Scope image interpretation

[nsubulysses]

I was going through some old pics and came across this scope image. It is the output of a Traynor YBA1-1A Bassmaster Mark II. Doing about 70V pk-pk right before clipping into 8 ohms.

There is a slight crossover notch developing as well as some sort of offset where the negative signal reaches down farther than the positve goes up. Also seems like the negative side starts to thin out on the horizontal scale while the top is a little rounder and wider. Unfortunately since this was a "favor" repair for a friend I don't have any documentation on the bias or plate voltage or anything. I would guess I biased it around 60% and the plate voltage was high from what I remember -- probably around 560V. Although the sine wave was sllightly distorted I let it go after poking around for a bit and not finding anything. I wasn't gonna spend hours since it was a free repair and the amp sounded good. Now in retrospect i'm wondering, what does this image mean? Any thoughts?

Does this have something to do with the shared screen grid resistor, or that C7 and C8 are different values, causing a different frequency response for the + and - section of the waveform?

[g1]

If I saw this on a unit I was testing, the first thing I would suspect is that the power tube doing the top-side duty is weak. Then I'd swap the power tubes and see if it changed to the bottom side clipping earlier than the top.

[Tom Phillips]

You said the amp is putting out 70Vpp. That's 76 Watts into 8 Ohms.
Therefore, I think it's likely that a waveform showing those distortion characteristics is as expected from a Traynor YBA1-1A putting out that much power.
Something to consider anyway.
Cheers,
Tom

[Chuck H]

Forgive my lack of formula awareness... According to this chart I think that's 150W!?! In which case I would start by noting a possible error in the voltage division adjustment. Corrections on how to interpret this are accepted.

[Tom Phillips]

70Vpp = 24.8 Vrms That's the key Chuck.
Square 24.8 and divide by 8 and you come up with 76.6 Watts.

[g1]

Chuck, I think you took the 70Vp-p and halved it to 35V peak but didn't multiply that by .707 to get the RMS value.

[Chuck H]

Thank you gents. I guess I always figured the chart for accurate since it's resistances are given in common speaker load values. I see two things happening here. One is the possible explanation for some outrageous output power claims by both Traynor and proponents. The other is that they seem to be damn impressive just the same.

[nsubulysses]

Chuck, I think you took the 70Vp-p and halved it to 35V peak but didn't multiply that by .707 to get the RMS value.

Cool this thread got sidetracked and I now know this method too

My method is just taking pk-pk and multiplying by .3535 to get RMS. Then as Tom said RMS^2/load = Watts. I think it said this on allaboutcircuits.com

[Enzo]

NSU, first, this is a guitar amp, not a hifi. There are variations all through the amp, it is not designed for precision signal reproduction. That the push and the pull are not exactly the same is no surprise. The crossover distortion looks normal. You might be able to tune it out, but in my experience setting up guitar amps for no crossover distortion usually results in a poor sound.

To find out WHY the two sides are different, you'd have to scope the signal coming into each power tube, the signal coming into the phase inverter, and for that matter all through the amp. The signal waveform could already be asymmetrical back at V2 or even V1 and the power stage is just faithfully reproducing that.

The phase inverter could be causing it, any stage could, if nothing else, the two halves of the tube could have different gains. And so on.

C7 and C8 are different because they do different jobs. C7 is the input to the PI, it feeds the grid. C8 is a large value cap whose job is essentially to ground that other grid for AC signal. read up on the operation of an LTP phase inverter stage.

[J M Fahey]

I was going through some old pics and came across this scope image. It is the output of a Traynor YBA1-1A Bassmaster Mark II. Doing about 70V pk-pk right before clipping into 8 ohms.

There is a slight crossover notch developing as well as some sort of offset where the negative signal reaches down farther than the positve goes up. Also seems like the negative side starts to thin out on the horizontal scale while the top is a little rounder and wider. Unfortunately since this was a "favor" repair for a friend I don't have any documentation on the bias or plate voltage or anything. I would guess I biased it around 60% and the plate voltage was high from what I remember -- probably around 560V. Although the sine wave was sllightly distorted I let it go after poking around for a bit and not finding anything. I wasn't gonna spend hours on it since it was a free repair and the amp sounded good. Now in retrospect i'm wondering, what does this image mean? Any thoughts?

Does this have something to do with the shared screen grid resistor, or that C7 and C8 are different values, causing a different frequency response for the + and - section of the waveform?

In 2 words? :
* TUBE
* AMPLIFIER


Nothing wrong, nothing to correct, it's in the nature of the beast.

In fact, personally I prefer it less overbiased and with more of the improperly called "crossover distortion"

As of the apparent assimetry:

To find out WHY the two sides are different, you'd have to scope the signal coming into each power tube, the signal coming into the phase inverter, and for that matter all through the amp. The signal waveform could already be asymmetrical back at V2 or even V1 and the power stage is just faithfully reproducing that.

The phase inverter could be causing it, any stage could, if nothing else, the two halves of the tube could have different gains. And so on.

You want perfect sinewaves?
Go SS ... period.

EDIT: explaining by analogy:

what does this image mean?

Hey!!!
It's burnt !!!!
It actually has carbon in it !!!!
It's raw !!!!!
It oozes blood !!!!!!
It's unevenly, poorly cooked !!!!!!

So you want even, perfectly cooked meat, uniform colour, no visible blood?
HEALTHY????
Fine, BOIL it:



Pick one

[Chuck H]

So let's say it can only produce 70W clean instead of 78W! Still a feat for any standard two bottle guitar amp!

[km6xz]

The wave form is not out of the ordinary but has some information to tell: It has a pretty good dose of even harmonics, maybe 12-15% second which is not normal for a push pull amplifier with lots of negative feedback, non-reactive load and optimized for linearity........which is NOT what a guitar amplifier is designed for. The crossover notch is also harmonic distortion, as it the flattopping. If you took a perfectly linear amplifier operating well under maximum power and fed it a sinewave, it would look like a math sine function, then add a bit of twice that frequency, and a small amount of 3 times that base frequency and repeat with lower and lower levels all the way up to 10khz or so, the waveform would look pretty much like what you are seeing. You are seeing harmonic distortion, the kind that can sound "musical" but you are also seeing Intermodulation Distortion which is anything but musical. Intermod is something we try to design out of an amplifier but what causes the desired harmonic distortion also creates the horrible intermodulation distortion. The main culprit is non-linearity, which you see signs of near the zero crossing area and after the compression knee at the top. That non-linearity of a stage causes two or more fundamentals mixing together to create sidebands. Side bands are additional signals which are the sum and difference of the two original signals. But when a complex signal with many many harmonics goes through that non-linearity, every one of those artifacts created sums and differences additional signals. The result is not pleasant for hi-fi or instrument amps because the sound is a sea of additional signals not in the original. You only want the musical sounding distortion, but the other junk makes the note less distinct in a sea of sound energy, where a flute starts sounding like a guitar or become ambiguous in location in the listening room, and clarity of notes. So a great sounding guitar amp does not harm the original, just adds a mix of even and odd harmonics. The waveform you posted will not sound good with complex rich content but could be good for overdrive sounds. Harmonic distortion tracks signal level in our perception of it, weak softer sounds will have weak softer harmonics. IM distortion is very easily heard at high levels but worse at very low levels, save down around the level where the dominate distortion process is the cross-over notch. A signal that barely exceeds the amplitude of the notch, will be horrible sounding. The trick in getting an amp to get dirty when pushed but clean when backing off, "touch" is to minimize IM distortion everywhere except the high levels. So crossover notch should not be present at all at low levels and can get more and more apparent as the amp supply and tubes are getting pushed. If you back off the level to say 50% of that amplitude, make sure no notch is seen but allowing it to come in as compression starts is a good target for an amp that is going to be played clean, really clean and then pushed into dirt when bearing down on it.
Waveforms tell a lot but they have to be looked at using various levels and signal types. Two concurrent non-harmonically related signals fed at the same time will reveal the extent of IM distortion you have.
The most versatile amps seems to be those that feature power amp compression when pushed, that creates a strong odd order harmonic content being driving by preamp circuits that start to become non-linear at about the onset of compression in the power amp. That non-linearity in the preamp, which is almost always class A single ended, generates even harmonics that the push-pull output does not. It is a balancing act of levels and stage gain that really makes one amp with the same general topography sound sweet and another that sounds like a dog. Most builders don't seem to understand the basics of electronics for sound so think magic parts or tricks are the answer. They never are.

[nsubulysses]

These responses were everything I was hoping for and more. Thanks everyone.

JM, is this the the boiled meat you speak of



Anyone care to comment on this waveform too?

Searching for more info on these things I found "Harmonic Distortion and Negative Feedback in Audio-Frequency Amplifiers" on tubebooks.org - http://www.tubebooks.org/Books/bbc_feedback.pdf
If any other good resources on this come to mind please let me know.

[Jazz P Bass]

That wave form appears to be symetrically clipping.
Both the top & the bottom are flattening out at the same moment & the width of the 'flat part' looks the same.
(one of the key things to look at, the width of the peaks)

Further searches:
https://www.google.com/search?q=harm...JYeyJN49utA%3D

https://www.google.com/?gws_rd=ssl#q...ipped+waveform

[km6xz]

That one looks well balanced and having a stiff high filtered power supply, even though it is biased hot. It looks more like a hi-fi amp and have very low even harmonics.
You can see a lot more information if you can view frequency domain as well as time domain. The scope is the tool for time domain but a spectrum analyzer is the tool for frequency domain, so you can see the harmonics and their absolute amplitude.
If you have a sound card here is a software spectrum analyzer that works great:
Visual Analyzer at Download
It has a scope, function generator, meters and a nifty spectrum analyzer. If someone build their own probe interface with attenuator to keep input levels of the sound card down to within its limited range, this and a laptop is a complete test setup for bench or field work. I have HP spectrum analyzers and other lab style gear but this works great also on the bench