EEVblog #1326 - How Engineering Minds Think Alike

[EEVblog]

Two almost identical complex designs published at almost the same time?
How does that happen?
Let's go back to 1996 and see how similar Dave's PC Logic Analyser design is to one published in Elektor Electronics magazine.

[Kleinstein]

Using the external flip-flops has 2 more reasons:
1) The 74ACT version can accepts TLL and 5V or /3.3V CMOS
2) The 74AC.. chips at the input give some protection to the more special chips. So if damaged by something like ESD or over-voltage spikes, it would be relatively cheap chips to replace and the CPLD would likely survive.


Besides a direct copy of one design from the other, there is also the common predecessor or article that can influence independent designs. To some degree this could be the ispLSI starter kit.

[EEVblog]

Using the external flip-flops has 2 more reasons:
1) The 74ACT version can accepts TLL and 5V or /3.3V CMOS
2) The 74AC.. chips at the input give some protection to the more special chips. So if damaged by something like ESD or over-voltage spikes, it would be relatively cheap chips to replace and the CPLD would likely survive.

Yep, for sure.
Still, I would have traded off that if it was possible, but the LSI1016 just couldn't do it.

[Ed.Kloonk]

In the mid 90's I shelled out a bit of coin for a National Instruments ISA card which had 3 x 8IO.

Only used it for general purpose speaker and light relay controls and a few alarm sensor inputs. Was more handy to program rather than the LPT port, since it was just 3 byte registers and a control register for each.

I remember that article. I also remember Rod Irving Elect. I still have and use a tower computer case, the original guts of which have long been replaced.

With the NI card, I remember thinking it would be easy in software to make a decent-width logic analyzer -albeit rudimentry- if I wanted but never got around to it.

[EEVblog]

In the mid 90's I shelled out a bit of coin for a National Instruments ISA card which had 3 x 8IO.

Those I/O cards were great for all sorts of control stuff. Tons of companies made them, and still do.

[Kleinstein]

The standard LPT ports were also not that difficult to use, more or less a standard IO chip. I remember using 2 additional parallel port cards in a PC for this. One had a slight change to get an additional address.

Data transfer to the PC was still a bit slow with the LPT port. AFAIR no all lines could be used for both directions.

[golden_labels]

This video should also make one think about patents in the age when its not about a few competing companies from a single country, but millions of people from all around the world.

[Nominal Animal]

By the way, the green-screen floating head works really well when you wear a black (or dark) T-shirt!  I like it.

(This coming from a person who has to block all ads, and who even finds the animated icons in the message editor distracting.)

And although having the floating Dave-head at the bottom soothes my OCD, I do think you're right in that it works better when you move it to where appropriate for the content. 

[thm_w]

The standard LPT ports were also not that difficult to use, more or less a standard IO chip. I remember using 2 additional parallel port cards in a PC for this. One had a slight change to get an additional address.

Data transfer to the PC was still a bit slow with the LPT port. AFAIR no all lines could be used for both directions.

12 outputs 5 inputs: http://tiij.org/issues/issues/fall96/computer/printer/printer.html

I still use multiple PCI parallel ports for CNC machining. Its only within the past year or two that you can get low cost (<$100) USB based boards for control. Higher priced ones have been around for a long time though (~$180).

PCI-E parallel adapter $10: https://www.aliexpress.com/item/4001205199981.html

[benst]

Dave,

For completeness, I have a little bit more history on the Elektor version. I only have the Dutch ("Elektuur") pages here, but they usually equalled the English ones. (Elektor / Elektuur was a Dutch-only publication originally way back in the '60's.)

It started in the 7-1995 issue with a design contest. This was won by Lamesh with the Logic Analyser design (then only 32 ch) in 12-1995. I've attached a few pages from the Dutch versions: the contest , the initial design , and the final one you showed in your video.

Ben

[Carl_Smith]

I learned a long time ago that engineering minds think alike. And I am reminded of it every time I see one of my ideas independently thought of by someone else.  It's happened so many times I've often joked that I should wear a tin foil hat to keep people from stealing my ideas right out of my head.  But the reality is that us engineers all have a similar education and skills, and a similar toolbox of ideas and technologies to pull from when we see a problem that needs solving, and we come up with similar solutions when we see an opportunity.

I have a folder in my web browser with links to all the things I thought of first but never acted on and eventually someone else did. There's probably like 50 things in that folder now. Some are just project ideas like Dave's logic analyzer, but others have been turned into actual products by other people and companies.  I thought of making a multimeter with a wireless detachable LCD years before Fluke actually made that a product. 

The latest one someone else has pulled straight from my mind was an idea to make a flexible sensor that can detect how it is bent using flex PCBs and capacitive sensing.  My idea was to put it into a glove for detecting finger positions.  Just take a couple flex PCBs with capacitive sensor pads and probably a blank layer between and when you bend it one flex PCB will slide on the top of the other and slightly misalign the pads because of the slightly longer path around the outside.  I figured you could detect this with capacitive sensing methods.  I never made a prototype for a couple reasons.  The main one being I get lots of ideas and never get around to acting on many of them.  But also, I don't know if my analog design skills are up to the task of designing circuitry that can detect fractions of a picofarad of capacitance change as a couple plates slide maybe a millimeter one way or the other.  But someone else has designed this exact thing and it seems to work very well.
https://hackaday.com/2020/06/26/slipping-sheets-map-multiple-bends-in-this-ingenious-flex-sensor/

I've actually thought about just posting all my ideas somewhere even if I never actually work on the projects.  Like just post them on Hackaday.io as a project concept and just say that it's only an idea that I'm not currently working on.  Some people might think "why just give up all your good ideas?"  But if someone else does it I get the satisfaction of people knowing I posted the idea first even if I never actually acted on it.  And I might get other people to do the actual hard work of engineering my ideas.      And it's very unlikely I'd ever make any profit off "hording" my ideas anyway.

[Mr. Scram]

I learned a long time ago that engineering minds think alike. And I am reminded of it every time I see one of my ideas independently thought of by someone else.  It's happened so many times I've often joked that I should wear a tin foil hat to keep people from stealing my ideas right out of my head.  But the reality is that us engineers all have a similar education and skills, and a similar toolbox of ideas and technologies to pull from when we see a problem that needs solving, and we come up with similar solutions when we see an opportunity.

I have a folder in my web browser with links to all the things I thought of first but never acted on and eventually someone else did. There's probably like 50 things in that folder now. Some are just project ideas like Dave's logic analyzer, but others have been turned into actual products by other people and companies.  I thought of making a multimeter with a wireless detachable LCD years before Fluke actually made that a product. 

The latest one someone else has pulled straight from my mind was an idea to make a flexible sensor that can detect how it is bent using flex PCBs and capacitive sensing.  My idea was to put it into a glove for detecting finger positions.  Just take a couple flex PCBs with capacitive sensor pads and probably a blank layer between and when you bend it one flex PCB will slide on the top of the other and slightly misalign the pads because of the slightly longer path around the outside.  I figured you could detect this with capacitive sensing methods.  I never made a prototype for a couple reasons.  The main one being I get lots of ideas and never get around to acting on many of them.  But also, I don't know if my analog design skills are up to the task of designing circuitry that can detect fractions of a picofarad of capacitance change as a couple plates slide maybe a millimeter one way or the other.  But someone else has designed this exact thing and it seems to work very well.
https://hackaday.com/2020/06/26/slipping-sheets-map-multiple-bends-in-this-ingenious-flex-sensor/

I've actually thought about just posting all my ideas somewhere even if I never actually work on the projects.  Like just post them on Hackaday.io as a project concept and just say that it's only an idea that I'm not currently working on.  Some people might think "why just give up all your good ideas?"  But if someone else does it I get the satisfaction of people knowing I posted the idea first even if I never actually acted on it.  And I might get other people to do the actual hard work of engineering my ideas.      And it's very unlikely I'd ever make any profit off "hording" my ideas anyway.

Good ideas are worth nothing. Good executions of those ideas are.

[Carl_Smith]

Good ideas are worth nothing. Good executions of those ideas are.

Which is why I have a tendency to want to keep my ideas to myself, with the idea that someday I will actually do the "good execution" part.  But I seldom do.

[thm_w]

There are so many half finished or incomplete concept type projects on Hackaday. It has value, for me at least, if I search and find your project it might give me some good ideas, ICs to use, etc.

But, just a sentence or two idea, without schematic/sketches/whatever only really has value for yourself IMO.

[LaurentLamesch]

Two almost identical complex designs published at almost the same time?

Hello Dave,

I'm the author of the Elektor design. Fascinating video, I really enjoyed watching it. Obviously, when I saw your design some years ago I also thought: how come that the concept is practically the same, is this based on my design or what?

Some minor remarks:

• If I remember well, I only addressed 4K because of the size limitations of the ISP1016. As a side product it's also a bit easier to clock a 12 bit counter with 50 MHz than a 15 bit one. Achieving 50 MHz as system clock was not so straightforward if I recall correctly.
• Yes the reason why I used the ISP1016 was that they were the first CPLDs which were easily available, and could be programmed with a simple LPT adapter and free SW. I did not have the starter kit, but I think the programmer circuit was published in Elektor.
• The LS245 buffer was chosen for protecting the PLD and because it has Schmitt Trigger inputs, for better reliability of the communication between PC and analyser. Also, the PC reads back the control shift register to check if the transferred settings are correct.
• The original 2 layer PCB for the 1995 contest had larger pads and vias for easier DIY manufacturing, and the prototype PCB was also DIY. (2 layer PCBs were quite commonly made by hobbyists in the 90s in Europe, at least to my knowledge.)
• Finally, the project was not open source because you had to buy the SW and programmed PLDs from Elektor.

[EEVblog]

Two almost identical complex designs published at almost the same time?

Hello Dave,
I'm the author of the Elektor design. Fascinating video, I really enjoyed watching it. Obviously, when I saw your design some years ago I also thought: how come that the concept is practically the same, is this based on my design or what?

Thanks for joining Laurent. How did you find out about my video?

Some minor remarks:

• If I remember well, I only addressed 4K because of the size limitations of the ISP1016. As a side product it's also a bit easier to clock a 12 bit counter with 50 MHz than a 15 bit one. Achieving 50 MHz as system clock was not so straightforward if I recall correctly.

Ah, I didn't guess that RAM choice was based on the LSI limitations.
IIRC those 15 bit counters really pushed the limits of the chip, so much so I think I was pushed into that external 595 solution.

• Yes the reason why I used the ISP1016 was that they were the first CPLDs which were easily available, and could be programmed with a simple LPT adapter and free SW. I did not have the starter kit, but I think the programmer circuit was published in Elektor.

Interesting, didn't know that.
I had to buy the kit and the software was proprietary.
I did write my own programmer software for the ispGDS chips though:

• Finally, the project was not open source because you had to buy the SW and programmed PLDs from Elektor.

Mine was available on the EA BBS for free download, but people did buy programmed devices from me as well.

[LaurentLamesch]

Thanks for joining Laurent. How did you find out about my video?

Oh I just wanted to browse through the forum and saw the video on the main eevblog page.

• Yes the reason why I used the ISP1016 was that they were the first CPLDs which were easily available, and could be programmed with a simple LPT adapter and free SW. I did not have the starter kit, but I think the programmer circuit was published in Elektor.

Interesting, didn't know that.
I had to buy the kit and the software was proprietary.

I also considered using the intel iFX740 or 780 a few years before https://www.intel-vintage.info/intelmcs.htm#710768282 which as far as I remember was also easy to program and had a free SW (PLDshell).
However, parts were unobtainable for me. (In 1994, Altera bought Intel's PLD section, renamed the iFX740 to EPX740 and dropped the family some years later.) The interesting feature was that the iFX740 had already a 12 bit maskable identity comparator integrated, see page 5 in the datasheet https://docs.google.com/file/d/0B9rh9tVI0J5mUmlqeUhqb1RhYjA/edit.

I did write my own programmer software for the ispGDS chips though:

Nice! Did you have the bitstream specification, or did you reverse engineer it?

[EEVblog]

Nice! Did you have the bitstream specification, or did you reverse engineer it?

I totally don't recall! 

[MadScientist]

Just to point out , Elektor is a Dutch publication , not a uk one, and has been produced in many languages since the 1960s . In my youth it was seen as very high end , basically a product aimed at EE professionals and expert amateurs.

The fact it almost uniquely survives , is a testament in itself

[bsfeechannel]

I am surprised to know that Elektor was unknown in Australia in the 90s. It became to circulate in English in the 70 in many countries (it became being translated to German a few years earlier). Since the ads where from UK and the prices were in pounds, it give the impression that it was in fact from the UK. Later they started publishing it in other languages like Portuguese, Spanish and Italian with the respective localization. Elektor was the first electronics magazine I know to use glossy pages. Their style was very consistent. You spot a schematic or a PCB design published by them at a distance. Their projects were intended to take advantage of the latest achievements in electronics with obtainable components. Pretty neat.

I also liked their TUPTUNDUGDUS concept which I use to this day.





 

[ozcar]

I am surprised to know that Elektor was unknown in Australia in the 90s. It became to circulate in English in the 70 in many countries (it became being translated to German a few years earlier). Since the ads where from UK and the prices were in pounds, it give the impression that it was in fact from the UK. Later they started publishing it in other languages like Portuguese, Spanish and Italian with the respective localization. Elektor was the first electronics magazine I know to use glossy pages. Their style was very consistent. You spot a schematic or a PCB design published by them at a distance. Their projects were intended to take advantage of the latest achievements in electronics with obtainable components. Pretty neat.

I just checked, and I still have the number one issue of the English edition of Elektor - December 1974.

What you maybe don't know, is that back in 1983, Elektor started publishing an Australian edition. That did not last very long, maybe only a year or so? I have one of the Australian ones, and it does have local ads, but for only four companies - maybe that is why they gave up on it.

Also of note in that issue was a letter they (Elektor) received from Leo Simpson, then editor of Electronics Australia, in which he apologised for EA publishing a rip-off of an Elektor project.

[EEVblog]

What you maybe don't know, is that back in 1983, Elektor started publishing an Australian edition. That did not last very long, maybe only a year or so? I have one of the Australian ones, and it does have local ads, but for only four companies - maybe that is why they gave up on it.

The thing back then was that if that your local newsagent didn't carry a particular magazine then you never saw it, and hence never knew it existed. Especially as a kid when that was your only outlet to find things.

Also of note in that issue was a letter they (Elektor) received from Leo Simpson, then editor of Electronics Australia, in which he apologised for EA publishing a rip-off of an Elektor project.

I need to see that! 

[ozcar]

What you maybe don't know, is that back in 1983, Elektor started publishing an Australian edition. That did not last very long, maybe only a year or so? I have one of the Australian ones, and it does have local ads, but for only four companies - maybe that is why they gave up on it.

The thing back then was that if that your local newsagent didn't carry a particular magazine then you never saw it, and hence never knew it existed. Especially as a kid when that was your only outlet to find things.

Also of note in that issue was a letter they (Elektor) received from Leo Simpson, then editor of Electronics Australia, in which he apologised for EA publishing a rip-off of an Elektor project.

I need to see that! 

[bsfeechannel]

That certainly took "how engineering minds think alike" to another level. I had a cursory read of both articles published by Elektor and EA. EA's circuit seems not only to be a rip-off but also a stripped down version of Elektor's circuit (they omitted the overload indicators). To their credit they redesigned the PCB (to accommodate for the omitted overload indicator circuit and to add the power transformer), added a range switch (on Elektor's you choose the range using a jumper, but a switch is also suggested), adapted values and some ICs (due to local availability, perhaps), showed more details about construction and stated explicitly that their project was "based" on one by Elektor.

[ozcar]

That certainly took "how engineering minds think alike" to another level. I had a cursory read of both articles published by Elektor and EA. EA's circuit seems not only to be a rip-off but also a stripped down version of Elektor's circuit (they omitted the overload indicators). To their credit they redesigned the PCB (to accommodate for the omitted overload indicator circuit and to add the power transformer), added a range switch (on Elektor's you choose the range using a jumper, but a switch is also suggested), adapted values and some ICs (due to local availability, perhaps), showed more details about construction and stated explicitly that their project was "based" on one by Elektor.

There are hints that was not the first time EA copied something from Elektor - use of the words "another ELEKTOR design" and ""this time" by the Elektor bloke. Perhaps they had mentioned something before. That was the fourth Aussie issue, and the only one I got.

Edit:
I don't recall ever seeing the EA article before - did not know that they acknowledged Elektor in it.