Another prototype

ch.JPG

op.JPG

n.JPG

Out.JPG

Some things do get easier. I've used lumped constant delay lines to get non

-overlapping drives for complementary fast MOSFETs, and built my own 2nsec monostable with a pair of discrete broad-band transistors. I should have us ed an emitter coupled multivibrator, but did it the easy way first, which w orked well enough that I didn't get the chance to do it right.

Ghiggino, K.P., Phillips, D., and Sloman, A.W. "Nanosecond pulse stretcher" ,Journal of Physics E: Scientific Instruments, 12, 686-687 (1979).

Reminds me of bad memories from grad school...except you don't have enough timers, and they aren't triggering each other. ;)

Cheers

Phil Hobbs

Reminds me of university times. At one institute they were of the opinion that the capabilities of a freshly minted engineer is inversely proportionate to the number of timers he used in his final project. So here we had a guy who was close to bursting into tears because his stuff didn't work and the professor insisted that an agreed upon schedule be kept. I ripped out a massive chunk of sauerkraut-wired logic and replaced it with four 74121 timers. The system roared to life and joy came over the land. But ... now we had a huge rack-sized board in there that was largely empty except for those four lone timers. So we fabricated and mounted a "shield" for it.

I had an associate at GenRad, Ed Greenwood... his title was Digital Guru, mine was Analog Guru ;-) He'd go ballistic anytime I wanted a timer in my logic, but I did convince him to give in on just a couple of occasions.... like a PLL that would lock to floppy data while I swung the drive in crazy circles over my head (we were dealing with portable testers and drive speeds wandering all over the map). ...Jim Thompson

Nowadays I often encroach on their turf. At one client I only got a coffee and came back the hallway with my cup. One of the firmware guys stuck his head out and exclaimed "Hey, dudes, run. George is coming and I bet he wants another one of those 16-bit timers".

If all signal propagation proceeds from left to right, it is by definition not a hairball.

I can't download the pics at home. I've used a HV Pockels cell, and know squat about the HV driver. (You plugged in a length of coax to set the pulse width.)

But where's a good place for gold-plated proto types? (I may be doing wire bonding onto pcb's and the wire bonder guy said to get gold plating.)

George H.

I had Cirexx gold plate two square feet of blank 0.062 thick FR4, for hacking prototypes. I can send you a chunk if it would help.

Most PCB houses will do boards with gold-over-nickel pads. The gold is "immersion gold", something ballpark 10 microinches thick, but they'll do any thickness that you can afford. Ask the bonder guy maybe.

Do you know the different types of gold finish, how they are made and what they are good for? I've never found a site with a good explanation. ENIG is what you are talking about I believe, right? I've seen "Hard" gold and I think there is at least one other type of gold finish. I've never seen anything saying what each is good for.

I don't know much about that. I know that copper diffuses into gold and makes it tarnish, so on PCBs you have to use really thick gold (like, 1 mil) or use a nickel barrier layer and very thin gold. We specify "immersion gold over nickel", which is great for ROHS BGA work.

When you solder the gold, the gold dissolves in the solder, so you are really soldering to the nickel. But nickel doesn't solder! Explanations welcome.

akes it tarnish, so on PCBs you have to use really thick gold (like, 1 mil) or use a nickel barrier layer and very thin gold. We specify "immersion g old ove nickel", which is great for ROHS BGA work.

ally soldering to the nickel. But nickel doesn't solder! Explanations welco me.

My guess - and it's only a guess - is that bare nickel oxidises, and the so lder can't get through the nickel oxide to diffuse into the nickel metal.

Gold-plated nickel is protected from oxidation by the gold, right up to the moment when the gold dissolves in the solder, so the bare nickel would nev er get a chance to oxidise before the solder has "wetted' it.

This strikes me as plausible. I'd need to do a bit of reading before I coul d convince myself that it might be correct.

Nickel is solderable, it just takes a lot of time and heat, and an aggressive flux. Which supports the oxide idea.

Tim

-- Seven Transistor Labs Electrical Engineering Consultation Website:

My guess - and it's only a guess - is that bare nickel oxidises, and the solder can't get through the nickel oxide to diffuse into the nickel metal.

Gold-plated nickel is protected from oxidation by the gold, right up to the moment when the gold dissolves in the solder, so the bare nickel would never get a chance to oxidise before the solder has "wetted' it.

This strikes me as plausible. I'd need to do a bit of reading before I could convince myself that it might be correct.

-- Bill Sloman, Sydney

I seem to recall that unlike many finishes, nickle oxide is conductive even if it is not very solder-able. I learned that from model railroads. The brass or copper plated tracks (which ever is the orange/yellow looking stuff) oxidize and don't conduct until sanded a bit. Nickle plated tracks still oxidize but conduct anyway.

Battery contacts are usually nickel-plated steel, difficult to solder.

The gold over unoxidized tin theory sounds good, but I wonder how people manage to plate it that way. Some tricky chemistry maybe.

This partially explains it.

Ni-NiO-Ni makes a tunnel junction with a barrier height of about 0.2 eV. Those MIM (metal-insulator-metal) junctions respond quite literally from DC to daylight. I've used them as detectors at 200 THz myself, and other folks have measured visible-light response.

Down at DC (i.e. below 1 THz) they cause all sorts of harmonics and IMD, which is why you don't use nickel connectors for radio if you can possibly help it.

Cheers

Phil Hobbs

The Giga-camp rules of the finnish hams include a stocks punishment for talking of frequencies below 50MHz, with the exception of DC and IF. May I quote the statement above for the next camp ?

Everything below 1GHz is just slowly varying DC. :-)

All the speculations about NiO sound perfectly reasonable to me.

Say John, Can you give me any pointers abut PCB which are safe 'around' food and people? I have perused the ROHS directive, but what kind of PCB would be suitable for people facing equipment?

Appreciate any thoughts and directions. jb