The assembly instructions are better than the oscilloscope ones, by leaps and bounds. Here's a picture of the finished article:
Triangles can go to 6Vpp, while sine waves only go to about 5.2Vpp before the peaks start flattening out.
I had high hopes for this function generator -- hopes which were admittedly higher than they should've been. This is, after all, a $35 function generator. Some things which really started to grate on me:
- Frequency selection is really really really really twitchy. You use the coarse pot to get close, and then use the fine pot to get closer. Trying to get within 10Hz of 1kHz or 10kHz is annoyingly hard. You get the coarse pot within a thousand Hz or so, and then use the fine pot to zero in. Only the fine pot moves the frequency really really fast. What feel like microscoping movements are enough to send the frequency zooming off by a hundred Hz or so. I'd give my left arm for some sort of calibration (and lower-resistance pots, though I could presumably do that myself) so I could easily hit a given frequency without having to spend 5-10 minutes fighting the pots.
- Once you finally get the desired frequency, sometimes the function generator will drift off to another one -- 100Hz or so away. Which means you get to do it again.
- The square wave is actually just TTL out, as mentioned above. So it's 0V to +5V. That's great as long as all you need is a square wave offset by 2.5V, but it doesn't work so well if you want -2.5V to +2.5V. And there's no offset knob.
- It wants 15V DC to work properly. It's hard to find a 15V DC wall wart -- I ended up having to get one from Mouser, and the one I got was noisy enough as to throw off the function generator.
- Now I'm just being greedy, but there's no sweep. I missed sweep when I started working on RC filters. The chip can apparently do sweep, but the circuit doesn't implement it.