IIRC just slightly more than 2 samples per cycle, but doing so requires a reconstruction filter. A zero-order hold dac output by itself produces stair steps. Maybe the generator has a filter though.As one sine wave cycle can be generated from 2 samples...
Also wondering if I can ask about the scope bandwidth, 100Mhz?
Last edited:
Scope bandwidth is 200 MHz, but probes are the real limiting factor here. Ringing present on the square wave is caused by probe. Connecting generator to the scope with short BNC cable, results with waveform without ringing.
The 3312A and newer HP/Agilent/Keysight/Whatever-name-this-week 33120A are excellent too. If you buy used, buy one that's calibrated or at least thoroughly tested.I have an HP 3311A function generator that does pretty well. Only catch is it has a manual DC zero adjust.
Tom
A 200MHz scope can display a sine wave of that frequency although the amplitude will be attenuated. Since a square wave consists of a fundamental and odd-harmonics the first odd harmonic will be at three times the fundamental, the next harmonic at five times, etc. For a 60MHz square wave the first odd harmonic would be at 180MHz. That harmonic would be attenuated somewhat on a 200MHz scope. Probably the next odd harmonic at 300Mhz would be more or less missing. The waveform might appear as a little bit distorted sine wave. Could be that the 3rd harmonic would be a little phase shifted relative to the fundamental as well. There is some discussion about bandwidth on page 8 of the document at: https://download.tek.com/document/02_ABCs-of-Probes-Primer.pdf Interesting that Tek keeps revising the linked document. Didn't notice anything about spring grounds in this version.
Last edited:
These work well.
https://www.digikey.com/en/products...LrFEWsOzJeIlUMyPoS4VTlkCdI-GEgNhoClDYQAvD_BwE
Also there are pcb sockets that a probe (less clip head and ground lead) can plug directly into.
https://www.digikey.com/en/products...LrFEWsOzJeIlUMyPoS4VTlkCdI-GEgNhoClDYQAvD_BwE
Also there are pcb sockets that a probe (less clip head and ground lead) can plug directly into.
Regarding the probe spring grounds at digikey, they fit a certain diameter ground ring. Ordered a set to try but they don't fit any of my probes. Got a silver plated beryllium-copper coil spring of about the right diameter and made my own out of that for one probe. Some of the other probes came with their own spring grounds, but low-cost throwaway probes never seem to.
Seems that a relatively simple request of post #1 has turned into a complicated and extremely deep subject with the nitpickers chomping and chewing, discussing unrelated frequencies in the RF ranges to death.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.
Thank you!
Alas, it's been the usual case of some threads, eh?
Here is what I have been using.
https://www.amazon.com/Tenma-72-505...mzn1.fos.304cacc1-b508-45fb-a37f-a2c47c48c32f
https://www.amazon.com/Tenma-72-505...mzn1.fos.304cacc1-b508-45fb-a37f-a2c47c48c32f
Yes, it did get quite complicated. All in good conversation. They are free do that now that I already decided to buy the $115 Koolertron function generator. It is inexpensive, has a decent interface and looks like it should handle most anything I would ever want to do very well. I think the highest frequency that people test with on Class A amps (rarely) is only around 200khz. I bought the 15mhz version. I should get it today. I will post my impressopns here once I play around with it.
I am going to use it to set the oscillation cap value on a new Stasis front end that I put on my Threshold S/500. I ordered a few different value caps from Mouser as well.
I am going to use it to set the oscillation cap value on a new Stasis front end that I put on my Threshold S/500. I ordered a few different value caps from Mouser as well.
Uh, once built a simple mosfet buffer. Initial testing showed it didn't even start rolling off until up around 80MHz. Sounded great though. Just shows that individual circuit areas can operate and or potentially be unstable at very high RF frequencies even though an overall amplifier may roll off at 200kHz - 300kHz. Think about it: If every amplifier stage rolled off at such low frequencies we couldn't have a dominant closed-loop pole, instead we would have an oscillator. So just because its an audio device doesn't mean we don't need to check at much higher frequencies.Seems that a relatively simple request of post #1 has turned into a complicated and extremely deep subject with the nitpickers chomping and chewing, discussing unrelated frequencies in the RF ranges to death.
Last edited:
My Tenma 72-455A, the bigger model, suits me just fine and is entirely reliable for audio work.
They make decent stuff.
I also have a Tenma O'scope that plays nice on my service bench.
eBay has a lot of different function generator options. The $20 DDS ones would probably be good enough for your tasks. https://www.ebay.com/itm/3137412379...420+vdt4WumLiosHfraaVgHUc=|tkp:Bk9SR6zDzbbqYA Its unlikely your amp's risetime will be within 10X of what they can do. Then going through the options-
The DDS based China generators like the Uni-T seem pretty good. There is a lot of discussion about them on EEVblog https://www.eevblog.com/forum/testg...rbitrary-waveform-generator-220394/?topicseen The Siglents are probably a better choice especially if you plan to get a Siglent scope. The traditional generators from Wavetek, HP and Krohn-Hite are analog based and do work pretty well. The KH's can be adjusted to .1% THD, something I never accomplished with HP or Wavetek. But those are all old and may need service.
The DDS based China generators like the Uni-T seem pretty good. There is a lot of discussion about them on EEVblog https://www.eevblog.com/forum/testg...rbitrary-waveform-generator-220394/?topicseen The Siglents are probably a better choice especially if you plan to get a Siglent scope. The traditional generators from Wavetek, HP and Krohn-Hite are analog based and do work pretty well. The KH's can be adjusted to .1% THD, something I never accomplished with HP or Wavetek. But those are all old and may need service.
I like sweeping up to at least 30 MHz on circuits to check things stay stable even if provoked, so IMO wanting that sort of frequency from a function generator isn't out of order for audio work. Next I'll try Stupid Modulation Tricks to see whether output could be affected by RFI, as TI mentions here.Uh, once built a simple mosfet buffer. Initial testing showed it didn't even start rolling off until up around 80MHz. Sounded great though. Just shows that individual circuit areas can operate and or potentially be unstable at very high RF frequencies even though an overall amplifier may roll off at 200kHz - 300kHz. Think about it: If every amplifier stage rolled off at such low frequencies we couldn't have a dominant closed-loop pole, instead we would have an oscillator. So just because its an audio device doesn't mean we don't need to check at much higher frequencies.
A more complete guide for testing opamp EMI/RFI immunity: https://www.ti.com/lit/an/snoa497b/snoa497b.pdf?ts=1663567984979&ref_url=https%3A%2F%2Fwww.ti.com%2Flit%2Fpdf%2Fsnoa497
Also, attached below is a related article from a different author.
Also, attached below is a related article from a different author.
Attachments
Last edited:
Most useful resources, thanks! All the inputs on my stuff are already balanced and lowpass filtered at 160 kHz to keep out various RF nasties, although some of the other techniques in the Cherry Clough document are news.
Even speaker cables can send crud back to the power amp. A quad run of fat RG-8 sounded better than twin lead at a friend's downtown place: the audio seemed to have less highs, but switching back to the twin lead showed its extra treble was mostly grunge. For lack of a better term, the sound of the coax had "blacker backgrounds" as audiophile reviewers might say. We speculated that RF energy was getting in the power amp via the feedback send to the input differential pair.
Especially in cities RF can be an annoyance: Wendy Carlos found enclosing her Manhattan studio in a Faraday cage made gear which had been too noisy for consideration now fit for purpose. This was back in the early 1980s, mind, before cell phones and wifi and satellite TV and Starlink and all the other "contributions" gracing the electromagnetic landscape, so imagine how much noise eagerly awaits entry into your audio gear. These days design for EMI resistance is a necessity, not a luxury.
Mods: maybe a new thread is needed?
Even speaker cables can send crud back to the power amp. A quad run of fat RG-8 sounded better than twin lead at a friend's downtown place: the audio seemed to have less highs, but switching back to the twin lead showed its extra treble was mostly grunge. For lack of a better term, the sound of the coax had "blacker backgrounds" as audiophile reviewers might say. We speculated that RF energy was getting in the power amp via the feedback send to the input differential pair.
Especially in cities RF can be an annoyance: Wendy Carlos found enclosing her Manhattan studio in a Faraday cage made gear which had been too noisy for consideration now fit for purpose. This was back in the early 1980s, mind, before cell phones and wifi and satellite TV and Starlink and all the other "contributions" gracing the electromagnetic landscape, so imagine how much noise eagerly awaits entry into your audio gear. These days design for EMI resistance is a necessity, not a luxury.
Mods: maybe a new thread is needed?
Regarding speaker cable, if you haven't seen it before an interesting document attached below. Again, RF type considerations are involved.We speculated that RF energy was getting in the power amp via the feedback send to the input differential pair.
Also some published work on line level cables and headphone cables. The latter article is behind a paywall: https://www.aes.org/e-lib/browse.cfm?elib=20555
Attachments
Last edited:
Once again a timely article, since this week I'm replacing the cable on my HD-800 headphones with a pair of these:Regarding speaker cable, if you haven't seen it before an interesting document attached below. Again, RF type considerations are involved.
Also some published work on line level cables and headphone cables. The latter article is behind a paywall: https://www.aes.org/e-lib/browse.cfm?elib=20555
https://shop.sommercable.com/en/Cab...-Mikrofonkabel-SC-Cicada-SO-D14-200-0451.html
The capacitance change noted in the first article for Belden 8412 and 8422 might (really speculating here) be reflected in this listening test:
Although 8412 described as the "new classic" is a laugh -- I was making mic cables with that stuff in 1979!After some fiddling I found: 1) the 8422 and the 8412 share the same basic sonic imprint, but the 8422 sounds lighter on its feet and has somewhat smaller images than the 8412; 2) Using 2 pairs of 8422 the sound may need a little more weight, just as using 2 pairs of 8412 the sound needs some more litheness.; 4) the better balance was obtained by using the 8412 at the front and the 8422 from preamp to amp.
Speaker cable capacitance causing amplifier instability might be a non-issue with class D amplifiers since their lowpass filters have capacitors greater by a couple orders of magnitude than even pathological cables might present, for example 0.68 uF in the UcD400 vs 500-1500 pf/foot, and at 1 MHz the impedance of a 0.68 uF cap is about a quarter ohm so any RF reflections would get swallowed up pronto.
I wanted to share that I received the little Koolertron GH-CJDS66-A today. Hopefully within the next couple days I'll have some time to play around with it. My initial impressions are that it is very lightweight, and there's not very much going on inside of it. However, the screen seems to be easy to read, the interface seems nice and the buttons feel good.
I will post my impressions hopefully within the next couple days after getting a chance to use it. Thank you for all of your suggestions. Don't mind the dirty speaker and the gutted wall. I am in the middle of a remodel.
I will post my impressions hopefully within the next couple days after getting a chance to use it. Thank you for all of your suggestions. Don't mind the dirty speaker and the gutted wall. I am in the middle of a remodel.
Found another one-buck square-box with symmetry and with DC coupling.
The R & C values are same as explained above or online.
The missing pin number is 12.
I just noticed the publication date; I don't think this is a joke.
The R & C values are same as explained above or online.
The missing pin number is 12.
I just noticed the publication date; I don't think this is a joke.
Don't try to listen to square waves.....I wanted to share that I received the little Koolertron GH-CJDS66-A today. Hopefully within the next couple days I'll have some time to play around with it. My initial impressions are that it is very lightweight, and there's not very much going on inside of it. However, the screen seems to be easy to read, the interface seems nice and the buttons feel good.
I will post my impressions hopefully within the next couple days after getting a chance to use it. Thank you for all of your suggestions. Don't mind the dirty speaker and the gutted wall. I am in the middle of a remodel.
View attachment 1092513
- Home
- Design & Build
- Equipment & Tools
- Looking For a Good Function Generator to Produce Square waves