Suggestions Thread 0

[talex004]

Hi Dave! Great video blog, I like it

Quick tip about the forum itself. I just registered and it sent me the activation e-mail. This first mail also contained the password I had set for my account. This is unsafe and a bad practice, I'd suggest you do something about it.

[EEVblog]

Hi Dave! Great video blog, I like it

Quick tip about the forum itself. I just registered and it sent me the activation e-mail. This first mail also contained the password I had set for my account. This is unsafe and a bad practice, I'd suggest you do something about it.

Someone else mentioned this, but I cannot find any option in the forum to disable it, so if anyone knows how then please let me know.
I can't read your password, so it's pretty secure, unless someone intercepted your email!

Dave.

[talex004]

I just found out how. You'll need to edit one of the source files of the forum. The file is here:
[SMF_ROOT]/Themes/default/languages/Login.english.php
The line you need to edit is:
$txt['register_activate_message']
Once you do it, I recommend you register a new user account, just to make sure everything is alright, and then delete it later.
I hope you understand php syntax Good Luck!

[mightyohm]

I'd like to see a discussion of multimeter "counts", how they affect accuracy, etc.

[Dago]

I'd like to see a discussion of multimeter "counts", how they affect accuracy, etc.

I'm pretty sure Dave has already covered this in some video

[EEVblog]

I'd like to see a discussion of multimeter "counts", how they affect accuracy, etc.

Your wish is my command!
http://www.eevblog.com/2009/08/21/eevblog-26-multimeter-counts-accuracy-resolution-calibration/

Dave.

[mightyohm]

My apologies, I still have 20+ back episodes to watch.  I've been getting through 3-4 a day this week.  Will check it out!   

[TrentO]

Dave,

Could you talk about the hows- and whys- of PLL's? My PIC32 uses a /2 then *20 PLL mechanism to obtain the necessary 80MHz from an 8MHz oscillator... WHY?!! Why not simply use a 80MHz oscillator, and only /2 later for the peripheral bus? This sounds kinda "Rube-Goldberg" to me!

Also, in my quest to measure the clock on everything electronic I own, I found a neat little IC from Cypress, a CY22050ZX in an old satellite modem that I had lying around-- 50Mhz in, 100Mhz, 10Mhz (x3) out-- and apparently, they're "field programmable!"  Perhaps you could white-board how this works?

Thanks,

-Trent

 

[GeekGirl]

Dave,

Could you talk about the hows- and whys- of PLL's? My PIC32 uses a /2 then *20 PLL mechanism to obtain the necessary 80MHz from an 8MHz oscillator... WHY?!! Why not simply use a 80MHz oscillator, and only /2 later for the peripheral bus? This sounds kinda "Rube-Goldberg" to me!

http://en.wikipedia.org/wiki/PLL is a great place to learn the basics of PLL (it even goes into the maths required for the filters etc), one of the best way to learn about things like a PLL is to play with one, the easiest way to do this at home is to play with the 4046B (this is a standard CMOS chip made by every man and his dog), You could go and find a discreet PLL circuit and build it, but you will not learn much more than you would playing with the 4046B.

Also, in my quest to measure the clock on everything electronic I own, I found a neat little IC from Cypress, a CY22050ZX in an old satellite modem that I had lying around-- 50Mhz in, 100Mhz, 10Mhz (x3) out-- and apparently, they're "field programmable!"  Perhaps you could white-board how this works?

http://www.cypress.com/?rID=13742 Is the Cypress page for this part, it explains everything you need to know to play with this part It actually looks like a great part till you read the datasheet and find you can only program it 100 times (I am sure you could program more than this, but this value is the MINIMUM number of times they guarantee There are some clock generators out there that are serially programmed (eg I2C, SPI etc) that you have to set them up EVERY time you power the system up, these can be changed for ever, I was actually thinking of using one of these as the heart of a frequency generator.

[cd]

A chat about your product's lifetime and the reasons for it would be interesting.
Reviews of your favoured hand tools and why.

[cygnus2112]

Hey Dave,

Great blogs, I love them.
As a 'sidebar' podcast, how about a little tour of your bench behind you?  I'd be interested to see what you got hanging out back there.
Like that little yellow panel that's mounted under the left corner of your benchtop...   

Bryan

[Dago]

Like that little yellow panel that's mounted under the left corner of your benchtop...   

Bryan

I'd assume it's the grounding point for ESD straps and such.

[rfrey]

Hi.  Dave, I love that your blogs assume a certain level of professional experience, so I almost hesitate to request this subject, but as a newb it's something that would really interest me.

When faced with unfamiliar territory, i.e. specs that have requirements you have no experience in, how do pros go about researching and choosing parts?  It seems to my uneducated eye that there are about 1.24x10^23 chips available out there... how do you keep track, how do you know you're not spending 3 days on a portion of a circuit when there's a 30p chip available from XYZ corp to do the same thing, etc?

Obviously with experience you amass a library of favourite parts in your head, but what resources do you turn to when your internal catalogue fails you?

[armandas]

how do you know you're not spending 3 days on a portion of a circuit when there's a 30p chip available from XYZ corp to do the same thing.

Simple, you just spend those three days looking for that part in various catalogs!

[GeekGirl]

how do you keep track, how do you know you're not spending 3 days on a portion of a circuit when there's a 30p chip available from XYZ corp to do the same thing.

I use various resources, RS and Farnell catalogues, the flyers that companies like Maxim and Ti send out (these product briefs are great to get you in the right direction)

If you really are stuck, suppliers reps can sometimes provide a good lead, I sometimes find it hard to describe the function I need the way that the manufacturers call it, it also helps to think outside of the square, you may need an operational amplifier, but you have to remember that just because a part is marked "Video Amp" does not mean you can not pass an audio through it.

[EEVblog]

Hi.  Dave, I love that your blogs assume a certain level of professional experience, so I almost hesitate to request this subject, but as a newb it's something that would really interest me.

When faced with unfamiliar territory, i.e. specs that have requirements you have no experience in, how do pros go about researching and choosing parts?  It seems to my uneducated eye that there are about 1.24x10^23 chips available out there... how do you keep track, how do you know you're not spending 3 days on a portion of a circuit when there's a 30p chip available from XYZ corp to do the same thing, etc?

Obviously with experience you amass a library of favourite parts in your head, but what resources do you turn to when your internal catalogue fails you?

Great question, and this is something I've been meaning to do a blog on for a long time now.

Basically it all comes down to hard work doing the research. Probably a good 80% of my design time is spent finding the right parts.
Not only trying to find out what chips are available to do certain jobs, but what ones are available in the qty I need, at the price I need, in the package(s) I need, with the specs I need, with available brand substitutes!

It 's not uncommon the change the specs of your design because of what parts you can actually get or can afford.
Cost driven designs are harder because you can't just pick the first chip you find that meets your requirements.

Not as hard as it sounds, but very very time consuming. My tools are usually the big component suppliers first (Digikey, Mouser, Farnell) to find actual parts based on various functional categories. i.e. Opamp -> Dual - > Low Power -> 5V single supply -> SO8 package

Then findchips.com and octopart.com to find the best price and how many people stock that part.

Then the manufacturer websites if needed for more exotic part searches.

Dave.

[hakko]

Regarding Gossen multimeters compared to cheap multimeters, here is what Gossen has to say about cheap multimeters http://www.gossenmetrawatt.com/english/seiten/cautiondangerousmultimeters.htm - Including a video of a burning meter. A related video is at http://gps.sozialnetz.de/global/show_document.asp?id=aaaaaaaaaaaakar

Let's see if Dave can spark the same flames :-)

That's on the the to-do list!
I'm having trouble finding info on the exact test pulse conditions used to get that result, might have to actually ask Gossen.
They are also supposed to be sending me a couple of old meters to destroy too 

Dave.

Why instead destroing all of then why not to make a raffle of some of it to forum members One per month of reweied instruments should be nice for forum users!

B.R

Matheus

[armandas]

Why instead destroing all of then why not to make a raffle of some of it to forum members One per month of reweied instruments should be nice for forum users!

I didn't know it was raining equipment in Dave's lab
Seriously though, the talk is about destroying a *cheap* multimeter.

[bmwm3edward]

How about a blog on EE design best practices.   I loved the blog about "Worst Practices", or how things are typical in an EE design firm and how managements screws things up, etc., but what are the best practices?

I have been in software for 15+ years and know a lot about best practices there and have coached a lot of newbs on how to be successful with software design while reducing headaches by using some simple project life cycle techniques/methodologies, but I would love to hear about the "ideal" way to do it in the EE world.  What are the typical roles (besides the actual engineer), etc?  How is quality verified (QA)?  How are changes/revisions managed?

[rfrey]

Basically it all comes down to hard work doing the research. Probably a good 80% of my design time is spent finding the right parts.

Thanks for the rundown - If I had been to guess a % of project time I'd have guessed in high 10s or low 20s.  This one post is a real perspective changer for me.

[rfrey]

If you really are stuck, suppliers reps can sometimes provide a good lead, I sometimes find it hard to describe the function I need the way that the manufacturers call it, it also helps to think outside of the square, you may need an operational amplifier, but you have to remember that just because a part is marked "Video Amp" does not mean you can not pass an audio through it.

I guess that's why design is at essence a creative activity and why people like you and Dave are deservedly driving your Lotus Esprits to your personal helicopters to commute to work.  As it should be. 

Seriously, it sounds like the parts search is a much bigger bag of worms than I thought, and may even be at the core of the profession in some sense.

[yoshiki]

I would love to see a comparison between electronics simulation software, some people say B2-Spice is the best, I personally agree with that but I know there is more out there, which one do you recommend, why?

[GeekGirl]

I have a good topic for a blog, SMPS, in things like CFL's, We have all had a CFL just die with a "POP" I have yet to pull one apart, I have the ideal candidate, it is a 130W CFL (power consumed not incandescent equivalent, that is around 500W lol)

I always love small discrete SMPS, no controller, just a bridge with bulk cap, 2 transistor multivibrator, transformer and then the output bridge and filter cap.

[Simon]

how about a blog about HD44780 displays ? they are probably the most eleasing thing for a pic/uC user but perhaps not the most obvious things to use ?

[kodon]

Hello,

I e-mailed to dave about a week a go with the gyroscopes... But it seems that it was not interesting enough to blog about.
Well here's an another idea:
How to improve a switching mode power supplies efficiency? What are the things to watch out etc... I'm interested in this because I desigined and built a push/pull SMPS and ran out of ideas how to make the FET's to run cooler. Well, the heating problem is now solved with a better heatsink, but there's obviously some things that I am not aware of. I studied how to clamp that thing but that just reduces the efficiency. So how on earth are you supposed to have 80% plus efficiency out from those things? My built is running now about 55 to 60 percent. The FET gate Voltage is nice and clean, I'm using a power mosfet driver +/- 1,5Amps to the gates @30kHz, I tried various deadtimes etc... I'm trying to rise 12VDC to ~600VDC for CDI.

On the LED hacking blog Dave said something about another transistor with the buck converter... what was that all about?
And generally, I just started to wonder how can you improve a SMPS efficiency? It seems that the greater the efficiency, the higher the FET's Drain to Source Voltage toleration should be? Am I totally wrong?? Also it seems that the transformer's coils should be winded in some particular order to get the best efficiency out of it So there's a lot of mysteries behind and I think this would be very useful information to anyone who watches your blog.

So how about it?