Reasons for hacking DSOs

[Fungus]

Is painting your 'scope a different color exactly equivalent to unlocking extra bandwidth and features from a legal and/or moral point of view?

[tggzzz]

You never look at square waves higher than 7 MHz or so?

What does frequency have to do with it? All square waves are infinite.

Square waves are infinite but the ability of your oscilloscope to display them is not.  That limitation is due not just to the frontend bandwidth, but also to things like the display itself.

The factor of 10 thing is a somewhat arbitrary cutoff.  It may be that you actually want to examine irregularities in the square wave that have a frequency characteristic greater than 10x of the primary frequency.  10x is just an easy frequency to still see in a square wave when you're viewing a few cycles of the wave on the scope, whilst 100x probably isn't (even if the scope has the bandwidth and sampling rate for that, the display's resolution probably prevents you from seeing frequencies that high as anything more than vertical lines until you zoom into the waveform).

The slower the base frequency of the wave, the more component frequencies you'll be able to see on the screen before you run into bandwidth limits.  Of course, you might run into display resolution limits first.

Let me guess, you come from an RF or audio background.

The frequency is completely irrelevant.

Consider, for example, a very realistic and practical digital waveform that you can find in virtually any digital circuit since the late 1970s: a waveform with 1MHz frequency (i.e. << your 7MHz) and a 1% duty cycle, i.e. a 10ns wide pulse.

In digital circuits the only relevant parameter is risetime and the ability to observe it. A 100MHz scope will enable you to see the 3.5ns transitions and the pulse top. A 50MHz scope won't.

[kcbrown]

Let me guess, you come from an RF or audio background.

Actually, I come from a software background. 

I'm actually relatively new at this.  In college, I started out going for a EE degree but switched to computer science because I found that I was spending all my free time messing around with computers.  Best decision I ever made, except that in hindsight, I think it may be easier to learn computer science stuff on the side than EE, so I often wonder if I would have been better off sticking with EE and pursuing computer interests on the side.  Dunno.

The frequency is completely irrelevant.

Consider, for example, a very realistic and practical digital waveform that you can find in virtually any digital circuit since the late 1970s: a waveform with 1MHz frequency (i.e. << your 7MHz) and a 1% duty cycle, i.e. a 10ns wide pulse.

In digital circuits the only relevant parameter is risetime and the ability to observe it. A 100MHz scope will enable you to see the 3.5ns transitions and the pulse top. A 50MHz scope won't.

Suppose your square wave is unstable.  If the instabilities have a base frequency characteristic greater than 50 MHz, will the 50 MHz scope show them?

Digital still operates in the analog domain -- it's just an abstraction based on a convention that specifies the analog characteristics that implement the abstraction.  If you're looking at a digital signal with a scope, isn't there a good chance that you're doing so because the digital abstraction has broken down and you're trying to figure out what's going on?

[tggzzz]

Let me guess, you come from an RF or audio background.

Actually, I come from a software background. 

I'm actually relatively new at this.  In college, I started out going for a EE degree but switched to computer science because I found that I was spending all my free time messing around with computers.  Best decision I ever made, except that in hindsight, I think it may be easier to learn computer science stuff on the side than EE, so I often wonder if I would have been better off sticking with EE and pursuing computer interests on the side.  Dunno.

That was the way I did it, and I agree with your assessment. Most software is glorified CRUD, and hence very very boring.

The frequency is completely irrelevant.

Consider, for example, a very realistic and practical digital waveform that you can find in virtually any digital circuit since the late 1970s: a waveform with 1MHz frequency (i.e. << your 7MHz) and a 1% duty cycle, i.e. a 10ns wide pulse.

In digital circuits the only relevant parameter is risetime and the ability to observe it. A 100MHz scope will enable you to see the 3.5ns transitions and the pulse top. A 50MHz scope won't.

Suppose your square wave is unstable.  If the instabilities have a base frequency characteristic greater than 50 MHz, will the 50 MHz scope show them?

A spectrum analyser would be the right tool to analyse such phase noise.

[kcbrown]

Suppose your square wave is unstable.  If the instabilities have a base frequency characteristic greater than 50 MHz, will the 50 MHz scope show them?

A spectrum analyser would be the right tool to analyse such phase noise.

Hmm...Perhaps so.  I guess it would depend on how reliable the instability is.  If it's an occasional glitch that you want to capture, then a scope's mask feature could well be just the thing to catch the instability in the act, but that only works if the scope can see the glitch in the first place, which is really what's in question here.

And the spectrum analyzer would be able to tell you that the instability is there, and perhaps to help you characterize its shape.  It wouldn't help you in the slightest to detect when the instability is occurring, at least, not in any way that I can think would be useful.   I'm actually not sure about that last.  If spectrum analyzers have the ability to perform a mask test sort of like how the oscilloscopes can, and are able to sample the relevant portion of the spectrum quickly enough, I suppose it might be able to trigger a scope to capture other signals that could be relevant to the diagnosis of the problem.  Are spectrum analyzers capable in that way?  I've never used one myself, so I'm rather in the dark about what they can do these days.

[kcbrown]

Actually, I come from a software background. 

I'm actually relatively new at this.  In college, I started out going for a EE degree but switched to computer science because I found that I was spending all my free time messing around with computers.  Best decision I ever made, except that in hindsight, I think it may be easier to learn computer science stuff on the side than EE, so I often wonder if I would have been better off sticking with EE and pursuing computer interests on the side.  Dunno.

That was the way I did it, and I agree with your assessment. Most software is glorified CRUD, and hence very very boring.

The problems that can arise from it being crud in the first place can be quite interesting, actually, but I think one has to be something of a generalist in order to derive any real satisfaction from that.   And I happen to be a generalist (I've done a fairly diverse set of things in the computing world), so I'm able to solve a reasonably wide range of problems.  That keeps things interesting for me.

I recently regained my interest in the hardware side of things, and nowadays is possibly a better time to pick that up than ever before, what with the easy availability of inexpensive test equipment (that such segways nicely into this thread is pure coincidence, I assure you.  ), forums such as this, etc.

[Fungus]

I would consider it extremely unlikely that it is illegal to modify the standard box (at home) to become the deluxe box. I would draw a parallel here with a 50MHz/100MHz scope that has a similar keycode or jumper link system. It would only be illegal if it was actually illegal for a consumer to 'own' a 100MHz scope in the UK.

Exactly.  And this is all because you're not violating any laws, particularly copyright laws.  There is no copyrightable content that you'd be accessing that you didn't have rightful access to before.

Arguing what today's written laws say is pointless.

If something becomes widespread enough to affect a large company's business model then laws will appear. Bet on it.

[vodka]

 At 99% times when  you bought a software or hardware, the EULA is inside the box or it shows when you are installing the software . So Is it legal hide  EULA or Conditional Contract?

I believe that no, i think that this kind contracts are a FRAUD OF LAW, by two motives:


The first motive , the corporation imposed to consumer a series rules or norms that the consumer can't negociate, so at my country(Spain) will be a abusive clausule for foul the reciprocity.

Second motive, to hide a EULA or Conditional contract without the customer knowledge  it can be considered a scam,beside  with agravating for foul proportionality, because the corporation

with great resources have abused to consumer  with low resources. So this too is a CLAUSIVE ABUSIVE.




I am pro to hack the electronic items, when the corporations try to put chains to the neck and make me a captive customer furthermore to abuse of his position power for appling me unfair rules .

When i bought an electronic item , i have all the rights. And anybody can't say me as i must to use.

[G0HZU]

I would consider it extremely unlikely that it is illegal to modify the standard box (at home) to become the deluxe box. I would draw a parallel here with a 50MHz/100MHz scope that has a similar keycode or jumper link system. It would only be illegal if it was actually illegal for a consumer to 'own' a 100MHz scope in the UK.

Exactly.  And this is all because you're not violating any laws, particularly copyright laws.  There is no copyrightable content that you'd be accessing that you didn't have rightful access to before.

Arguing what today's written laws say is pointless.

If something becomes widespread enough to affect a large company's business model then laws will appear. Bet on it.

Are you sure about that? I seriously doubt that large companies can change the law if they mess up a business model for a product like an oscilloscope. This thread is getting silly again...

[Lightages]

Is painting your 'scope a different color exactly equivalent to unlocking extra bandwidth and features from a legal and/or moral point of view?

It is a change from what was originally purchased. What if painting your scope silver made it more tolerant of being used in sunlight so it didn't overheat as badly. This would be an increase in capability, no? What if you cut a circuit trace and added a better voltage regulation circuit to reduce the nose in the input amplifiers? More capability and moral and legal. What if you change the cooling fan to make the scope less noisy? Again not part of the original specification. What if you hack away a bit inside and add a 50ohm switchable resistor so you can have 50ohm impedance inputs when wanted? Again, no problem right? What if you erased the original firmware and wrote your own from scratch that makes the scope work the way you want? Any problems there?

If anyone has any problems with any of these modifications I have just mentioned then you truly believe that when you buy something it is not yours. You have a boss or many bosses, the company(ies) that manufactured it, that tell you how to use your things and how to live.

[tggzzz]

Actually, I come from a software background. 

I'm actually relatively new at this.  In college, I started out going for a EE degree but switched to computer science because I found that I was spending all my free time messing around with computers.  Best decision I ever made, except that in hindsight, I think it may be easier to learn computer science stuff on the side than EE, so I often wonder if I would have been better off sticking with EE and pursuing computer interests on the side.  Dunno.

That was the way I did it, and I agree with your assessment. Most software is glorified CRUD, and hence very very boring.

The problems that can arise from it being crud in the first place can be quite interesting, actually, but I think one has to be something of a generalist in order to derive any real satisfaction from that.   And I happen to be a generalist (I've done a fairly diverse set of things in the computing world), so I'm able to solve a reasonably wide range of problems.  That keeps things interesting for me.

My background ranges from low noise optoelectronic systems, through generic digital and semi-custom arrays, to  networking stacks WLANs and cellular systems, by way of hard real-time, soft real-time and generic modelling software, and e-commerce systems.

When graduating I made a policy decision to avoid databases, and very largely kept to it - to the extent of removing trad databases from some product lines.

I recently regained my interest in the hardware side of things, and nowadays is possibly a better time to pick that up than ever before, what with the easy availability of inexpensive test equipment (that such segways nicely into this thread is pure coincidence, I assure you.  ), forums such as this, etc.

I'm simultaneously delighted and appalled that so little has changed since the early 80s. The major differences are that high-end industrial embedded development systems now cost tens of pounds, and the recent emergence of nanopower systems and GS/s ADCs and DACs.

[Zero999]

What's moral or not varies from person to person, country to country, culture to culture etc.

Yep, so somebody coming in here claiming "it's clear" is mistaken.

If you want a very clear example of something you're not allowed to do with your 'property': You're not allowed to put a "DS1104Z" sticker on it and sell it for a markup.

Just out of curiosity. If someone did that, then what law would they be falling foul of? Probably trademark.

What if someone hacked their oscilloscope, then decided to put it on ebay? Would it be fine, as long as they listed it as the DS1054Z and nothing else?

I know someone who hacked their  Rigol DS1052 from 50MHz to 100MHz and sold it as a Rigol DS1052 but listed 100MHz in the specification: is this legal?

[KL27x]

So you've never used it to check a digital logic signal is behaving properly (i.e. signal integrity) from any logic family since the late 70s?

Truthfully, no. I have played very little with any sort of logic chips.

You have unusually limited requirements.

Yes, my requirement are limited. (I use my scope maybe 10 times a year, even.) But maybe not so unusually so. Even 10 MHz and 25MHz scopes were very popular 5 years ago. Or even in the 70's. I wonder why it's necessary to have a 100MHz scope to use logic chips? Back when I actually used them, I didn't even have a scope.

If you scoped a TTL logic circuit on a 50MHz scope, you would see nothing useful, then?

[Fungus]

If you scoped a TTL logic circuit on a 50MHz scope, you would see nothing useful, then?

The original question was whether or not you'd see a difference between 50MHz and 100MHz.

The answer is yes, you would see a difference on any square wave. The rise time for a square wave on the DS1054Z is 7ns. The rise time is 3.5ns on a 1104Z.

[kcbrown]

Arguing what today's written laws say is pointless.

If something becomes widespread enough to affect a large company's business model then laws will appear. Bet on it.

Today's laws make the framework we currently operate under.  If you want to pretend that you operate under a more restrictive framework than that, you can certainly do so, but if you do that, then how far are you going to take it?  Are you going to make no modifications to your automobile, for instance, to improve its capability?  Are you going to refrain from improving anything you have, just because a future framework of law might forbid it?

It's your call, but I'd advise living for today (while ensuring, of course, that what you have remains sufficient for tomorrow).  Get the most of what you have while you can, because in all of this, time is your most precious and limited resource.

[kcbrown]

My background ranges from low noise optoelectronic systems, through generic digital and semi-custom arrays, to  networking stacks WLANs and cellular systems, by way of hard real-time, soft real-time and generic modelling software, and e-commerce systems.

When graduating I made a policy decision to avoid databases, and very largely kept to it - to the extent of removing trad databases from some product lines.

Actually, databases are very interesting things IMO, and incredibly useful for organizing data and accessing it efficiently, as well as enforcing integrity.  Some are better at it than others, of course, and I guess it's not much of a surprise that one of the worst ones out there is also the most popular. 

Using a database properly takes some skill and understanding.  A lot of people who make use of them just use them as containers without really considering what the real relationships within their data are and how best to model them in the database, and the end result is a mess that is inefficient at best, and often makes the data difficult to get at through anything but code.

I'm simultaneously delighted and appalled that so little has changed since the early 80s. The major differences are that high-end industrial embedded development systems now cost tens of pounds, and the recent emergence of nanopower systems and GS/s ADCs and DACs.

There seems to be one thing that has changed significantly, thanks to the massive increase in processing power: radio.  It's now software defined in large part, which seems like it should make things very interesting in that area.  I suppose we've had DSPs since the 80s so there's always been some element of that, but I'm under the impression that the key difference is that we now have what amounts to direct control over the output waveform itself, and that has to open up some very interesting possibilities, no?

Because I'm just now getting into this, I haven't studied any of that in great detail.  I'm just sort of letting my imagination run wild with the concept. 

[KL27x]

The original question was whether or not you'd see a difference between 50MHz and 100MHz.

The answer is yes, you would see a difference on any square wave. The rise time for a square wave on the DS1054Z is 7ns. The rise time is 3.5ns on a 1104Z.

Well, my question was where you have actually seen a practical difference. Between 3.5ns and 7ns rise time, this obviously makes a difference... but at what frequency/clock speeds does this difference become significant? Certainly at a specific frequency, this difference could result in a 10 fold difference in efficiency... or erroneous signal transmission. But at lower frequency, this could be as meaningless as the 10th significant digit of the voltage of a logic signal. And is not the ultimate goal increased efficiency and/or signal integrity, rather than what the signal looks like on a scope, in which case even without the most perfect representation of the wave form possible, it is still more than enough information/starting point to achieve the desired result?

Aside from what little I have come up with, now I have logic chips added to this list... at what speeds of operation does the difference between 3.5ns and 7ns make a significant difference in the ability to tune/debug/develop such a circuit?

IOW, in which specific applications that you work on would you give two shits about this difference? Because to me it seems like in the specific bandwidth where this matters there is not a whole lot of practical use. Going much higher in frequency, the 100MHz model will be as blind as the 50MHz model. Going much lower, there is no practical difference. So there is a specific range of frequency/application where it does, indeed, make a big difference. To who and for what is where my curiosity lies.

If it sounds like I'm being intentionally obtuse, please consider the simpler explanation that I am actually ignorant.

[kcbrown]

I agree with having copyright and don't believe that all software should be free but there's a very big difference between hardware and software, especially when scale is concerned.. It is true that both software cost money in non-recurring engineering costs. The difference is how the cost reduces vs the number of units produced: as the number of units rises towards infinity, the cost of the hardware reaches the lower limit of the cost of the raw materials, energy, labour, transport etc. but with software, the cost per unit falls towards zero. I feel that this point is often neglected, when people think about software vs hardware.

This is all true.  This is exactly why it's critical to point out that copyright is not being violated by entering any of these magic codes into these scopes.  Because there is no copying of copyrighted material going on, nor violation of the DMCA, there is no violation of law (at least in the U.S. -- other jurisdictions may have laws that would prohibit these things).  More importantly, there is no compromise of the value of the code by way of the creation of additional copies of it, because no additional copies are being created in the first place.

Copyright's purpose is to make it possible for creators of copyrightable works to impose artificial scarcity of the works themselves, thereby causing their works to be subject to the normal laws of supply and demand.  That is the only legitimate purpose it has.  Anything beyond that is an unjustifiable restraint upon the free market, and amounts to a claim that creators of copyrightable works are somehow more special than anyone else who creates things, and that results in higher prices and a less efficient market for all.  Because copyright law goes beyond mere prevention of unauthorized distribution of copies, a less efficient market with higher prices is exactly what we've gotten in a number of cases.  That is exactly what occurred with Microsoft -- they leveraged copyright law to impose themselves into the operating system market in such a way that other entities could not reasonably compete with them, even when those entities released their products for free (note how Linux, to name but one example, still does not come close to Microsoft in terms of adoption on the desktop).

[tggzzz]

So you've never used it to check a digital logic signal is behaving properly (i.e. signal integrity) from any logic family since the late 70s?

Truthfully, no. I have played very little with any sort of logic chips.

You have unusually limited requirements.

Yes, my requirement are limited. (I use my scope maybe 10 times a year, even.) But maybe not so unusually so. Even 10 MHz and 25MHz scopes were very popular 5 years ago. Or even in the 70's. I wonder why it's necessary to have a 100MHz scope to use logic chips? Back when I actually used them, I didn't even have a scope.

If you scoped a TTL logic circuit on a 50MHz scope, you would see nothing useful, then?

You would see some useful things, but miss important ones such as voltage overshoot, bad edges, and runt pulses. Collectively known as signal integrity.

Modern logic is faster, even jellybean components have >700MHz components.

[kcbrown]

The original question was whether or not you'd see a difference between 50MHz and 100MHz.

The answer is yes, you would see a difference on any square wave. The rise time for a square wave on the DS1054Z is 7ns. The rise time is 3.5ns on a 1104Z.

Well, my question was where you have actually seen a practical difference. Between 3.5ns and 7ns rise time, this obviously makes a difference... but at what frequency/clock speeds does this difference become significant? Certainly at a high enough speed, differences here could result in 2 fold differences in efficiency... or erroneous signal transmission. But at lower frequency, this could be as meaningless as the 10th significant digit of the voltage of a logic signal. And is not the ultimate goal increased efficiency and/or signal integrity, rather than what the signal looks like on a scope, in which case even without the most perfect representation of the wave form possible, it is still more than enough information/starting point to achieve the desired result?

Aside from what little I have come up with, now I have logic chips added to this list... at what speeds of operation does the difference between 3.5ns and 7ns make a significant difference in the ability to tune/debug/develop such a circuit?

I might be able to answer this question.

Suppose you're trying to figure out why your logic circuit isn't working.  What you see on your 50 MHz scope is a rise time of about 7 ns, and everything looks normal, so you end up failing to determine what the root cause is.

Turns out that the root cause is ringing in the signal, that the actual rise time of the waveform is less than 7 ns, and the reason for the ringing is that your board layout causes the digital connection to be improperly terminated.  Because your 50 MHz scope can't see frequency components smaller than 7 ns, it can't see the ringing that has a frequency characteristic of 3.5 ns.


A bit contrived, certainly, and I might actually be wrong about what a 50 MHz scope would see in that case (and perhaps about what a 100 MHz scope would see as well).  But it seems to me you'd have a better chance of seeing such artifacts with a faster scope (the faster the better, of course).

If it sounds like I'm being intentionally obtuse, please consider the simpler explanation that I am actually ignorant.

You and me both.   

[tggzzz]

The original question was whether or not you'd see a difference between 50MHz and 100MHz.

The answer is yes, you would see a difference on any square wave. The rise time for a square wave on the DS1054Z is 7ns. The rise time is 3.5ns on a 1104Z.

Well, my question was where you have actually seen a practical difference. Between 3.5ns and 7ns rise time, this obviously makes a difference... but at what frequency/clock speeds does this difference become significant? Certainly at a specific frequency, this difference could result in a 10 fold difference in efficiency... or erroneous signal transmission. But at lower frequency, this could be as meaningless as the 10th significant digit of the voltage of a logic signal. And is not the ultimate goal increased efficiency and/or signal integrity, rather than what the signal looks like on a scope, in which case even without the most perfect representation of the wave form possible, it is still more than enough information/starting point to achieve the desired result?

Aside from what little I have come up with, now I have logic chips added to this list... at what speeds of operation does the difference between 3.5ns and 7ns make a significant difference in the ability to tune/debug/develop such a circuit?

At any clock rate whatsoever.

If there are runt pulses, or if signal integrity problems cause poor transitions or voltage overshoot, or if there is setup or hold time violation, then the circuit may malfunction. It doesn't matter a fig if the next transition is in 1ns, 1us, 1ms, 1s, or 1 year. The clock frequency is completely irrelevant.

IOW, in which specific applications that you work on would you give two shits about this difference? Because to me it seems like in the specific bandwidth where this matters there is not a whole lot of practical use. Going much higher in frequency, the 100MHz model will be as blind as the 50MHz model. Going much lower, there is no practical difference. So there is a specific range of frequency/application where it does, indeed, make a big difference. To who and for what is where my curiosity lies.

If it sounds like I'm being intentionally obtuse, please consider the simpler explanation that I am actually ignorant.

See above. Even in 1980 a 100MHz scope was the minimum necessary. For jellybean modern logic even 350MHz is slow. Even amateur products can have clock rates of 1ns bit periods and need 10GS/s sample rates to verify signal integrity.

[Fungus]

Are you sure about that? I seriously doubt that large companies can change the law if they mess up a business model for a product like an oscilloscope. This thread is getting silly again...

You're not been paying attention to copyright law lately, have you?

...or all those secret trade agreements they're working on ("TTIP" or whatever the latest version is called).

[G0HZU]

Are you sure about that? I seriously doubt that large companies can change the law if they mess up a business model for a product like an oscilloscope. This thread is getting silly again...

You're not been paying attention to copyright law lately, have you?

...or all those secret trade agreements they're working on ("TTIP" or whatever the latest version is called).

Show me an example where a scope manufacturer has managed to 'change the law' because people are hacking the unlock codes on their scopes and I'll take you seriously.

[Fungus]

Show me an example where a scope manufacturer has managed to 'change the law' because people are hacking the unlock codes on their scopes and I'll take you seriously.

Here: http://bit.ly/4kb77v

[G0HZU]

The original question was whether or not you'd see a difference between 50MHz and 100MHz.

Even in 1980 a 100MHz scope was the minimum necessary.

If you wound the clock back to 1980 and looked at something like a typical MC6800 microprocessor PCB layout then I can't imagine you would see much difference. The PCB layout would probably be quite poor and I doubt there would be any PCB tracks with controlled impedances.

The timing requirements for these things was fairly lax and the risetimes slow so I would expect that a 50MHz scope would have been fine.

If a hobbyist was playing with one of these chips then you could probably be looking at a bird's nest arrangement for the wiring rather than a proper PCB.

Where are you going to attach the scope ground to monitor those all important (3.5ns   ) risetimes?