cotton-pickin Pulse Picker

Some reading for a lazy Sunday afternoon.

A typical femto-second laser might give 65kW 0.1ps light pulses at a rate of

80MHz. But some applications can't handle a powerful light pulse every 12.5ns, for example fluorescence-lifetime measurements (we don't want to re-excite any molecules before achieving 100% decay). The solution is to add a light modulator, and run it from a pulse-picker. The picker should have adjustable pulse width and delay or phase, to exactly enable the modulator in the middle of a selected pulse. Pulse picker synchronizers typically cost $6000, but I figured it shouldn't be hard to design and make them for say, under $100.

Here's my recently-completed RIS-761 divide-by-N pulse picker.

formatting link

It combines fast logic gates with a few analog-circuit functions to provide synchronization and shaping. In practice the phase and width settings would only be made once per laser setup, so knobs are not provided. The delay / phase is adjustable over a 20ns range, more than needed, and the width from 5 to 18ns. The 99cm pcb slides into a Hammond 2x4x4-inch extruded box, which is the most expensive component. Parson my poor cell-phone photo quality.

--
 Thanks, 
    - Win
Reply to
Winfield Hill
Loading thread data ...

Looks fun. I gather you're mounting those BNCs on both sides of the board. I'm way too chicken to use SMT BNCs, because grad students are always stepping on BNC cables and tearing them out of the plug.

These are the ones I like--they're solid zinc alloy and pass the stomp test with no worries:

formatting link

Of course it's a bit tough to mount them right opposite each other.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
 Click to see the full signature
Reply to
Phil Hobbs

We use the Bomar V-BITE connectors on some of our boxes.

formatting link

They are sort of edge-launch, with the connector centerline almost in the PCB plane. They really grab onto the board.

--
John Larkin         Highland Technology, Inc 

lunatic fringe electronics
 Click to see the full signature
Reply to
John Larkin

Yes, I've long used those beasts, but have become a recent fan of stacked-pair BNCs. Interestingly, all manufacturers use 0.6-inch vertical spacing, but other than that, they differ substantially. I'm using the molex 73415-3980, which uses strong metal brackets and features a pair of 2-56 threaded panel-securing mount holes.

formatting link

--
 Thanks, 
    - Win
Reply to
Winfield Hill

It combines fast logic gates with a few analog-circuit functions to provide

They are sort of edge-launch, with the connector centerline almost

I looked at those, but iirc they're getting hard to find, and I don't really trust the mounting pins on the back--they're actually a loop of wire press-fit into a slot in the connector body. Plus they're only supported by the board.

Something with a nice bulkhead thread that transfers the stress to the box instead of the solder joints is very reassuring. (You do have to let the board move around a bit.)

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
 Click to see the full signature
Reply to
Phil Hobbs

As I mentioned, the ones I'm using have such a feature, and the wire is soldered. BTW, I added a pic of the pcb, panel and connectors from the back to the dropbox folder.

--
 Thanks, 
    - Win
Reply to
Winfield Hill

BTW, John, note my use of 65LVDS2-series LVDS receivers throughout, instead of comparators, inspired by your post a few years ago. They're great! Fast, stable, in sot-23 package. I added a Bourns CD143A-SR3.3 TVS (in a sot-23-sized package) to protect their fragile inputs.

--
 Thanks, 
    - Win
Reply to
Winfield Hill

I have some data on offset voltage and jitter vs cm voltage... somewhere. They are pretty amazing fast CMOS-input rrio comparators, not advertised as such. Their outputs are pretty fierce, too. About

1/4 the price of an official comparator at this speed.

I've used them as customer-input triggers and they don't seem to get damaged. A parallel RC in series with the input will protect the esd diodes from gross DC overloads.

You might save a few ns by using the NC7SV74K8X flip-flop. Tpd is 1 ns, faster than old ECL. 16 cents!

We've done a few custom pulse picker boards, usually with a bunch of other functions, like photodiode amps, multiple delay/width outputs, other i/o.

Here's one, really a laser controller, but its basic function is pulse picking.

formatting link

--
John Larkin         Highland Technology, Inc 

lunatic fringe electronics
 Click to see the full signature
Reply to
John Larkin

Oh, what are those BNC "header" type things? Are they custom?

George H.

(Thanks for the schematic.)

Reply to
George Herold

Were you asking me? My parts were Molex 73415-3980,

formatting link

--
 Thanks, 
    - Win
Reply to
Winfield Hill

Yeah anyone.. .thanks. (I guess I need to send more time browsing parts catalogs... that's (in some ways) harder to do online these days.)

George H.

Reply to
George Herold

Nice, but my picoseconds wouldn't make it around the bends.

Calling this a "75 ohm" connector is interesting too.

--
John Larkin         Highland Technology, Inc 

lunatic fringe electronics
 Click to see the full signature
Reply to
John Larkin

Probably more about the pin diameter than anything.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
 Click to see the full signature
Reply to
Phil Hobbs

BNCs are good to roughly 3 GHz, 100 ps maybe, so those connectors might be OK. Maybe a little inductive.

We don't use BNCs on anything faster than about 500 ps. A good BNC is actually pretty impressive at high frequencies.

--
John Larkin         Highland Technology, Inc 

lunatic fringe electronics
 Click to see the full signature
Reply to
John Larkin

A lot of my stuff goes on optical tables, where BNC is *de rigueur*, being both traditional and very practical, even for low frequency stuff.

For use inside larger systems, I've been designing in a lot of Displayport connectors lately. You get four lanes of shielded twisted pair plus control wires, it has a retaining catch so it doesn't jiggle loose, and it's cheap like borscht.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
 Click to see the full signature
Reply to
Phil Hobbs

Nice idea and thanks for sharing

Just a question, you are paralleling the 74LVC2G04 inverters. Is that safe to do?

Cheers

Klaus

Reply to
klaus.kragelund

Yes, it's 75 ohms. As slow as 2ns rise/fall time is fine for a pulse picker, although one does want less drift and variation than that. The short distance of impedance mismatch isn't even noticeable with a 500MHz scope. I'm happy to be able to double the number of BNCs onboard.

--
 Thanks, 
    - Win
Reply to
Winfield Hill

I'd say yes, indeedie! Note they're both in the same package, and likely have identical propagation delays, etc.

--
 Thanks, 
    - Win
Reply to
Winfield Hill

We parallel three or even six gates to get more drive. Seems to work.

--
John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  
 Click to see the full signature
Reply to
John Larkin

$6000 sounds extreme.

Our little delay generator can do pulse picking. It has an input divider and N-of-M burst triggering.

formatting link

Some pulse pickers need more outputs, for firing multiple Pockels cells and stuff. Sometimes we don't pick the next pulse, but one several pulses out in the train, so we need a semi-complex set of pretty accurate delay+width outputs.

Closer to the geometric mean of $100 and $6000.

--
John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  
 Click to see the full signature
Reply to
John Larkin

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.