Podcast: Play in new window | Download
Subscribe: Apple Podcasts | RSS
Thanks to our sponsor, Via Designer and the folks at Triad Semiconductor! Check out ViaDesigner.com/TheAmpHour for a free trial of the software for a year and start designing your custom, low cost ASIC for mixed signal designs!
- Dave has been cleaning up his lab. Do you have a messy or clean lab? Send us pictures!
- It seems like with the amount of stuff Dave grabs from the trash room, he could appear on the US show Hoarders one day.
- It was tax day when Chris and Dave recorded. How do you deal with business income?
- Gold has been dropping recently on the markets. Is that a good thing for buying connectors?
- What about buying parts with lots of copper? Does that lower the Seebeck Effect, the primary mechanism in thermocouples?
[tube]http://www.youtube.com/watch?v=AYblSfpKRUk[/tube] - Many chips are moving towards a copper metalization layer, not just high end processors anymore.
- Dave and Chris both agree this is interesting tech, but is not moving towards chip printing.
- When there are more robots, will there be less human resource issues? Will robot suicide be as big a problem at Foxconn as human worker suicide?
- UBM has a slight shakeup lately. Test and Measurement World is shutting down.
- EE Times no longer has a print edition and hasn’t since the beginning of the year apparently. Design East (in Boston) has also been cancelled.
- Though they are primarily going to an online model, Dave reminds bloggers to not sell out.
- The MAKE Hardware Innovation Workshop has a great speaker list again this year. They’re also taking applications for people to show off previously unshown prototypes to investors and others at the workshop. You have until April 19th to submit your idea.
- If you’re really looking to dig into a design, you can design an analog circuit in your garage for less than $3K (but you’ll have to wait 4 months!)
- How do you teach the importance of iteration?
- Listeners asked about how to get started with FPGAs. Aside from pinging many of the resources already out there (Papilio, Xess, Digilent), Dave suggests to start with existing projects. Chris suggests to try and get ahold of softcore processor kits if you can afford them.
- If none of those work for you, head over to the EEVblog forum and ask your specific questions
- Kenny asks about choosing an op amp for a ultrasonic application.
- App notes are hard! Dave’s post about buying a new computer took 2.5 hours and that was without testing or graphs or anything!
- This video about the origin of negative feedback, as told by its inventor, Harold Black is one of many awesome ones available on the ATT Tech Channel.
[tube]http://www.youtube.com/watch?v=iFrxyJAtJ7U[/tube]
The t-shirts are coming! Hopefully by this week or next! Thanks again to everyone for your patience!
Thanks to epSos.de for the picture of the gold coins.
Fort Knox is full of gold plated tungsten. Ask Chinese if you don’t trust me.
T-shirt slogan watch: “The large stick of reality”
Regarding the bond wires… When I was involved in doing some hybrids, chip on board, and some IC packaging, the gold bond wires we used were 1mil diameter (0.001″) solid gold wire. There were some circumstances where 1.3mil wire was used, and other high-frequency applications where flat ribbon wire was used. The technology was thermal-sonic bonding. This is a combination of heat and ultrasonic “scrubbing” to create the eutectic bond. We used either ball-wedge bonding, or wedge-wedge bonds. The ball bonding used an arc to form a ball at the end of the wire by melting it. This ball was then bonded to the chip’s bond pad – the the wire was drawn to the package/board bonding pad and then wedged between the head and the substrate to create the wedge bond.
Regarding op amp selection: You mentioned LT Spice, which is an awesome tool. However, I always caution the young players (as Dave says) that you have to tread cautiously. Simulation models are designed to operate properly over a limited set of “correct” operating conditions, and will behave unpredictably (and usually wrong) when the device is setup/used/biased incorrectly. Simple examples – the simulator has no problem putting 1000A through a 1N914A diode, or may not behave predictably or correctly if the op amp inputs are taken beyond their valid common mode range.
Regarding circuit board manufacture, mikeselectricstuff has a 1947 video of an enormous machine that manufactured circuit boards with inductors and condensers (now called capacitors) directly embedded in the board. http://www.youtube.com/watch?v=3QHizDkrFqg
I guess I forgot that Tungsten was even used for the lower metalization layers; always thought of Aluminum and more recently Copper per IBM’s process breakthrough in the 90’s.
Question: What would be THE trade show to attend for test and measurement? On the Agilent and Tek sites, I don’t see much of anything. Rigol’s big show this year seems to be the Dayton Hamvention. For semiconductors, the situation seems even worse, though there are some events outside the US. Boy, money IS tight Chris, yikes!
Regarding Op-amp selection, yes how long is a piece of string? :o) However there are some practical guidelines before even getting to the technical stuff…
1. Would any of the Op-amps your company already use (i.e. stock) meet your requirements? If so use that!
2. If no, which of the most common types would do, i.e. LM324 etc.? (cheap, available, lost of app notes, etc.)
3. Still no luck, is there any other application specific Op-amp or one used in an app note similar to your application
that you can use?
4. Now look for an expensive, very specific part…