Isola microwave laminate

I only needed 3 square inches, but they sampled me a full panel. Good, I'll have some around for fast protos.

formatting link

This is much better than FR4, and the board houses say it handles like FR4.

Reply to
John Larkin
Loading thread data ...

That is a lot better.

Cheers

Phil Hobbs

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

Nice crisp step, no ugly drool.

I'll try a 4-layer board for real. My SFP link thing is really suffering from the FR4 losses.

Reply to
John Larkin

Simon and I were TDRing a transmission line sample on an FR4 board today. It's a time stretcher for geophysical lidar--24 T/Hs with about

50 ps aperture time, made of SAV-581+ pHEMTs. Each hold cap sits right on an input of a simultaneous-sampling ADC. The gates are driven by FIN1002 LVDS receivers via a very small Schottky diode, and are forward-biased a bit in track mode. One input of each FIN comes from a DAC output, and the other from a ~10-ns RC ramp. (There are two ramp generators.)

The ramp lines are actually the parlour trick--ringing caused by all the input reflections has to be managed carefully.

(Positively Larkinesque.) ;)

Cheers

Phil Hobbs

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

I can send you a hunk of the Isola. I have a whole panel. It seems to have a working dielectric constant about 3.5, which makes traces wider, which is generally good.

Hang on, I'm counting how many rules you are breaking.

Send me a schematic, just for fun.

I was just talking to Miguel about this: nobody wants to do analog design any more, and the guys in the big companies who used to do it have mostly retired. More biz for us, I guess.

Fast stuff is especially exotic. People quake at the n-word, and absolutely panic at the p-word. We have f-worded some.

Crossover, using RF/microwave specified parts in time domain, is special fun. Let's not tell anyone that it's possible.

Reply to
John Larkin

I did some of that around 1984, in Cambridge, UK. Word has finally reached California?

Come to think of it

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

used a BFT95 5GHz broadband transistor in a simple emitter-coupled monostable. I was working in west London at the time, but it was for a group at the University of Southampton.

John Larkin is trying to shut the stable door forty years after the horse had bolted.

--
Bill Sloman, Sydney
Reply to
Bill Sloman

Hey SL0W MAN, you just can't RESIST posting ad-hominem bullshit about John even when it is TOTALLY out of context, can you? Just shows what an ASSHOLE you are!

Reply to
Flyguy

Good,

w4Fa?dl=0

s like

ut

ht

a

the

hed California?

her",Journal of Physics E: Scientific Instruments, 12, 686-687 (1979).

stable. I was working in west London at the time, but it was for a group at the University of

se had bolted.

n

What makes you think he is human?

Flyguy doesn't understand the context."Using RF/microwave specified parts i n time domain" is clearly quite beyond him. John Larkin may think that he h as discovered this trick himself, but in fact he's decades behind the rest of the field. I can document it with things I've done. If we'd thought we w ere doing anything new, somebody would have tried to patent it, but nobody bothered, so it has to have been around for rather longer.

Flyguy has his permanent and inflexible conclusion. He's not quite up to re asoning his way to it.

--
Bill Sloman. Sydney
Reply to
Bill Sloman

+1
Reply to
John S

Someone as weak at recognising context as Flyguy. They used to be thinner on the ground around here. Maybe Donald Trump is putting something in the water to make more people silly enough to vote for him.

--
Bill Sloman, Sydney
Reply to
Bill Sloman

Sent.

Assuming my health holds out, I'm planning to keep doing it for at least another 15 years. Gizmo-building is way more fun than golf or TV-watching or model trains or whatever it is that retired guys do. (Okay, Joerg brews beer. That has its points, but it's a bit like watching paint dry. Of course he gets his suicidal thrills on his bike.)

Before boat anchors became cheap, that was not an entirely unreasonable stance.

Yeah. Trouble is, when doing that you have to buy volume, and a reel of SAV-551+es is four grand or something. I'm going to have a whack at using the 40-GHz Renesas pHEMTs one of these times. I bet they're not as stable as a SAV-551+. ;)

Cheers

Phil Hobbs

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

FG,

It would be nice if you could refrain from reposting BS's BS. It's sort of like cross-modulation due to a corroded antenna that lets it get by the trap. ;)

Thanks

Phil Hobbs

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

Cool.

Ah, you are fine-tuning a power supply with a DAC. I'm doing that too. Desperate times and all that.

We did sort of the reverse of multiple s/h channels, for the NIF beam modulators. We summed 120 gaussian impulses, one every 250 ps, each with a DAC for amplitude, to make an arbitrary waveform generator. I learned about Gibbs Ears.

formatting link

Nowadays there are fast DACs that can make the waveform they want. It would be fun to redesign.

--
John Larkin         Highland Technology, Inc 

Science teaches us to doubt. 
 Click to see the full signature
Reply to
jlarkin

PIM is interesting. All sorts of things happen at -150 dB down.

Reply to
John Larkin

Got some Rogers samples to test.

formatting link

It's a lot shinier than the Isola. I'll carve a TDR/TDT sample for this too.

Reply to
John Larkin

We do that fairly routinely for Class-H TEC drivers.

This board has three very nice octal 8-bit DACs (DAC088S885, $2ish), that set the thresholds for each of the 24 FIN1002 LVDS receivers that control the SAV-581+ sampling gates. Each hold cap sits right on an input of one of the four LTC2351CUH-14 6-channel, 14-bit simultaneous-sampling ADCs.

It's strictly the bigger-hammer approach, but still a whole lot cheaper than a 14-bit, 4 GS/s transient digitizer when you need range resolution of an inch or two, but your rep rate is a kilohertz. We're getting ten of them stuffed right now--the initial use is for a DoD SBIR project we're subbing on, but I expect it'll come in handy for jobs as diverse as geophysics and mapping swarms of honeybees. (We're waiting for word on a grant proposal to do that.) The DoD thing is Phase 1, and assuming Phase 2 and 3 get funded, we'll be helping some actual IC folks implement the same general idea in an ASIC. (No, you can't have pHEMTs on silicon ASICs, but the strays are so much lower that you almost certainly don't need them for this sort of job.)

Fun!

Cheers

Phil Hobbs

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

It wasn't BS. "Crossover, using RF/microwave specified parts in time domain, is special fun. Let's not tell anyone that it's possible" on the other hand, was.

Having to flatter John Larkin is a different kind of trap.

--
Bill Sloman, Sydney
Reply to
Bill Sloman

Gibb's oscillations

formatting link

I found out about it - the hard way - back in 1978. I had to put a Hamming window on my FIR filter - realised as a long string of very precisely trimm ed resistors, which meant re-doing the lot of them. The bats listening to t he signal didn't care much, but making the published spectra look nice was important.

--
Bill Sloman, Sydney
Reply to
Bill Sloman

Somebody still designs the multi-GHz scopes and remaining gear. One example:

formatting link

Best regards, Piotr

Reply to
Piotr Wyderski

I wonder if the 110 GHz version costs a million dollars.

Sure, some outfits are doing heavy-duty electronic design, but I think there are fewer analog/circuit designers than there were. One of my EEs is a recent grad and he says that his classmates avoided analog electronics. They all want to type code, as in c or python or verilog.

Just my impression. I don't have serious statistics, just some personal anecdotes.

--
John Larkin         Highland Technology, Inc 

Science teaches us to doubt. 
 Click to see the full signature
Reply to
jlarkin

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.