Hi all,
I just got my first 'scope, free from an uncle. It's a really old one, a "Scopex 4D 10B", quick search seems to suggest it's 10MHz. I'm not sure how good it is and all that, but I can't afford to buy one really so I will have to make do! It seems to all work ok, though I think the probe connectors might need cleaning or something. They need a bit of a wiggling about to get a stable image
I was just wondering if anyone would be so kind as to point me in the direction of 'scope pictures of any sorts of oscillation, so I know what I am supposed to be looking for.
I've only been brave enough to play a 30Hz tone, with a spare set of speakers connected (probably not wise, but it worked...), and connected the ground of the probe to the speaker terminal -ve, then probed the speaker terminal +ve. All I can say is that I seemed to get a very clean sine wave show up.
I presume that if it were oscillating it would stand out and wouldn't be a clean wave? I presume that I should be testing a whole load of frequencies though, and possibly not just sine waves (I fear I may be quite limited here, as I am using a computers sound card for the tone generation, I measured a 1KHz square wave straight from the sound card output and it had some 'ringing' and 'overshoot').
Any good links for a newbie would be much appreciated. I tried searching for ages for any pictures of an actual scope showing some form of oscillation, but couldn't really find anything!
I just got my first 'scope, free from an uncle. It's a really old one, a "Scopex 4D 10B", quick search seems to suggest it's 10MHz. I'm not sure how good it is and all that, but I can't afford to buy one really so I will have to make do! It seems to all work ok, though I think the probe connectors might need cleaning or something. They need a bit of a wiggling about to get a stable image
I was just wondering if anyone would be so kind as to point me in the direction of 'scope pictures of any sorts of oscillation, so I know what I am supposed to be looking for.
I've only been brave enough to play a 30Hz tone, with a spare set of speakers connected (probably not wise, but it worked...), and connected the ground of the probe to the speaker terminal -ve, then probed the speaker terminal +ve. All I can say is that I seemed to get a very clean sine wave show up.
I presume that if it were oscillating it would stand out and wouldn't be a clean wave? I presume that I should be testing a whole load of frequencies though, and possibly not just sine waves (I fear I may be quite limited here, as I am using a computers sound card for the tone generation, I measured a 1KHz square wave straight from the sound card output and it had some 'ringing' and 'overshoot').
Any good links for a newbie would be much appreciated. I tried searching for ages for any pictures of an actual scope showing some form of oscillation, but couldn't really find anything!
If you get your signal from the soundcard you will see no oscillations. Speakers cannot oscillate. Oscillations need some active circuitry like amplifiers.
Oscillations occur if the output from an amp gets back to its input in such a phase that it reinforces the input signal; the output gets higher and higher until it hits the supply limit, then it starts to fall until it hits the lower supply limit etc. You always need to feed in energy to sustain an oscillation.
Does that help?
Jan Didden
Oscillations occur if the output from an amp gets back to its input in such a phase that it reinforces the input signal; the output gets higher and higher until it hits the supply limit, then it starts to fall until it hits the lower supply limit etc. You always need to feed in energy to sustain an oscillation.
Does that help?
Jan Didden
janneman said:
Oh, sorry. I probably badly phrased something (or don't understand what you're telling me yet!).
The 30Hz tone I measured from my newly built GainClone amplifier outputs, basically (at the speaker terminals, with spare speakers connected). I was using a sound card here as a tone generator to feed the amplifier.
The square wave I measured from the soundcard output just experimenting
But basically, from what you're saying, it sounds like oscillation should be really quite obvious....?
A sine or square wave IS oscillation. It sounds like you are interested in seeing whether your amp is oscillating due to instability.
If your amp is oscillating at a very high frequency, 10s - 100s of MHz, you probably won't see it on a 10 MHz BW scope. If it oscillates at lower frequencies, you should be able to see it clearly, even without injecting a signal from your sound card. Try shorting the input of the amplifier and look at the output with the scope. If the amp is working properly and your grounding is done well, you should see just a tiny bit of noise on the amplifier output. If the amp is oscillating, you will probably see some large amplitude waveform like a sine or square or something more complex. You may see some evidence of poor grounding in the form of a 60 and / or 120 Hz wave of probably complex shape. That is what makes a buzz or hum in the speaker.
I_F
If your amp is oscillating at a very high frequency, 10s - 100s of MHz, you probably won't see it on a 10 MHz BW scope. If it oscillates at lower frequencies, you should be able to see it clearly, even without injecting a signal from your sound card. Try shorting the input of the amplifier and look at the output with the scope. If the amp is working properly and your grounding is done well, you should see just a tiny bit of noise on the amplifier output. If the amp is oscillating, you will probably see some large amplitude waveform like a sine or square or something more complex. You may see some evidence of poor grounding in the form of a 60 and / or 120 Hz wave of probably complex shape. That is what makes a buzz or hum in the speaker.
I_F
I_Forgot said:
Ah yes, sorry. This is what I meant!
I saw someone state that a 100MHz scope was good for up to about 10MHz, so is what I have probably roughly good for about 100KHz?
I've also read that such high frequency oscillation can destroy tweeters apparently. After how long might you expect this to happen if I were to play at a fairly loud volume constantly?
Oh I see! Well I believe my grounding is ok. AndrewT was very helpful there and I can't hear anything at all coming from the speakers with the inputs shorted
I'll give that a go and see if there's anything to see there anyhow, thanks very much!
By this reasoning, your scope would be good up to 1MHz. However, you should be able to see 100MHz on your scope, it just might appear a little shorter on the screen than it actually is.markiemrboo said:
If your amp is prone to oscillation, it may need a push to get it started. An oscillation is the repeated amplification of something and if there was no something to begin with, an unstable amp may sit there like a stone. Often the noise that all amps have is enough to trip an oscillation, sometimes a turn on thump and sometimes it just takes a while to warm up. Sometimes touching the inputs will do it and leaving them open may help in this respect.
An oscillation usually happens at one specific frequency. It is usually higher than 20kHz. It usually causes full scale deflection so that you can either see it cover the oscilloscope screen from top to bottom, or you can see the amp's upper and lower limits on the screen usually in the form of a slightly brighter region at the top and at the bottom.
An oscillating amp will destroy a tweeter in a matter of seconds or even less. Crossover networks to tweeters have high-pass response (that's why it works as a crossover) so it will easily pass the high frequency oscillation, including one that may be well above audible range. Tweeters are made with thin wire coils to keep the mass down so they can respond to high frequencies. Music doesn't normally require a lot of power in the highest registers, so a tweeter in a speaker system rated for 300 Wrms may only be rated for 1 or 2 watts. An oscillating amp can easily deliver quite a bit more than the 1 or 2 watts required to cook the tweeter.
I_F
I_F
Apart from the useful info here, have a look at this scope tutorial.
https://www.cs.tcd.ie/courses/baict/bac/jf/labs/scope/
/Hugo
https://www.cs.tcd.ie/courses/baict/bac/jf/labs/scope/
/Hugo
lndm said:
Well spotted
Netlist said:
Thanks very much!
I do have the oppertunity(sp) to buy myself a Tektronix 475A second hand from ebay (collection only and lives about an hour away from me). Do you think it would be worth me buying one? I hear the Tektronix are very, very good, but I am not sure how often it would be used or even if I really do need it as it sounds like I am not suffering from any very high frequency oscillation as my tweeters are not on fire yet.. I can't decide whether it would be worth it, to a newbie this one seems nearly adequate... but I am sure a higher bandwidth one would pick up more detail?!
lndm said:
A 1 MHz BW scope will limit your thinking as well as your measurements. You will wonder about something that you would be able to check with a decent scope, but then you say to yourself, "I won't be able to see it with my crappy 1 MHz BW scope", so you will dismiss the possibility of ever finding out what you were wondering about.
A wide BW scope is an "enabling" technology. If you have a 200 MHz BW scope, you will find yourself doing things beyond audio simply because they are much easier when you can see what is going on. If you can get a deal on the Tek, go for it.
I_F
I_Forgot said:
Is there any way that it wouldn't be putting full power in to the speakers, or oscillation wouldn't end up with me holding a pair of dead tweeters? I have a Zobel network on the speaker outputs, but no Thiel(sp?). The Zobel is 2.7R and 100nF.
I used a X1 (EDIT: sorry, 1X!) probe on unloaded speaker outputs (as in, no speaker or power resistor connected), and disconnected but *not* shorted (yet) inputs. Set to 10mV/cm and 1uS/cm I was seeing two really small and rather fat sine waves. The first repeated about 4 times, the second maybe 10 times. The amplitude of both were probably around 1cm.
I'm really not very good at maths, at all (I used to be, until I got to College, then it all just seemed to go out of the window!), but in order to get the frequency of this wave would that just be, for example:
There are 8x1cm blocks horizontal across the screen of my scope, so presumably the total time period across the whole screen 8us (set on 1us/cm)? If it repeats 4 times across the whole length of the screen, then this would mean the sine wave repeated 1 time every 2us, or 0.5 times per 1us? Converting us to seconds should then be multiplying by 1,000,000, so this would make the frequency of the wave 0.5x1000000 = 500,000Hz = 500KHz? Since it was ~1cm high this would make it 500KHz at 10mv?
I'm sure it's more complicated than this, isn't it? I can get a poor quality webcam shot (digital camera is off for RMA) of the scope screen if needed.
Curiously, this exists on both channels, and on one channel the amplitude is slightly (but noticably) higher. To put more of a twist on the situation, the channel with the smaller amplitude actually runs quite hot with no signal, while the channel with the slightly higher amplitude runs barely noticably warm. The heatsinks are slightly different though. The cool running one is an old PII heatsink, whereas the hotter running channel is a really large PII Xeon (I think) heatsink cut down a bit, so it's roughly the same size as the PII heatsink.
Is *this* oscillation, or just harmless electrical noise of some sort? I ask because, as far as oscillation has been described to me, it couldn't be oscillation. My tweeters would be long gone after 3 days of abuse at higher volumes?
If it were to limit my thinking then I should get into a different game. First I will wonder to myself, then I'll go to the library (or the internet these days) and learn (and succeed).I_Forgot [/i]If the amp is oscillating said:
Like a (hypothetical
) driver, that lives in a city. The roads are so well constructed you'd have to fall asleep in order to crash, and the speed limits are very low. This driver comes to the country where speed limits are more liberal, and hazards abound, and they end up in a ditch.I do like to use a modern scope, and I like to build crossovers by way of computer simulations. I'd say that technology may have improved my products and it has certainly improved my turnover times. As a hobby, it is the self improvement that gives me greater satisfaction.
<lndm gets off soapbox>
markiemrboo said:
I suppose if it were oscillating at a high enough frequency it might not put full power into your tweeter. The tweeter's own inductance will protect it from a very high frequency oscillation.
Count again. There are usually 10 divisions across the scope, but yes, your technique for estimating the frequency is correct.
If you are seeing a periodic (repeated) waveform it IS oscillation. If the chips are running hot with no signal, THAT is probably a sign of oscillation. Here's an idea- put a cheesy transistor radio near the amp, turn it on, then turn on the amp and see if you hear the radio squealing. Another test- put a resistor on the output of the amp (try 4-10 Ohms) and see if it gets hot with no signal input. Check for DC on the output of the amp while you're at it. Recheck your wiring to make sure both channels are wired correctly.
You are very lucky if your tweeters are not toast yet. Normally when you build an amp, you should test it into a resistor before you connect it to your relatively expensive speakers.
I_F
I_Forgot said:
Here we would be talking MHz?
Nope, definately only 8 on this one!
Well, what's strange to me is that only one is getting hot, yet they both display the same waveform. The one that runs ever so slightly warm ("normal") even has the higher amplitude.
Haven't got one of those sadly
Power resistor? 10R@10W ok? I've actually done this before and it didn't get hot at all, but I *will* try again.
Checked already. I added a cap to the feedback in series with, and after the resistor to ground (it's a LM3886 chip amp if I haven't mentioned, sorry). DC offset is -1mv on one channel and -3mv on the other. Without the cap one was -30mv and the other was -90mv if I remember rightly.
It has been said that these chips do like to oscillate with no input, which is why I wanted to try really. Just so I could see what it would look like. I should try with the input shorted too.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.