Podcast: Play in new window | Download
Subscribe: Apple Podcasts | RSS
Our friend and previous guest, Jeri Ellsworth, returned to the show to talk about her latest projects. And holy wow, they are pretty great!
- Jeri discusses the process of making transistors in your home using household chemicals, silicon wafers and a small kiln or oven. More info on her process of making the chips over at her recently created “JeriEllsworthJabber” channel.
- The making of a silicon ingot (we called it a boule on the show)
- Later, the Japanese industry perfected some of the processes.
- These days, discoveries come in the form of creating an all organic processor.
- In the silicon process used today, they are creating 20 nm line widths while still using 193 nm technology!
- Jeri is talking to Intersil about doing some work for them. Dave mentions that it’s important for chip companies to get their names out there
- An EEVBlog forum member pointed out that the Fluke87 removed expired patents on the latest rev.
- This Day In Nerd History: Intel was incorporated in 1968. The founders were Andy Grove, Gordon Moore and Robert Noyce.
- Jeri explains more about her Software Defined Radio mentioned on a previous episode of The Amp Hour.
Not as many links as usual, but that’s because we were too busy talking about electronics (as requested!). Keep an eye here for a video later today.
This was the worst show ever. Less electronic talk!
+1
I wish the guys would at least let you finish your sentences. Raised in a barn, both of them.
The Best show ever! I’ve gotten tweaked a few times from this podcast, Cosmic rays and radiation effects, Software defined radio, and now home-made semiconductors, Wow!
I studied Physics in school and stopped with a Bachelor’s so I’ve ended up doing a lot of electronic stuff through the years. Closest I ever got was Space radiation effects testing back in the late 80s. All the electronic boxes were viewed as “Black” so when one didn’t work at at somebody’s particle accelerator we would throw them back in the shipping crate and get another. Very frustrating and made me want to understand the signal processing electronics for the radiation detectors and counters.
Now Jerri talks about “tabletop semiconductors”! A wonderful new opportunity to get into the guts of stuff and tweak it.
Thanks for the great show and the shout outs for Bob Pease and Jim Williams, what wonderful treasures the world has lost.
One of the best shows yet. Definitely two nubs up.
———————————
In the uC space I see warm fuzzies (or curmudgeons comfort) also coming from whatever they were exposed to and what was available at whatever institute of learning they went to. Clever vendors will try to get their gear into these places. Free boards and parts is good. Free compilers/debuggers/IDEs with open source hardware debugger/programmer interfaces is better.
If i’ve got a warm fuzzy it’s with Freescale. It’s just what I’m used to. But i think it’s important to be vendor agnostic. Otherwise you might miss a better solution
————————-
I know Intersil have a bunch of product categories but i always think of them as candidates for RS232/422/485 parts in a project.
JD
ICL7106 is what I think when I hear Intersil. The bog-standard multimeter IC that made 3 1/2 digital handheld multimeters affordable. And that, or clones of it, are more often than not still to be found in $5 multimeters.
It’s spelled silicon “boule”
In proper english its spelt “spelt”, American
Those who such thing matter too, have for some time now determined thet both “spelt” and “spelled” are correct usage.
No points lost for Jeri at all. 🙂
I’ve used intersil parts before, but I never really put any real thought into it. I really like linear technology parts too. Philips (NXP) really pisses me of, because they tend to ignore hobbyists and students.
Now THAT was an excellent show. Good job.
Happy Birthday Amp Hour!!
Congratulations guys 🙂
Happy b-day guys!!!! good show like allways
Really liked this episode. Some interesting background on the chip building process. The three of you should do more shows together. You guys brought up a lot of good topics.
And please stick with electronics, just like today. It was a good show.
I *think* next week will be the anniversary. The first programme was effectively the “zeroth” episode, not 1 week after the start. If you see what I mean. At the moment, TI have quite a lot of goodwill from me, but Analog and Atmel are very much on the wane, as every contact now seems to go via sales and PR people rather than directly to engineers, and they seem to exist purely to say no, and lose your emails. I agree about NXP, they just aren’t interested unless you are in the 100,000+ unit territory.
I have to agree! You have to complete 52 weeks to complete a year so the second year will not start until the beginning of week 53 (next week) I wonder if MoonPig have an appropriate card? http://www.moonpig.com/uk/default.aspx?adid=GUKmoon&gclid=CIWblfSJkKoCFQEf4QodnCTlwQ
I loved all the info Jeri brought to the show. It felt like she had a high information density, requiring a higher SNR (more attention on my part) at a given symbol rate (rate of speech in phonemes per second). It really made Shannon’s Law come to mind.
I actually like you discussing topics that are only tangentially related to electronics, esp if you have some facts or an expert guest. I would enjoy intentional tangents like engineering view of unrelated things like finance, dating, behavior of 1st anatomically modern humans, automation and jobs, Punic Wars, or whatever. I’m probably alone in that. The most important thing, IMHO, is to remove dead time when you’re looking something up, discussing what you should talk about, thinking, calculating, etc. A little dead time is okay; the show is still great.
When Chris talked about a cartoon showing “the guy dropping acid…” it reminded me of how many times last we I uttered a variant of “I think I smoked the pot” regarding a digital potentiometer.
My warm fuzzy IC vendor is TI. Maybe it’s the free seminars or the support they get from disties. I actually think the job of rep is very important for reps who realize they have to be excellent at making connections and bringing together people who might use their part together because just looking up parts and sending samples won’t cut it.
No this is the 52nd week so in terms of number of shows… this is their 1 year anniversary. The date was August 7, 2010 ….. i rember it like it was yesterday….HAHAHA Later guys!!!
I’m sorry I said anything now. I’ve started a war of pedantry.
Forgive me.
lol Adam, I just like to read correct info.
Hi, I don’t mean to be rude, but Where is the damn promised video??!!
yea thats what i am talking about!
Well, I’ll give the full story next time, but basically I saved it to a proprietary format by accident and now I can’t get the damn software to render it, no matter what I try (different versions, different computers, every combinations of settings I can try). So basically the plan is to make this an ad-hoc radio week. We can try video again next time (and are always willing to have a video expert help us out).
Who cares, anyway? The best looking of the three of us was frozen the whole time! No, seriously, I pulled a face muscle right at the start of the show, could barely move!
Sweet! Chris, looking foward to the video next week! pay attention next time, I mean you’re a damn engineer! you guys dont make simple mistakes like that!!! lol just kidding man. looking foward to next weeks show!!!!!
Normal people argue about solar panels. Not Jeri she smiles and tells you exactly how its made step by step and where the energy in passes the energy out and states facts.
SDR Comments
1-Direct conversion has the LO frequency right at the RF frequency you want. However, you need to have I/Q mixers to get a complex signal, useful for other then AM or FM demodulation. I think Jerri was describing a lo-IF approach, where the IF frequency is at an audio frequency. You can do this as well, but the image frequency can cause trouble
2 – Digital filters can be sharper than analog filters, but without analog filters in front of the ADC, you can and will pick up unwanted signals that can jam your desired signal.
3 – Superhet’s don’t throw away half the information, you can do I/Q in the digital domain after the ADC.
Cool things to play with..
Dave
Dave
1- I was talking about IQ demodulators and there is not an image problem if you work with both the real and imaginary part of the signal
2- My point was the there is a big advantage to the phase relationship of IQ data digitally. You can work with this in the analog domain, but it’s difficult to make circuits that are linear across a wide IF bands(don’t introduce phase errors)
3- Half of the information (the sum) in a typical superhet is filtered out(thrown away). You can use balanced mixers, but issue #2 becomes a problem.
-Jer
You’re right, I went and looked at the first of your youtube videos, and you’re doing direct conversion. As long as you’re doing direct conversion, there’s no image problem. I believe all cellphone chipsets use direct conversion at this point, so it’s pretty popular.
For a low IF approach, where the IF frequency is at “Low” frequencies, you get image suppression, but it’s dependent on how well I and Q match, in both amplitude and phase. Better than 30 dB is tough without calibration.
WIth a superhet, the sum and difference output’s of a mixer have the same information content, with just a spectral inversion between the two. When you filter out one or the other, you simple lose a little signal power.
Dave
Damn proprietary format (Homer’s style) 🙂
Immensely enjoyed No.52. Thanks for having Jeri on again and for a fascinating session. Dave “motormouth” is slowly learning to listen more but it may take quite a while 🙂
I don’t actually mind not having video for the Amp Hour, having used up almost all my monthly bandwidth downloading Dave’s HUGE tutorial videos.
Best part of Software Defined Radio: To add a new feature, or “upgrade” your radio, or improve your filter, etc., etc., you usually just have to download a software patch – or do one yourself. It helps save the little solder babies.
Analog Devices earned the warm fuzzies from me when they sent free samples of their rate gyros, via Fedex overnight, to me for a college project with virtually no production potential. I’d still go out of my way to spec them for a design, if given the chance.
Interesting thing mentioned there about older technology coming round again.
Organic semiconductors were being worked on in 1974, so there’s certainly been enough time for them to grow up with silicon. Whether or not they’re simply a missed opportunity or fundamentally flawed in terms of speed versus silicone, time still has to tell I suppose. There are certainly applications where they can already be forecast to better silicon, but I don’t believe speed is one of them due to structural nature of a polymer versus a crystal and the velocity of the carriers.
As for home chip fabrication becoming everyday common, ha! Most people can’t set the clock on their VCR. Why would they ever go to the effort of making the chip to make the clock function? If it cost about a thousand dollars for Jeri to make A transistors, A transistor will have to rise to the price of at least a thousand dollars to offset the cost of anyone else bothering for anything other than academic lols.
Where’s my jet pack and holiday to space I was promised?
Oh, and the kind of furnace used for oxide growth on silicon has been available for around a hundred years. And yet barely anyone has one. Christ knows how long it’ll be before MBE becomes a common household item. Most people don’t even know what it means.