These are quite amazing for the money.I have been reading but haven't been able to wrap my brain around what to get. I would like a piece of equipment to test square waves at 1khz and 10khz for testing amplifiers for oscillation etc.
Looking for something in the $100-$200 range. I was hoping some of you may have some good suggestions. Maybe something I can order on Amazon. Sorry for the newb request. This will be my first Function generator and I feel like I am running around in circles trying to get something to produce a nice square wave to test with.
Thank you!
2 channel and can do sweep and burst (most useful) as well.
https://www.ebay.com/itm/133511828215
You already have the 1Khz square wave generator, the probe test port on the oscilloscope.
Calibration output circuit on my oscilloscope. attached
Should work if if what you are testing is not DC coupled. If it is, add a capacitor in series and a resietor to gnd.
Calibration output circuit on my oscilloscope. attached
Should work if if what you are testing is not DC coupled. If it is, add a capacitor in series and a resietor to gnd.
Attachments
Last edited:
Inserting a capacitor in series with that output will compromise its square wave accuracy depending on the loading.You already have the 1Khz square wave generator, the probe test port on the oscilloscope.
Calibration output circuit on my oscilloscope. attached
Should work if if what you are testing is not DC coupled. If it is, add a capacitor in series and a resietor to gnd.
It was designed to only calibrate probes.
Remember, a cap causes a phase shift which is dependent on frequency and impedence.
I can recommend a FY6900 DDS that you can get for around 100$.
With two channels available you can easily use it to test balanced gear (just offset one channel by 180º).
I just wish it had slightly lower output impedance (it is standard 600R).
With two channels available you can easily use it to test balanced gear (just offset one channel by 180º).
I just wish it had slightly lower output impedance (it is standard 600R).
This Koolertron signal generator looks pretty neat. I ordered it. Free returns if It doesn't work out and I will receive it in 5 days. I will play with it a little bit and report back. Here is a 2nd video talking about the software that you can use on your computer to make arbitrary waves etc. Thanks for all of the input.
Yeah, good, but. Note the dual diodes and trim-pot. This trims-off the asymmetry of naked CMOS. In the factory they gave the women a marked-up oscilloscope screen saying which trim for what error, maybe 10 seconds per unit. But setting up that calibration jig is time and brains, maybe more than it is worth. A revised plan sets the oscillator at double frequency and flip-flop divides by 2 to get a symmetric output. 50 cents more parts against hours of jig-making.... one scheme works better in mass production, the other for a one-off.Calibration output circuit on my oscilloscope.
Yes, a series cap adds phase but the alternative is a dual-polarity supply. Or the insight that you don't need ZERO phase, you only need to be much-much better than your most mediocre amp. 10uFd 22k gets down to a part-Hz, much better than 20Hz the nominal audio limit, and insanely less phase shift than any speaker or headphone. The paranoid or low-Z can easily find 100uFd; or feed a 50k amp input with a DC-coupled output.
For that matter we should not need to worry about the low-end on most transistor amps. The request for 1kHz and 10kHz suggests that the high end is more important, and that's usually where you find trouble.
Much sine-wave "error" can be allowed for by in/out comparison. Say the raw square has 10% pip and 5% tilt, is the DUT output similar or worse? Square-wave is a quick-check.... if you need measurements, probably better to use sweep techniques (why the HP 200A/B has the big knob).
I think I've got similar requirements as the OP. I've used sound output from the Mac mini headphone port for bode plots - this works but is hard limited by active filtering in the Mac at 20KHz. I use a python script to simulate the siglent AWG where the scope tells the script the frequency required for the bode plot. This worked surprisingly well!
I think I've hit the limit as I'd need a sig get for digital clocking prototyping (24.756MHz), DSD prototypes and I want a 455KHz (IIRC) for checking the tuning for my Phillips tube radio - the Silent SDG1032X would tick the boxes and it can be hacked to 60MHz. The key issue with most of modern AWGs is that the inbuilt wave forms work at values limited by the internals - for example sine at 60MHz, square at 60Mhz but triangle less and memory based AWG waveforms substantially less (they don't publicise that). However I think it has all the waveforms I want in the hardcoded set (sine, triangle, square, AM & FM modulated) and it's supposedly got decent square wave transitions compared to the Rigol and older models.
There's plenty of synth generators out there, cheap and easy. There's also plenty of discrete schematics starting with crystals and maintaining a non-digital approach - hell if you can do it with a tube and a crystal there's got to be some great kit out there waiting to be loved again!
I think I've hit the limit as I'd need a sig get for digital clocking prototyping (24.756MHz), DSD prototypes and I want a 455KHz (IIRC) for checking the tuning for my Phillips tube radio - the Silent SDG1032X would tick the boxes and it can be hacked to 60MHz. The key issue with most of modern AWGs is that the inbuilt wave forms work at values limited by the internals - for example sine at 60MHz, square at 60Mhz but triangle less and memory based AWG waveforms substantially less (they don't publicise that). However I think it has all the waveforms I want in the hardcoded set (sine, triangle, square, AM & FM modulated) and it's supposedly got decent square wave transitions compared to the Rigol and older models.
There's plenty of synth generators out there, cheap and easy. There's also plenty of discrete schematics starting with crystals and maintaining a non-digital approach - hell if you can do it with a tube and a crystal there's got to be some great kit out there waiting to be loved again!
How are the square waves being evaluated? With a scope? Does the scope probe ground wire ring? Does the generator including any test cable need to be terminated in its characteristic impedance to control ringing, etc.?
The thing about square wave testing is that its good to have the rise time fast enough to adequately test an audio device (along with a nice flat top of course), but not so fast that that you have to work too hard to control ringing in the test equipment. Some of the DDS arbitrary waveform generators on ebay are probably faster rise-time than you might need or want.
The thing about square wave testing is that its good to have the rise time fast enough to adequately test an audio device (along with a nice flat top of course), but not so fast that that you have to work too hard to control ringing in the test equipment. Some of the DDS arbitrary waveform generators on ebay are probably faster rise-time than you might need or want.
Last edited:
How are the square waves being evaluated? With a scope? Does the scope probe ground wire ring? Does the generator including any test cable need to be terminated in its characteristic impedance to control ringing, etc.?
The thing about square wave testing is that its good to have the rise time fast enough to adequately test an audio device (along with a nice flat top of course), but not so fast that that you have to work too hard to control ringing in the test equipment. Some of the DDS arbitrary waveform generators on ebay are probably faster rise-time than you might need or want.
Agree. That is one of the very nice things about using one of the older HP function generators - the rise and fall times, as well as the pulse duration and offset, are all adjustable, The sine and other waveforms generated are also very nice to have in one unit. Makes it very easy to check for oscillations, frequency response and overall stability of the amplifier under varying input conditions.
Hal
All digital signal generators are arbitrary waveform sources. So, precise rise/fall time adjustment is available. I’m very satisfied with this one:How are the square waves being evaluated? With a scope? Does the scope probe ground wire ring? Does the generator including any test cable need to be terminated in its characteristic impedance to control ringing, etc.?
The thing about square wave testing is that its good to have the rise time fast enough to adequately test an audio device (along with a nice flat top of course), but not so fast that that you have to work too hard to control ringing in the test equipment. Some of the DDS arbitrary waveform generators on ebay are probably faster rise-time than you might need or want.
UNI-T UTG962E
I suppose it depends on what rise times might be of interest. If they aren't too fast the generator's stair-step output should smooth out reasonably well. The one linked to appears to use a 5ns clock period?...precise rise/fall time adjustment is available.
Output is smoothed well. This is actual fastest rise time of about 50 ns. No stair-steps. 🙂
Anyway, anything within analog audio is glacially slow compared to this.
Anyway, anything within analog audio is glacially slow compared to this.
Hmmm... Max generator frequency is given as 60MHz. Sample rate is given as 200M samples/sec (which would seem to imply a 5ns generator clock period).
The period of a 60MHz wave is 16.666ns/cycle. That would seem to imply a 60MHz synthesized waveform would consist of roughly 3 samples?
To make things simpler how about let's say maximum generator frequency was given as 50MHz? That would seem to imply 4 samples per cycle. A sine wave though? Maybe approximately (+1,0, -1,0), but would it still work as well if we asked for 45Mhz?
At least it seems like a 50MHz square wave could plausibly make sense. Am I missing something here? Maybe I'm having a bad day...
The period of a 60MHz wave is 16.666ns/cycle. That would seem to imply a 60MHz synthesized waveform would consist of roughly 3 samples?
To make things simpler how about let's say maximum generator frequency was given as 50MHz? That would seem to imply 4 samples per cycle. A sine wave though? Maybe approximately (+1,0, -1,0), but would it still work as well if we asked for 45Mhz?
At least it seems like a 50MHz square wave could plausibly make sense. Am I missing something here? Maybe I'm having a bad day...
Last edited:
Much like PRR, I've often used a makeshift Schmitt trigger oscillator, preferably with 74HC gates running at 5 V to get a fairly short rise time. You typically get ringing due to the probe ground lead, but you can usually get around that by using a short ground connection instead of the normal ground lead, or just ignore any ringing that clearly changes frequency when you bend the probe ground wire.
Hmmm... Max generator frequency is given as 60MHz. Sample rate is given as 200M samples/sec (which would seem to imply a 5ns generator clock period).
The period of a 60MHz wave is 16.666ns/cycle. That would seem to imply a 60MHz synthesized waveform would consist of roughly 3 samples?
To make things simpler how about let's say maximum generator frequency was given as 50MHz? That would seem to imply 4 samples per cycle. A sine wave though? Maybe approximately (+1,0, -1,0), but would it still work as well if we asked for 45Mhz?
At least it seems like a 50MHz square wave could plausibly make sense. Am I missing something here? Maybe I'm having a bad day...
Declared max. frequency of those generators is for the sine wave only. Square wave frequency is limited on the control interface to 20 MHz. Though, I find acceptable square wave limit at about 6 MHz, where square wave definition is still good. At 20 MHz, it becomes practically a sine wave.
As one sine wave cycle can be generated from 2 samples, 60 MHz signal is perfectly clean.
I know this is diverging from the OP's post but from the SDG1062X user manual:
So you pay the piper - as many of the cheaper alternatives don't provide any characteristics and the marketing department slaps a 30Mhz on it even if that is barely a sine wave.
It also lists square wave maximum of 60MHz, althoughBenefitting from a special square-wave generating circuitry, the Square from the SDG1000X breaks the 60 MHz bandwidth barrier, reaching rise/fall times of less than 4.2 ns, and frequencies up to 60 MHz.
So you pay the piper - as many of the cheaper alternatives don't provide any characteristics and the marketing department slaps a 30Mhz on it even if that is barely a sine wave.
- Home
- Design & Build
- Equipment & Tools
- Looking For a Good Function Generator to Produce Square waves