EEVblog Electronics Community Forum
Electronics => Beginners => Topic started by: Rooster Cogburn on March 26, 2019, 11:07:11 am
-
I bought one of those dirt cheap function generators. Doesn't really work, the output seems completely broken above a few Khz. I've read in other places that this is a common problem. Also read a comment by one user about replacing the TL072 opamps with LF412 ones, which made a significant different for him. I could do that (yay!).
Is that a good idea? Has anybody else here managed to fix / improve this el-cheapo function generator? I'd love to give it a shot and see if I can turn it into something useful.
-
Doesn't seem worth the effort, at least now you've got a spare ATmega48, 1602 LCD, BNC, some switches and a project case..
https://hackaday.com/2018/05/17/review-fg-100-dds-function-generator/
-
Yes, most cheap gadgets are worth buying for the parts alone :)
But here, a comment from that review you linked:
"I have replaced the TL072’s with LF412’s. Now the signals are quite good up to 200 KHz than the Amplitude goes down to 5Vss at 500 KHz, but stll as SINE."
If that's the case I'd happily tack on two bucks worth of opamps on my next component order and spent 5min swapping them out. That's absolutely worth it to me.
-
If 200 KHz is all you need, fill your boots. For me, shipping those puny parts would cost a lot more than 2 bucks..
-
It would be a ~100x improvement over what it does now :)
I could get some parts from AliExpress, but who knows what I'll get. Shipping cost from a reputable local vendor would be >5EUR, but that's why I said "happily tack on two bucks worth of opamps on my next component order".
But you think doing a TL072 -> LF412 opamp swap is a sound idea? Looking at the data sheets (I'm not very knowledgeable about opamps) it seems like it should work. If there are other components that could be replaced with better performing ones or small modifications I could make to the circuit, I'd want to give that a try as well.
-
I don't think there's any harm in linking this thread (https://www.eevblog.com/forum/beginners/improving-the-fg-100-dds-function-generator) in a single post on the FY6600 thread (https://www.eevblog.com/forum/testgear/feeltech-fy6600-60mhz-2-ch-vco-function-arbitrary-waveform-signal-generator/) asking for simple upgrade help?
The guys posting there have been modding that DDS for ages.. (and if you're unsure, send a PM to Simon asking if it's OK..)
-
I looked a bit more and it sure doesn't look great.. In this video (at 13 mins) there's crossover distortion even at 5khz.
https://www.youtube.com/watch?v=r_il0MVB_Rs (https://www.youtube.com/watch?v=r_il0MVB_Rs)
-
Oh yeah, it's a <20EUR piece of crap. I'm just playing around, learning etc.
On the YT video you linked there's another comment suggesting an opamp replacement:
"I have the same function generator and got the same problem with bag output. Tl072 are counterfeit (working in class C=Crossover distortions) and after replaced them with ne5532, the signal it's perfect until 500 khz."
Who's this Simon fella I should hit up? (sorry, still kinda new here)
-
He's Moderator, be brief, he does a lot. (I PM him usually only when a few bad apples impinge the health of the forum).
-
A few years ago I developed alternative firmware for the FG-100 (see https://github.com/timsavage/fg100alt-firmware (https://github.com/timsavage/fg100alt-firmware)).
From the feedback I have received the quality of these units seems to vary greatly. The early units are good, but there are now knock offs made from much cheaper parts that have pretty awful output.
The unit I started with doesn't exhibit any of the distortion shown in reviews of later units, even at 500Khz it's still clean.
-
I realize this is a stale thread, but in a recent thread BillyO linked a couple of recent videos about upgrading this cheap-and-not-worth-it device so I thought I would link it here and post my experience as well :
https://www.eevblog.com/forum/beginners/how-would-i-go-about-creating-a-variable-tone-generator-30khz-to-50khz/msg4950109/#msg4950109 (https://www.eevblog.com/forum/beginners/how-would-i-go-about-creating-a-variable-tone-generator-30khz-to-50khz/msg4950109/#msg4950109)
I basically found the same thing - mine was completely useless out of the box (even worse than in that video), so I replaced the opamps with opa2992 (10.6 MHz, 32 V/us, 65 mA output drive, $2.50 each). Was a little disappointed that the 3 dB bandwidth was only 350 kHz or so, since I was expecting better with these opamps. While I had it apart I also figured out the configuration of the output stage. The sequence of opamp stages seems reasonable: first remove DAC DC offset, then an optional lowpass filter (Sallen-Key), followed by the AC amplitude adjustment and finally adding the desired DC offset. I do think a passive lowpass filter would have been better, but whatever.
The filter (engaged when button is up) does reduce distortion and artifacts on sinusoids at the higher frequencies and adds a little overshoot in square waves. With the new opamps it can produce 4.2Vpp sinusoids at 500 kHz with about 0.6% THD, and significantly higher voltage but distortion is > 3% so is pretty poor. I have attached some figures just to show what the 'improved' output looks like. Most of the plots are for the fg-100 connected directly to my scope (picoscope 5244b) directly with a bnc cable with no 50-Ohm feed-through; the last plot is a 20 kHz sinusoid at 7Vpp with a 50-Ohm feed-through, which is about as much current as the opamp can provide and the distortion was the same as when delivering 14Vpp into high impedance. I also ran 7Vpp square waves into 50 Ohms, but after a minute I felt the output opamp and it was getting pretty warm.
The main reason I got it was so that I would have a battery powered function generator that I could use without regard to circuit and scope ground references. And for the $25 I have invested in this ($20 delivered; $5 opamps) it is an okay deal.
jason