EEVblog #893 - Mailbag

[EEVblog]

More Mailbag Monday



SPOILERS:
Infinity PV organic printed solar cells:
http://www.infinitypv.com/
Panasonic 840 JE-840U calculator teardown
LED controller car computer thingo teardown
Old school parallel port software protection dongle teardown
World's first logic IC! The Fairchild µL900 series, as used in the Apollo guidance computer.
Dave is taken to task about his religious rant in a previous video. Can he destroy Dave with a logical argument?
Expert witness court case letter.
Cheap ebay soldering iron of death teardown
Cheap BEC brand voltage regulator for model airplanes, is it any good? Will it meet it's claims?
Silego GreenPAK Dual-Supply Programmable Mixed Signal Matrix
http://www.silego.com.cn/web/uploads/Products/product_442/SLG46532_DS_r100_03222016.pdf

[Carl_Smith]

Back in the late 1990's my employer at the time used PCAD Master Designer (for DOS) for circuit board layout.  It used the Rainbow Sentinel parallel dongle for license control.  It has two connectors because you were supposed to be able to plug your parallel port printer into the dongle and it wouldn't interfere with normal printer operation.  But we had issues with that.

PCAD eventually came out with a Windows version around 2000 or so which lasted a few years until they were bought by Altium and everyone was supposed to migrate to Altium, but the company I worked for went under before they got that far.

[nwvlab]

Glad to see Dave having the same organic PV modules on which we are performing reliability tests

[SeanB]

Got the older version of that dongle. Has a small ATMEL 8 pin micro on it, and an A9323MAV 20 pin which I assume is either a bus switch or a buffer, plus a few odd support components. Was used to verify that a certain logging program was legitimate. too bad that it was not used by the supplier according to the OEM terms.......

With the LH0033 you really want to look for the older Nat Semi databooks, which had a lot more appropriate name for them. I think the appnote as well had the naming.  Still have a few around, they were surplus parts........

[rolycat]

Dave,

Mattia's analogy is undoubtedly flawed, but so is your logic.

The fact that many religions are mutually incompatible has no bearing on whether one of them could be correct. I would agree that it is more probable that none of them is correct, based on other evidence (or lack of evidence), but that is an entirely separate argument.

In your rant you claim (of religions)

...at best only one of them is true but because they contradict each other and that alone is in effect evidence that they are actually all made up.

This is simply wrong.

Let me attempt a simpler analogy. Suppose that rumours are circulating about a buried chest somewhere on an uncharted island. One group believes it contains a heap of gold, another that it contains a set of flawless jewels. Is the fact that they contradict each other alone evidence that the chest is really empty (or doesn't exist)?

[LazyJack]

That BEC. The output voltage is fixed. Don't calculate the power by multiplying the input voltage with the output current.

[iamdarkyoshi]

That BEC. The output voltage is fixed. Don't calculate the power by multiplying the input voltage with the output current.

I commented on that in the youtube/google- comments

[EEVblog]

Glad to see Dave having the same organic PV modules on which we are performing reliability tests

Are they any good?
What's your app?

[EEVblog]

Dave, Mattia's analogy is undoubtedly flawed, but so is your logic.
The fact that many religions are mutually incompatible has no bearing on whether one of them could be correct. I would agree that it is more probable that none of them is correct, based on other evidence (or lack of evidence), but that is an entirely separate argument.

That was entirely my argument.

Let me attempt a simpler analogy. Suppose that rumours are circulating about a buried chest somewhere on an uncharted island. One group believes it contains a heap of gold, another that it contains a set of flawless jewels. Is the fact that they contradict each other alone evidence that the chest is really empty (or doesn't exist)?

It makes the story more likely to be made up.

[Jr460]

Sorry but that "micro circuit" is not what was used in the Apollo AGC.

First two versions of AGC existed the block I and the block II.  The block II was a lot more that just made with different chips, they made some changes to the hardware as they found out things they needed/missed in the early specs.  The block I did fly, just not on any maned missions.

The first problem with that package that was sent in was both the block I and block II machines were made with 3 input NOR gates, not 2 input gates.  We look at it now and think that was stupid, but in fact it was a great idea for the time.  One part.  Fairchild could ramp up production, tweak the process and get the MTBF way down.  Check some of the documents on the net, there are graphs and charts of failure rates of the 3 input NOR gates.

The block I chips were also a single gate per package, thus it took 4100 chips in a TO style case.

The block II on the other hand went to flat pack dual 3 input NOR gate, and took 2800 of those.

A few other things that need to be corrected, it was not 16bit machine, it was 15bit, but memory had a 16th parity bit.

Memory layout was strange because the code and variable store kept growing and you only had so many bits in the the instruction for addresses.  They had to do banking and add registers that controlled banks.  Even that wasn't enough and they added a bit in one output channel for a "super bank".

The other odd thing about the machine was it's math, it was one's compliment rather than two's.  This made for a few odd things like a positive and negative zero.  Fine for most stuff but not when the hardware bumps up or down counter in memory each time a gyro sends a pulse.  It think I saw some paper from one of designers that looking back the ones's compliment was a poor design choice.

[EEVblog]

That BEC. The output voltage is fixed. Don't calculate the power by multiplying the input voltage with the output current.

Yes, I know, completely wrong. Can't edit the video now, can only do a new one.

[Someone]

Casual racism, not funny.

[bitseeker]

Using a good ol' manual typewriter:

When the keyboard has no key for the digit '1', you're supposed to use lowercase 'L' in lieu of the digit, not capital 'I'.

e.g., for the number 15

    l5 rather than I5

(Yep, I had a portable manual typewriter back in the day.)

[Smokey]

NMB is a really good fan company.  Just saying.

[Fungus]

I used those dongles back in the day.

The way they worked was that you sent it a number and it encrypted it using an on-board key then sent it back. The idea was to send it several different random numbers and make sure all the responses were correct.

This means the response was different every time so you can't just sniff the bus and make something to fake it. You need the on-board key.

PS: Re "expert witness" request letters. Why not write back and say your fees are $80,000 (or whatever)? You never know. 

[daqq]

Regarding the dongles: USB software protection dongles still exist. They aren't exactly common (fortunately) but they do exist.

[Fungus]

Let me attempt a simpler analogy. Suppose that rumours are circulating about a buried chest somewhere on an uncharted island. One group believes it contains a heap of gold, another that it contains a set of flawless jewels. Is the fact that they contradict each other alone evidence that the chest is really empty (or doesn't exist)?

Such a big discrepancy means the original source was unreliable, ie. both of them are less likely to be true as a result of that discrepancy.

(and if there was another version that claimed it was full of pearls it makes all three less likely to be correct).

[elgonzo]

What kind of logic are we talking about here right now? Something like predicate logic or (modal) probabilty logic; or perhaps just wishy-washy human logic? 

It's pretty clear that Mattia's argument is based on predicate logic, whereas Dave's argument is rooted in probability logic.
(Imo, applying predicate logic is not really useful when dealing with questions/problems where uncertainties/probabilities are involved...)

[rrinker]

 About those dongles - DOS software was often stupidly simple in how it checked those dongles. Back in the day, I had to go to a class to learn CadKey, a 3D CAD program (before AutoCAD had integrated 3D - this was better than the equivalent AutoCAD at the time) in order to support our drafting department. It used one of those parallel port dongles for protection. I had a copy of the software on my home computer, and by running it through a debugger I was able to see when it made the DOS call to the parallel port. The routine did some simple checksum testing to validate it was plugged in, but rather than much with that, I was able to get the program to run by simply bypassing the call to the "verify dongle" routine. Yes, that stupidly simple.
 The few times those more modern USB ones show up, it's usually a problem because people are using application and/or desktop virtualization and those keys just don't work with the most common content delivery systems.

[bitseeker]

@rrinker: Yep, I remember those days. Just nop the call(s) and you were home free.

[German_EE]

I have been an expert witness a couple of times and it proved to be a useful source of income. The first case was appearing on behalf of a microwave oven manufacturer where the owner had done something stupid (that wasn't listed in the manual) and then sued the manufacturer. They lost because the judge and jury decided to use some common sense. The second case was appearing on behalf of my ex employer where somebody had been involved in an elevator accident.

In both cases I was paid for every minute that I spent working on the cases including time doing research, plus, after the first case ended, there was a retainer just in case the matter went to appeal. The secret is to work out how much you were paid in your last job per hour, double that rate, and look the lawyers straight in the eye when they ask you how much you will charge.

[bitseeker]

They lost because the judge and jury decided to use some common sense.

That's refreshing. Common sense just isn't so common in real life.

The secret is to work out how much you were paid in your last job per hour, double that rate, and look the lawyers straight in the eye when they ask you how much you will charge.

That's reasonable. As an independent, you have to factor in costs that your employer would otherwise cover (supplies, utilities, insurance, taxes, etc.).

[VinzC]

And for the reference

Toto, I've a feeling we're not in Kansas anymore

• , around 1981
• , 1939

I honestly thought it was a reference to the former — I have that album, it's absolutely fan-frickin'-tastic! — until I looked up on the internet and found the excerpt from Wizard of Oz . I'd bet Dave would love to tear down the enormous Serge Synthesizer shown in the leaflet.

[EEVblog]

Casual racism, not funny.

What are you talking about?
Are you saying I said something racist?

[Someone]

Casual racism, not funny.

What are you talking about?
Are you saying I said something racist?

New Zealand is a country and race too, but you don't see the problem there. Australian courts have found such "jokes" to constitute defamation of individuals:
http://www.abc.net.au/mediawatch/transcripts/s3960474.htm
So its walking a fine line. Try making such jokes about any other nation or race.

[boffin]

Casual racism, not funny.

What are you talking about?
Are you saying I said something racist?

Apparently not; according to the New Zealand Advertising Standards Complaints Board.
http://www.nzherald.co.nz/nz/news/article.cfm?c_id=1&objectid=10334691

[Brumby]

Casual racism, not funny.

What are you talking about?
Are you saying I said something racist?

Apparently not; according to the New Zealand Advertising Standards Complaints Board.
http://www.nzherald.co.nz/nz/news/article.cfm?c_id=1&objectid=10334691

Not quite the same scenario.  In that complaint the term "sheep shagger" was directed at a ram.  As such, the term expressed nothing more than a reference to normal, expected behaviour in the animal world - perhaps indelicately and with an indirect reference to something else.  Using the same term in reference to humans, is quite direct and has a bit more to it.

It's a phrase I don't use.

[nwvlab]

Glad to see Dave having the same organic PV modules on which we are performing reliability tests

Are they any good?
What's your app?

We don't have any application for them: we performed accelerated reliability tests as a part of a collaboration with DTU (Danmarks Tekniske Universitet).

Their reliability is good, especially if you consider they are organic and manufactured using roll2roll techniques. Still, don't expect a 10-year lifetime

By the way, InfinityPV had a successful kickstarter project some months ago (Heli-On), a solar-powered battery charger. So I expect they reached a good level of reliability (at least compatible with the 2-year warranty required by EU).

[apis]

Maybe you Australians should use that moniker for people from another part of the commonwealth
https://linkbeef.com/3-farmers-arrested-in-wales-for-running-sheep-brothel/

[TheSteve]

That BEC. The output voltage is fixed. Don't calculate the power by multiplying the input voltage with the output current.

And I've no doubts it will meet its rated claims - it's made by one of the best(if not the best) companies in the RC business.

[jb79]

Hi Dave!

As some viewers of the video allready stated, you mixed up input voltage and output current.
The BECs used in model airplanes/helicopters are used to power the servos and motor controllers to control the airplane/helicopter. So the output voltage is about 5V, in some cases a little bit more (6-8V in this case).

50V in a model plane/helicopter is not common, but large electric power models need such high voltage to get enough power. Usually this are 12S Lipo packs with 44,4V nominal voltage, that would be 50,4V when fully charged. For example a 12S, 5000mAh pack with 30C would deliver 150A with 44,4V, that are 6660W of power. That's enough for aerobatic model planes up to 12kg.

I think it's no complete bullshit, (only using a 50V capacitor for 50,4V input voltage). I only think the full output power won't be reached (maybe 2/3-3/4 of it might be possible).

If the voltage is set to 8V and we assume an efficiency of lets say 85%, the maximum output power and heat of the BEC would be:
16V: 15A*8V=120W => 21,2W
24V: 13A*8V=104W => 18,4W
32V: 11A*8V=88W => 15,5W
40V: 9A*8W=72W => 12,7W
48V: 8A*8V=64W => 11,3W

Can you do a short measurement with electronic load, to see if I'm right with the estimation of 2/3-3/4 of the rated output currents?

I've also found the datasheet of the controller, a LTC3824

[German_EE]

Regarding the 'racism' thing. As a schoolchild in the UK we had a German kid come to stay with us for a week or so. He spoke perfect English and my brother and I quickly discovered that all the jokes he told about the Poles were the same ones that we told about the Irish. It would not surprise me in the slightest if the same thing happens between the Canadians and the Americans, the Americans and the Mexicans, the Australians and the New Zealanders, the South Koreans and the North Koreans etc etc.

There is therefore a choice of methods if you find yourself on the wrong end of a joke like this:

a) Laugh it off and say something funny back

b) Grow a pair

c) If you really must, complain like a spoiled kid that someone called you names.

I am quite sure that if a New Zealander is called a sheep shagger to his face that there are an ample supply of jokes against Australians, they have had a few hundred years to practice since the first convicts arrived.

[roffvald]

Regarding the 'racism' thing. As a schoolchild in the UK we had a German kid come to stay with us for a week or so. He spoke perfect English and my brother and I quickly discovered that all the jokes he told about the Poles were the same ones that we told about the Irish. It would not surprise me in the slightest if the same thing happens between the Canadians and the Americans, the Americans and the Mexicans, the Australians and the New Zealanders, the South Koreans and the North Koreans etc etc.

There is therefore a choice of methods if you find yourself on the wrong end of a joke like this:

a) Laugh it off and say something funny back

b) Grow a pair

c) If you really must, complain like a spoiled kid that someone called you names.

I am quite sure that if a New Zealander is called a sheep shagger to his face that there are an ample supply of jokes against Australians, they have had a few hundred years to practice since the first convicts arrived.

We have the same thing in Norway with the Swedes and vice versa.

[bookaboo]

Regarding the 'racism' thing. As a schoolchild in the UK we had a German kid come to stay with us for a week or so. He spoke perfect English and my brother and I quickly discovered that all the jokes he told about the Poles were the same ones that we told about the Irish. It would not surprise me in the slightest if the same thing happens between the Canadians and the Americans, the Americans and the Mexicans, the Australians and the New Zealanders, the South Koreans and the North Koreans etc etc.

There is therefore a choice of methods if you find yourself on the wrong end of a joke like this:

a) Laugh it off and say something funny back

b) Grow a pair

c) If you really must, complain like a spoiled kid that someone called you names.

I am quite sure that if a New Zealander is called a sheep shagger to his face that there are an ample supply of jokes against Australians, they have had a few hundred years to practice since the first convicts arrived.

In Ireland we tell Irishman jokes 

[Brumby]

In Ireland we tell Irishman jokes 

It's true!

I'm on another (much smaller) forum and we have one Irishman there who rolls out jokes on a regular basis - and a lot of them are about Irishmen!

You've gotta love the Irish!

[bitwelder]

Isn't already racist to segregate New Zealand population into a 'race' ?

[Conseils]

Many years ago I wrote DOS and windows 3.1 software that was protected with one of those keys..
As I recall there were two ways to implement protection:
 1) Write stuff into your source code sending a number and getting something back, also make noise on the exchange for obfuscation
 2) Allow their software to add stuff to the .EXE file and manage the issue

There were at least two types of key, cheap and expensive (easy to crack and harder). There was also a master key if I recall in the developer pack with a clear window in it where you could see the IC package. I seem to recall there were a few bytes of storage on the more expensive model.
After a full 30 seconds of research...
https://sentinel.gemalto.com/software-monetization/sentinel-hl/

Some pirating software would remove some vendors 'added code'. At one place I worked at, my boss bought this thing that recorded what happened on the parallel port. Playing back to the original software the responses obtained for the relevant signalling. There was a DOS (Disk Operating System) driver that had to be installed on the machine in support of its operation. The dongle breaker thing worked, though not with the package that it was obtained for and not the advanced Sentinel key.

Engineering software was extremely expensive, mainly as the libraries for everything generally had to be developed from scratch. Graphics drivers were a nightmare and Windows drivers a pain the rear end (well Windows hasn't changed, at least some things stay the same).

Thanks for the entertaining blog posts, as you say 'back in the day' http://dictionary.cambridge.org/dictionary/english/back-in-the-day'.

[SimonR]

Dongles are still one of several valid ways for licensing software. I currently use a USB key for one of my development systems because its the best fit for the way I have to work.

[apis]

I don't see why a dongle would add any extra protection from crackers. A cracker will try to break the weakest link, that could be the dongle if it's simplistic, but typically it will be by circumventing the part of the software that does the check.  Dongle or not, there have to be some part of the software that handles the copyright section and that can always be circumvented. Dongles always seemed like gimmicks to me and I assumed that is why they went away.

[Fungus]

Dongle or not, there have to be some part of the software that handles the copyright section and that can always be circumvented. Dongles always seemed like gimmicks to me and I assumed that is why they went away.

Cracking isn't always easy though.

eg. You can do the check differently on different CPUs (hash the CPU ID). It will run perfectly on the cracker's machine but when he releases the crack he'll get 50% of people saying the crack doesn't work on theirs. Do it four different ways and that number rises to 75%. You need to crack it four times on four different machines.

Or checksum the part of the program that checks the dongle to see if it's been tampered with. Start doing this three weeks after installation (when the user has some data files) and you'll probably end up with a sale when the crack stops working.

Time delays are your friend. Do some checks only after the program has been running for a couple of hours or after a certain number of edit operations. A cracker will hate you for that. He doesn't want to sit using your software for hours/weeks to see if he found every last check.

Obviously nothing is crack-proof. But with a bit of deviousness and imagination you can make it take many months. For a game that's nearly as good as crack-proof.

It helps to have cracked a few programs yourself (ahem).

https://www.vg247.com/2016/01/08/latest-pc-drm-tech-too-hard-to-crack-no-more-cracked-games-in-two-years-says-cracking-group/

[sleemanj]

I am quite sure that if a New Zealander is called a sheep shagger to his face that there are an ample supply of jokes against Australians

Absolutely, if an Australian calls me a Sheep Shagger, I will call him, an Australian.  That alone is insult enough!

[Brumby]

Steady on ............

[EEVblog]

Using the same term in reference to humans, is quite direct and has a bit more to it.
It's a phrase I don't use.

It's a joke, get over it. Too much of that taking offence crap around these days.
It's also a term of endearment, like Septic Tank, or Pommy Bastard.

[ez24]

I agree with the last word in the letter - "bullshit"

Funny the author would end up with this word.


Bulls can eat a lot of stuff so their shit can contain a lot of different material, like the letter, but it is still bullshit.

Disclaimer: I did not read the letter (I fast forwarded the video) but it did look like bullshit (to me).

I suggest Dave stay away from these bullshitters.

[Brumby]

Using the same term in reference to humans, is quite direct and has a bit more to it.
It's a phrase I don't use.

It's a joke, get over it. Too much of that taking offence crap around these days.
It's also a term of endearment, like Septic Tank, or Pommy Bastard.

I'm happy to take the piss out of those on the other side of the ditch as much as the next guy (to a point) - but that phrase is one I would personally avoid.  That's all.

There's a lot more we share with each other than the rest of the world.  ANZAC for instance.

[apis]

Dongle or not, there have to be some part of the software that handles the copyright section and that can always be circumvented. Dongles always seemed like gimmicks to me and I assumed that is why they went away.

Cracking isn't always easy though.

eg. You can do the check differently on different CPUs (hash the CPU ID). It will run perfectly on the cracker's machine but when he releases the crack he'll get 50% of people saying the crack doesn't work on theirs. Do it four different ways and that number rises to 75%. You need to crack it four times on four different machines.

Or checksum the part of the program that checks the dongle to see if it's been tampered with. Start doing this three weeks after installation (when the user has some data files) and you'll probably end up with a sale when the crack stops working.

Time delays are your friend. Do some checks only after the program has been running for a couple of hours or after a certain number of edit operations. A cracker will hate you for that. He doesn't want to sit using your software for hours/weeks to see if he found every last check.

Obviously nothing is crack-proof. But with a bit of deviousness and imagination you can make it take many months. For a game that's nearly as good as crack-proof.

It helps to have cracked a few programs yourself (ahem).

https://www.vg247.com/2016/01/08/latest-pc-drm-tech-too-hard-to-crack-no-more-cracked-games-in-two-years-says-cracking-group/

Hmm, interesting, didn't know they were having problems now. We will see if it lasts. I didn't mean to imply it is easy, but a hardware-dongle should not make it harder.

[Smokey]

This is one of the most accessible explanations I've seen of modern type anti-reverse engineering measures.  It's a super interesting read.
http://www.oklabs.net/skype-reverse-engineering-the-long-journey/

[ngonthier]

Sorry but that "micro circuit" is not what was used in the Apollo AGC.

Hi Jr460,

actually I have not said this specific circuit and package was used in the Apollo guidance computer (AGC). In my letter I wrote that "MIT team that designed the Apollo guidance computer (ACG) selected the Fairchild Micro Logic RTL dual NOR as the main building block for ..."

I think Dave simply made a short cut

The first problem with that package that was sent in was both the block I and block II machines were made with 3 input NOR gates, not 2 input gates.

Yes, I agree with you. What I sent to Dave is a dual DUAL input NOR, not the dual TRIPLE inputs NOR used on the AGC. Still  the uL 914 is part of the same Micro Logic (uL) family.

The block II on the other hand went to flat pack dual 3 input NOR gate, and took 2800 of those.

Again, I agree with you. I didn't said the AGC used a TO package. They used ceramic flat pack.

A few other things that need to be corrected, it was not 16bit machine, it was 15bit, but memory had a 16th parity bit.

Yes, I took the liberty to say it was a 16 bits architecture , but yes technically speadking it is a 15 bits machine 


Thanks for your comments! If you are interested by the AGC and you want a real deep dive into the architecture and operation of this machine, I recommend the book " The Apollo guidance computer, architecture and operation" from Frank O'Brian at Springer.


Cheers

Nicolas

[SimonR]

I thought I saw that it said it was the same family as well. As far as I know it was the only family available at anyway.

I can also recommend Frank O'Brians book, its a really good read especially if you want to pick fault.

The AGC was not 15bit plus Parity it was 15bit plus SIGN. The sign result of a calculation is stored in this bit in order to detect overflow as there is no carry flag.