DG sort of pioneered giant PC boards, while most everyone else had lots of small boards on wire-wrapped backplanes. I heard that only one DG employee had his own parking spot, the guy who did the PCB layouts.
Fun book, "The Soul of a New Machine."
Part of Steve Wozniak's self-taught education was study of the very clever Nova design.
-- John Larkin Highland Technology, Inc picosecond timing precision measurement jlarkin att highlandtechnology dott com http://www.highlandtechnology.com
"standardvia" on this board is 15 mil drill, 25 mil OD. That's medium-sized these days. I guess they could have been bigger. There are some "powervias", 25 mil drills.
It's a "solder sample", which we wrangle out of the PCB suppliers; it's hung on my wall now. It may well be 100% functional, or it might have failed electrical test.
Some suppliers disfigure the solder samples somehow so we can't use them.
If we ever have via problems, we dump that board supplier. It's happened a couple of times in 30 years.
Do one and show us.
No need for that, and it would be a hassle to make another board. The ground and power pours will spread out the heat, not that I expect much heat.
The main busses A/B/C/D/E each measure about 1100 pF to ground! That will limit use to roughly 1 MHz, but that's OK for the stuff we plan to test.
-- John Larkin Highland Technology, Inc picosecond timing precision measurement jlarkin att highlandtechnology dott com http://www.highlandtechnology.com
So your English is correct from the outset?
-- \_(?)_
Why did relays make it hard to design? I would think they'd be no real trouble at all unless you are worried about ground bounce from switching currents.
I noticed one large BGA part. Is that the LPC3250?
I also see a couple of what look like land grid arrays. What are they? U39 and U40 Or are they BGA with a polarization by the missing balls? I've not seen that foot print before.
Having to set your design rules for the whole board by the fine pitch BGA seems a shame, but I guess that didn't run the cost up tons.
It does seem like the board could have been laid out a bit differently with the relays packed more efficiently and saved a lot of board space. Is there a particular reason to leave so much space? Putting the driver chips in the rows with the relays would have saved nearly half the board space in one direction at a small (or no) cost in the other direction. I expect cutting the size in half would have cut the price in half since you could get two per panel.
-- Rick
?? SMT relays?? What do you do when one needs replacing?
I don't know about the relays John is using, but the ones I just got a few tubes of need clearance between relays so they don't magnetically interact with each other, per the datasheet.
-- Cats, coffee, chocolate...vices to live by Please don't feed the trolls. Killfile and ignore them so they will go away.
Replace it. We use almost 100% surface-mount parts, so we have lots of rework gear. We rework all sorts of stuff like BGAs that are way nastier than relays.
The spacing of the relays is pretty low density, so rework should be easy.
-- John Larkin Highland Technology, Inc lunatic fringe electronics
If probably has the same procesing power in the relays alone if you disconnect the microcontrollers.
Without question it will be something horrible sounding/bad word plays like all those new names for healthcare-ish companies.
I'm going with "entrusta" or "Benchmarke"
I can't find where the relay was indicated. How much spacing is required for your relay and is it specified center to center or package outline to outline? If nothing else, these boards could likely be shrunk a lot by offsetting every other row and reducing the spacing between rows.
But then miniature boards are what I tend to do. I can't recall the last time I used a relay in a design, so that's not really my area of knowledge.
John's boards tend to have lots of room so routing can be done at leisure. I wish I got that sort of easy design.
-- Rick
By definition. Nobody is in charge of the English language, which is one of its charms.
'The question is,' said Alice, 'whether you can make words mean so many different things.'
all.'
I was explaining the conventions of writing manuals, in a meeting this week. One never uses personal pronouns or commands, and everything is in the passive voice. "Protections are provided to Q4..." and such. The passive voice nicely avoids assigning blame.
Two of my people are native Spanish speakers, and they say that there is no passive voice in Spanish. The closest they can come is to say that something does something to itself.
-- John Larkin Highland Technology, Inc picosecond timing precision measurement jlarkin att highlandtechnology dott com http://www.highlandtechnology.com
What is passive about calling someone a "fathead"?
-- Rick
My brother is fond of saying, ?Language is descriptive, not prescriptive.?
Of course he says that after I challenge some new word he makes up, and he makes some zingers.
The two connectors at lower-right are for testing SMA attenuators. We use attenuators in production test and we need to be sure they are accurate, not fried or broken or anything.
The idea is to use the tee model and figure out the three ideal resistor values, R1-R2-R1.
The measurement will be to apply 10 mA to one end of an attenuator (from the Agilent SMU) and then measure the voltage at both ends with the DVM, which will be 10mA*(R1+R2)and 10mA*R2 respectively. Then switch ends (via the relay matrix) and do it again. The whole measurement sequence will take about one second.
Just to see how that might work, I measured R1+R2 on a bunch of much-used MiniCircuits VAT-series attenuators that I have in my office.
These are amazingly good and consistent. The R1+R2 values were within about 0.2% mostly. Calculated values are in red. My measurements are systematically low, which could be caused by my setup.
-- John Larkin Highland Technology, Inc lunatic fringe electronics
-- If you call it a ruler, then no one is presuming to correct your misuse of English itself, but rather your choorly posen words of choice. The device in question is an architectural scale, not a ruler, and the legend on each face identifies the number of divisions, per inch of scale engraved on that face. I don't think anyone was presuming to correct your English on this occasion but, rather, was casting aspersions on your observational comprehension and your knowledge of things technical. For example, if you pan left and look carefully at the scale above the densely engraved one, you may notice that it's legended as "10". Then, carefully inspecting the divisions by counting the marks between inch divisions, you'll find that each inch is divided into 10 equally spaced segments; hence the "10" legend refers the number of segments per inch on that scale. Likewise, the "50" refers to the number of segmets per inch on the scale it refers to, which would be 50, not 5. John Fields
It was a joke- imagine the same half sigh any drafing teacher had to go over each year with everybody.
On Saturday, March 19, 2016 at 10:12:09 AM UTC-7, John Larkin wrote: ....
...
That cutaway looks more like a 'Pi' attenuator than a 'T'.
kevin
Doesn't matter. From outside, they are equivalent.
-- John Larkin Highland Technology, Inc picosecond timing precision measurement jlarkin att highlandtechnology dott com http://www.highlandtechnology.com
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.