to remove the printed circuit board. Well, no, it doesn't look like we can. For one thing, there is a wire that is soldered onto the board from this plug that goes to the front panel. So, and there's another one over here, which means that I will probably have to unscrew these connectors. And then I can just remove the board. So this is interesting. When I was removing this, I noticed that the little wire that was soldered to here that was supposed to be soldered to this tab apparently broke. Maybe as I was doing this or it was already broken. Anyway, it's simple enough to repair. There's an extra screw, eleven screws. Okay. And there we go. So that's the display and keyboard assembly. Okay. Well, I'm back after a few days and I got from Digi-Key these switches. They're not push button switches like the ones that are specified in the manual because I couldn't find any that were small enough or thin enough, I guess. This opening was rather small. I think it's something like 15 millimeters across and 10 millimeters wide. So really the only thing that I can think of is to use these tiny little toggle switches. Really all you need is just two positions and, you know, this is a single pole, single throw. So I thought that hopefully it will actually fit. I can try it. Yeah. Okay. Well, that didn't really work very well. So I just visited the surplus store in order to find just random push buttons that latch that could be used as the power button for this because I'm having a really hard time finding anything suitable. And I also picked up a toggle switch, which of course is too big, but I figured it's nice to have a toggle switch. So all of these buttons look pretty suitable. They've got the square sort of end. They all look like kind of the sort of switches that would go in here that I've seen on the web. This one is an interesting one. It's rather big and chunky and it would sort of fit right there. And in fact, it kind of looks like it would fit, you know, even with these little leaves here. They do line up with these little cutouts for screws. And of course it does stick out a little bit like it should because then a button would go on top of that. So this is certainly a possibility. I also found this in a bag. This is actually, it is an HP switch push button. It's not the right one though for this different part number, but that's kind of cool. Now I have an OEM HP switch push button. And then I found this which came with a little wire harness and it, you know, it's also a latching power switch. And it has some sort of amount. And I looked at this as well and it actually does line up properly with the holes as well. So it's sort of like this is some sort of a standard. And in case you're wondering, this is a C and K switch, C and K, ampersand K. C and K make some really good switches. I like them. And the switch type appears to be an NE-18 series. The picture on the web shows that there are actually pins coming out the bottom. This one doesn't have pins, but I couldn't find that. But other than that, it looks pretty much identical with the exception of this special mounting bracket, which I think is actually custom made. So it turns out that those holes were just slightly smaller than a 440 screw. That's a number 440 screw, which is a US screw. It's not a metric screw. Of course it's not a metric screw. So I just tapped the 440 threads through those holes and attached those screws in there. And now it's pretty sturdy. So there's the front. And I can push it in and push it out and it doesn't break. So now I'll just solder these two wires onto one of the sides. I guess the closer side, of course. And then I'll put the panel back and see what happens when I turn this thing on. Okay, so I went into the panel and made sure that the button was working properly. And I traced out the signals to the ribbon cable. And I think what was wrong was that the ribbon cable was actually on backwards. So let's try this again. Ah, there we go. Well, the fan went on. And we're getting the number 1000 hertz sine wave. So entry frequency. Well, let's measure the signal and see if we're getting anything out. Alright, so first of all, this is what is coming out of the sync output. This is not the frequency output. The sync output is a square wave that is in phase with the output signal. So I may have made a mistake and I believe the mistake that I made was that I was on the wrong time base on the oscilloscope. Because I checked the sync output and I also wasn't getting anything, which I know that I was before. So the problem was that the scale was off. So now there are a great many tests listed in the manual for this thing. Unfortunately, a lot of them have to do with things like harmonic distortion and so forth that I would need a spectrum analyzer for. And I don't yet have a spectrum analyzer, but we can do some basic tests. So one of the first tests that the manual wants us to do is to output a sine wave at 20 megahertz and an amplitude of 10 volts. And we would connect this up to an oscilloscope that has a 50 ohm input. Now I don't have a 50 ohm input, but I do have an oscilloscope. So I'm just going to set this to 20 megahertz frequency, 20 megahertz. And I'm going to set the amplitude not to 1 millivolt, but to 10 volts. Alright, so we should be getting something on the output. Let's take a look. Okay, so the first thing we're seeing is, well, we're definitely seeing a sine wave. However, the oscilloscope says it's 240 hertz. Now that's a thing with digital oscilloscopes. With an analog oscilloscope, you would just see hash over here. It would just be basically filled with light. What this is doing is it's doing aliasing because it's sampling the signal every, you know, maybe a few microseconds. So what I need to do is I need to expand the scale and you can see that we are getting a much better signal. And the frequency does appear to be at 20 megahertz. We are seeing a peak to peak voltage of 22, 23 volts or so. Again, we aren't terminated. So first of all, the signal integrity is not going to be great. And second of all, the voltage is going to be roughly approximately twice what the thing should actually be outputting. But this is pretty good for me. What I really should do is get either a 50 ohm load or what they call a 50 ohm load feed through, which basically is a resistor of 50 ohms that you connect up to the BMC cable. And then you can measure it from there. I did find one of those feed through connectors on eBay, but, you know, whatever. So the next test is to go to 1 megahertz or 10 megahertz. I'm actually not going to do that because I'm pretty convinced that this is working pretty much okay. Again, I cannot do any calibration because I don't have a spectrum analyzer. So all I really care about is that it's outputting, you know, something that's reasonably close to what it should be. One of the next things to do is look at a square wave. All right. What happened? Ah, there we go. Okay. Well, we've got a square wave. It does kind of look like 10 megahertz or thereabouts. The square wave rise time between the 10% and 90% points should be measured at 2 kilohertz. And the rise time should be less than 20 nanoseconds. That's pretty impressive. So let's change to 2 kilohertz. I'm going to have to dial this all the way back. Change to 2 kilohertz and one still 10 volts. Yeah. Okay. 2 kilohertz. 2 kilohertz. All right. And we should be measuring a rise time of less than 20 nanoseconds. So my oscilloscope claims that the rise time is 10 nanoseconds. That's pretty good. Let's take a look at the fall time. And we have a fall time of about 12 or so nanoseconds. So that seems to be pretty good. Next thing is to look at the triangle wave. So we are going to set a triangle wave for 10 kilohertz. So let's go ahead and change to triangle wave and 10 kilohertz and see what we're getting on the output. Okay. That looks like a triangle wave to me. Let me just go to the different triangle waves so that we can see them. Okay. So there's one triangle wave and here's the opposite triangle wave. There it is. So that seems to be working fine for me. Everything else basically requires equipment that I don't have. So things like harmonic distortion, amplitude flatness, things like that, which obviously I'm not going to do at this point, but I'm pretty satisfied with this right now. We can see that we have sine waves, square waves, triangle waves, all sorts of different triangle waves. Add approximately the correct frequency and voltage, which is all I really care about at this point. So I think we can count this as a success. So Rob tries to fix a thing and succeeds. Excellent. Well, see you next time. Rob tries to fix a thing.