signal indicator

Yes, that works. I might be wrong about the CML swing (gotta verify that) which is officially 400 mv p-p at each pin when externally terminated. I'm not terminating so I expect 0.8 p-p. The 2-diode voltage doubler would be a little wimpy if each one only sees 0.4 p-p.

My 1-diode thing works usably at 0.4 p-p.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics

On a sunny day (Wed, 08 May 2019 07:21:47 -0700) it happened John Larkin wrote in :

Never tried that above 40 kHz or so, but why not. But it only takes .7 V to open a NPN, 150 mV for a Ge. And you can bias the detector - side to reduce that.

I tried forward biasing my schottky diode, but it didn't give me any more detector output from RF off to RF on. I think there is theory to support that observation.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics

On a sunny day (Wed, 08 May 2019 09:48:00 -0700) it happened John Larkin wrote in :

These BAT15-99 have a very low voltage drop for low currents:

If there is more than one pulse then the filter C will be charged and a very low current should flow?

I meant this way of biasing the detector:

A few gigahertz, certainly. That's not the only option, though. Rectenna is a good keyword to search on:

I'm going to use a comparator (or an opamp as a comparator) between the detector and the LED. Less risk.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics

On a sunny day (Wed, 08 May 2019 19:51:37 -0700) it happened John Larkin wrote in :

Yes of course, maybe a LM324 ;-)

I am doing some audio phase shift thing with some TLC274ACN right now... 5 mV offset, does not matter because AC only. driving an AD8346 QAM modulator to make 2.4 GHz SSB.

Quick though.

--
Bill Sloman, Sydney

None of the RF specialists I've run into has been aggressively butch.

It isn't an area where gender or sexual preferences is likely to have a significant effect on performance. Cursitor Doom may differ but he has silly ideas about wide variety of areas. A renaissence half-wit.

--
Bill Sloman, Sydney

We just failed at doing a wideband I/Q modulator to simulate a jet engine blade-tip sensor. Certain Parties convinced me to go over the top, including a super-wideband all-pass 90 degree shifter which was too clever by at least half.

Basically, everything oscillated. There were some distortion issues too.

So we're laying out rev B, which will be much simpler and will work only at the customer's fixed frequency. Those big Phoenix blocks are versatile but are also antennas.

Pity. We usually get rev A right.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics

On a sunny day (Thu, 09 May 2019 06:57:57 -0700) it happened John Larkin wrote in :

Yes, PCBs make things difficult to change....

I have this DVBS transmitter, controlled by a raspberry Pi computer, works very well:

it gets it frequency from an ebay board with ADF4350 and TCXO, also controlled by same raspberry. This is not the same circuit as on my website.

This is for video, I thought: 'I have transport stream IQ from the root-cosine filter, why not generate IQ for SSB the antique way with a 90 degrees phase shifter?', so cut the coils to the AD8346 driver, and connected an audio socket to it. Made a little phase shifter board:

You may wonder why all the trimpots, well it requires precision resistors and I had a bag full of 10 turn pots from ebay, so, Ohm meter to 1 Ohm accurate. Did not have 1% caps at hand so tried +-10% Chinese ones Anna Lyser showed great carrier suppression (in the noise), but only 10 dB sideband suppression with those caps, and an LM324 with cross over distortion, got the TLC274ACN now, and some 1% caps yesterday, but those are so big... The 2.4 GHz output will have to be amplified, again ebay to the rescue with 5 pieces RF2126 chips: The wiring will be interesting, the cooling too, soldered a 2 mm screw on the back of the chip, sticks through the board, maybe heatsink on it?

PCBs ...

Now I need some courage to build the rest.

I tried to milk individual transistors in single-stage for what they got. RF types I had to socket. In Europe they cost so much in the early

70's that they had to be shared between homemade measurement tools. Sort of a musical transistor.

My first VHF amp had a Motorola RF transistor in it but it was only affordable because it was a reject, multi-emitter version with some of them blown. So instead of the typical 10W I only got 6W out of it which was fine and within budget. Because the beer budget was absolutely not to be sacrificed for stuff like that.

That is sorely lacking where I live. In Rancho Cordova where I used to work a long time ago that's now different. They started calling that area "Craft District" because there is almost a dozen breweries with tap rooms.

IIRC the cost difference was more than 0.1c/piece versus trimmed (Asian pricing) and that ended up making a substantial difference. Designing mass products can be fun.

--
Regards, Joerg 

http://www.analogconsultants.com/

Yes! I am helping with a similar project (not laser) right now. Most engineers roll their eyes when a boss or a manager asks "Can we shave another $2 from the BOM"? With me my eyes light up, it's fun.

Right now I had to pause, can't visit the client. The handlebar of my road bike snapped off at full speed during an after-church ride and ...

--
Regards, Joerg 

http://www.analogconsultants.com/

Yowsers, that sounds intense, I'm glad you're still able to type. Was there no early indication of imminent failure?

Clifford Heath.

Well, it's been a week now and still hobbling around. Can't work much. Hopefully the left foot will heal up. OTOH I am glad this didn't happen

1/2h earlier where I was roaring down a steep hill at 42mph with big pointy boulders on the side and a T-intersection into a busy road at the end of this hill.

The other good thing was that three drivers stopped and almost got into an argument about who gets to take me and my bike home, despite me bleeding a lot. There are many good people in our society. There was no way I could have made it home by myself.

Absolutely nothing. It just snapped out of the blue, like in his case:

The whole right side starting almost at the stem was suddenly gone. I will never use aluminum for a handlebar again. And no carbon fiber, of course. Most likely my road bike gets a straight steel handlebar now. First I'll have to check it's frame for impact damage. Once I can walk again ...

--
Regards, Joerg 

http://www.analogconsultants.com/

OK, drive a cmos gate with the clock, maybe add an extra capacitor on the output to ground. Vcc (and ground!) current will increase linearly with frequency.

That's cute. A small bypassed resistor from gate ground pin to PCB ground will make a small voltage that can feed two comparators. A few tenths of a volt there won't bother anything. It could be done on the high side too, let Vcc droop a bit as frequency goes up.

Nice idea, as long as it's a CMOS gate and not ECL.

It's a charge dispenser.

--

John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com

Hey Joerg, maybe it's time you slowed down and got one of those three wheeled bikes with a basket in the back. :^)

Ducks, George H.

Golf cart.

--

John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com

And you're doing this for your health. :(

May God heal and protect you.

Cheers

Phil