RF is horrrible

Jan 01, 2020 87 Replies

Am 03.01.20 um 20:50 schrieb John Larkin:

I did not have a GVA-63 here, only its grandpa, an ERA-2SM. So on this rainy Saturday, I did paint a wideband amplifier, printed it to foil, etched the board (with some other test structures) drilled the vias to GND 0.3mm, soldered the GND connections with wrapping wire and populated the board.

<
formatting link
> is the circuit, next picture to the right is the layout, and then the board.

Sorry for the soldering, these cell phone macro photos are so unforgiving. Probably it's a good thing that they don't let me solder space flight boards, I must call the qualified solder ladies for every change. :-)

Next picture is mag_db(S21), aka gain in blue and the input impedance in Ohms (red). It seems that in the left half of the plot, where the ERA2sm still works, the input impedance wiggles around 50 Ohms. The network analyzer calculates Zin from the measured S11.

The next picture is S11, and every curl is a resonance. It is more than probable that the ERA-2 is not the guilty party, apart of its input capacitance. As short as the input microstrip is, just holding the hand over it without touching has visible influence.

The last picture is the Agilent 54754A TDR. The outer side of the board's SMA barrel is exactly in the middle of the screen. The ringing just left of the middle is the port saver.

The strip lines have been calculated for 0.8 mm FR4, but I only had 0.5mm thin FR4 left over, so the line impedance will be somewhat low. A 200 mV step may be a little bit too much for a chip with gain that lives on 3.5V.

Have a nice Sunday, Gerhard

How? Since the post that you are responding to was directed to DLUNU and you've carefully snipped out the headers that would have made that clear, you are merely trying to be clever, and falling short (not for the first time).

Smart enough to notice when you try to be sneaky. Which doesn't take much intelligence at all.

Nobody in their right mind would take your advice seriously. It's there to be jeered at, and I'm happy to do the jeering. You deserve it.

Bill Sloman, Sydney

Nice work. Your S11 and TDR are inconsistent with what I'm seeing on the GVA part.

I woke up at 3AM [1] with a revelation, which may actually be true. There are two common MMIC types around:

1) One of them is the classic Darlington, like ERA-x. It needs a resistor from Vcc to the output pin to set the operating current. My experience is that they behave as expected, with the low-frequency response limited by the external coupling caps. Your TDR does suggest a Zin below 50 ohms. 2) The newer type, like the GVA-63, includes some sort of current regulating circuit, which allows a hard DC supply to be applied to the output pin through an inductor. I conjecture that this current regulator is slow, so it affects the low frequency gain and impedances. That aligns with Phil's observation that MMICS have bad low-frequency gain, which I thought strange because I hadn't seen that before. It also explains my weird TDR and impedance measurements, ditto not seen on ERA-type parts.

Some MMICS are rated for operation from "DC to xxx GHz" and some from maybe "10 MHz to xxx GHz" which suggests the slow regulator inside. This idea is complicated by the very loose way the microwave people use the term "DC", which in one case that I know of, is only valid above 100 MHz.

So I need to be careful to use the old kind in truly wideband circuits. We have some old-style Darlington SOT89 MMICs on the way. Luckily, my PCB layout looks OK, if I just use the right MMIC, so we'll go ahead and fab some boards.

I want Spice models!

[1] I know other people who also occasionally wake up around 3AM with new ideas. Or just get hungry.
John Larkin Highland Technology, Inc lunatic fringe electronics

Nice to know I'm not the only one who does this - and at precisely the same time, too! And not always about electronics. More often than not it'll be about someone's health that's endangered and I've come up with a solution to save their life.

Create some, then. It's good fun! And useful to others as well as yourself.

"The BEST Deal is NO DEAL"

There are several dozen people alive and well today who would otherwise be dead thanks to them acting on my advice. What have YOU got to show for the time you have spent helping others? (which I'm guessing is precisely nil).

"The BEST Deal is NO DEAL"

That sounds more worthy than MMIC ideas.

Are you an MD? I talk to my primary MD about problem solving. My puzzles are electronic and hers are medical. My advantage is that I can blow things up - sometimes deliberately - without serious consequences.

There is no medical (or mechanical, or hydraulic) equivalent to a scope probe.

I've seen the theory that pre-electricity people tend to go to sleep at dusk, wake up at dawn, but also get up for a while around 3AM, to stoke the fire or make love or something.

Being fundamentally lazy, I do a lot of design unconsciously, pushed into some invisible co-processor. I get answers delivered at 3AM sometimes, but more often in the shower in the morning.

John Larkin Highland Technology, Inc lunatic fringe electronics

I'm not much of a people person, but I can donate money to help people without the need for wet-ware contact.

formatting link

I decided that there were few things I could be less empathic about, so that's what I should do.

I've funded over 100 surgeries by now. Well, I get some very nice tee shirts too.

John Larkin Highland Technology, Inc lunatic fringe electronics

Certainly more worthy than the MMIC ideas I've spouted elsewhere on this thread, for which I apologise! I was attempting to describe something somewhat different.

Heck no. I couldn't do their job. If I'd gone into medicine it would only have been purely research with the aim to cure certain diseases. I can't do the inter-personal hands-on bit but greatly admire those who can.

I get a lot of men (overwhelmingly) who don't like to go to their doctor for some reason. I suspect it's fear of some sort on their part. Anyway, they'll tell me what's been bothering them and at the time it'll go in one ear and out the other and I'll think no more about until I wake up at 3am the next night or two having put the puzzle together subconsciously and with no deliberate effort. I'm a pretty good diagnostician whatever the subject matter may be.

"The BEST Deal is NO DEAL"

wouldn't a bead on the emitters' current source be just as effective? If you're feeding a high-level differential signal, it'll just damp a little of the common-mode response without having much effect otherwise. You're driving the differential mode, THAT shouldn't need any small tweaks.

I plan for a bead in series with the common tail-current sink (a fast NPN), and resistors or beads in each gate. I just hate it when things oscillate, so the emitter caps might be a useful option.

John Larkin Highland Technology, Inc lunatic fringe electronics

Exactly - the latter can usually be spotted by the low frequency roll-off.

However in addition to that, some MMICs (probably both those kinds) have internal matching circuits and some don't.

I wish the manufacturers would be more explicit.

Clifford Heath.

Gerhard does indeed do excellent work. Allow me to also compliment you for the unequivocal manner in which you embrace your insomnia. You're both inspirational to me, in different ways. :)

Thank you,

Don Kuenz KB7RPU There was a young lady named Bright Whose speed was far faster than light; She set out one day In a relative way And returned on the previous night.

Oscillators don't; amplifiers do. :)

"The BEST Deal is NO DEAL"

It's always worthwhile to put in a bit of belt-and-suspenders when doing that sort of thing. That way it looks like we knew what we were doing. ;)

I use a fair number of zero-ohm jumpers for that. They also help during bringup, especially the ones in series with the supplies.

Cheers

Phil Hobbs

Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC / Hobbs ElectroOptics Optics, Electro-optics, Photonics, Analog Electronics Briarcliff Manor NY 10510 http://electrooptical.net http://hobbs-eo.com

I didn't know about the built-in regulator--all the MMICs in my drawer are the older kind that need a resistor. It's mostly the internal matching, if any, that screws me up.

Cheers

Phil Hobbs

Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC / Hobbs ElectroOptics Optics, Electro-optics, Photonics, Analog Electronics Briarcliff Manor NY 10510 http://electrooptical.net http://hobbs-eo.com

Depends on the mechanism. Being able to control the phase of the coupling between the two transistors could turn out super useful.

I have some of those nice BLM15BA0050SN1 beads on order--check out the Z(f) curve here:

They're still pretty reactive up to ~700 MHz, but they're awesome above that.

Cheers

Phil Hobbs

Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC / Hobbs ElectroOptics Optics, Electro-optics, Photonics, Analog Electronics Briarcliff Manor NY 10510 http://electrooptical.net http://hobbs-eo.com

NT makes the same kind of claim. Every now and again a health guru goes to prison because somebody dies after taking their advice rather than going to somebody with proper medical training.

Giving people advice when you don't know what you are talking about is a dangerous habit - anything beyond "you should talk about that with the your doctor" is decidedly unwise.

Absolutely nothing, which follows the New Testament advice (Matthew 6.1) not that I expect to be rewarded in Heaven, since I don't think that there is one. I don't make a habit of seeking out other people to help, but I do help other people from time to time, and don't make a fuss about it.

Your guess about how I might spend my time is a malicious speculation, and not the first you've come up with. You do come across as a nasty piece of work.

Bill Sloman, Sydney

I can't accept that, Bill. I haven't said anything to warrant that description. I reckon there's something else that's got your knickers in a twist. Time will tell...In the mean time, the golf course beckons, don't you think? .tick-tock, tick-tock...

"The BEST Deal is NO DEAL"

Your social perceptions aren't up to much then, and it's very unlikely that giving you a detailed example would get through to you.

Of course you do. Psychopaths always find it hard to realise why other people find them obnoxious.

Bill Sloman, Sydney

Well, I've had you down for a few things on this group, Bill, but I never had you down for a snowflake. Dishing out all that supercilious condescension and yet you can't take a little polite criticism in return. Fancy having to tip-toe around the senstitivies you've never shown before. Hmmm. Nope. I don't buy it.

"The BEST Deal is NO DEAL"

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required