Bootstrap help (PD-TIA)

Sounds like your bootstrap's output impedance is too high.

If you don't care too much about getting the lowest noise, you can make a Darlington-style bootstrap by putting a CPH3910 in front of a

I often use a JFET/PNP series-feedback pair for this. You put about 600 ohms in series with the JFET's drain, and wrap a PNP around it, as in AoE3 Fig 2.92 with R5 = 0, Q1 = CPH3910, and Q2 = BFT92. (The BFT92 is EOL, but I have a couple of reels.) The idea is that Q2 tries to keep Q1's drain current constant. It works best when Q2 isn't running much quiescent collector current, maybe 2 mA. (It's different from a Sziklai pair in that the first stage takes most of the current.)

It might need some tweaking to get it to be stable with a much slower PNP--maybe a bead in the base. You can get fancier, e.g. by putting a diode in series with the drain and sense resistor, and a parallel RC in the emitter, and wring a few more decibels out of it.

The other thing that helps a lot is to use a stiff tail current source.

We sell and license a proprietary version of the White cathode follower idea--see .

Cheers

Phil Hobbs

I use a JFET with BJT, e.g., AoE3, Figure 8.83, page 548, and Figure 4x.25. Or see my RIS-617 designs, with my preferred PNP, see schematics,

I think that's right. I had to reduce the value of the bias resistor between PD anode and negative power supply...

OK thanks I can give that a try. (I don't know enough about jfets...)

Hmm sounds trickier, I'm not sure I understand it... which always makes it harder to get working. :^)

Nice... I'm thinking out of my price range.

Say, maybe if I put a bit of inductance in the line between PD anode and the negative bias supply? That may reduce the loading on my bootstrap.

Hey thanks, I've got a few things to try.

George H.

Right thanks. I saw that but didn't know why it was there.

Ahh and look an inductor in the PD bias line. :^)

George H.

Oh dear, this 'final' project just withered on the vine. I was fired today... (expected.) I'm off for a week at cape cod, looking after my mother-in-law who has some crappy neurological disease. And then I need a new gig.

George H.

Bootstrapping a photodiode over a cable gets interesting as the cable gets longer. I had to explain that to a customer who elected to not manage his cabling to his photodiode.

My first bootstrap was the noise of a 100k ohm resistor, down ~2 feet of cable and probe, to 77 K.

George H.

Huge fun for all concerned. :(

Send me your home email--I'm happy to be a reference.

Cheers

Phil Hobbs

Thanks, SED is my best/favorite collaborator currently. Which is nice, but also stinks.

snipped-for-privacy@gmail.com

I'm going to cape cod for a week. George stares at ocean and thinks about life, nothing new. :^)

GH

Next time go to Cape May, so you can swing by Burg Frankenstein here for a beer en route.

Cheers

Phil Hobbs

Brilliant idea on their part ... NOT !!! But a chance for you to branch out anew. I'll be glad to give a good recommendation.

Hi George, similar feeling here too, after years of plowing along steadily this seems to be a year of change.

piglet

Works fine as long as the cable is short compared to system rise time.

That particular customer liked to poke loose photodiode wires in a big pipe at random.

Yeah, mostly I feel relieved. I torpedoed this one new hire, and so I've been expecting it.

Let me share this piece of the severance letter.

"That Mr. Herold agrees not to share plans, designs, electronic circuits or any other Teachspin property with any commercial company or educational institution for a period of 20 years."

I have no idea what this means, but it seems crazy to sign on to such terms. (Property would seem to imply any circuit scribble I've done at work.)

I love the twenty years part. :^)

George H.

Hi piglet, Come on in, the water's fine. :^)

George H.

Thanks Phil, If it wasn't such a long drive I'd swing by on my way home.

George H.

You can agree to not share accepted-definition trade secrets, presumably in return for something of value to you. You have a legal right to "practise your trade" which includes doing normal stuff.

I wouldn't sign anything like that unless a giant severance is included.

I really enjoyed being fired. Got a nice fresh start on life. Unemployment and severance financed some shopping around for six months.

Maybe you can do some serious scientific instrumentation next, not the educational things.

The maximum term should be no more than 3 years.

But you SHOULD NOT sign such a vague document, especially one calling out "electronic circuits".

If something is to be protected, the protected item(s) need to be spelled out very clearly, otherwise your entire engineering career can be put at risk. For example, you know how to design a specific high-performance amplifier. Maybe they have one like that in a product, maybe you designed it, maybe not. But you should not be prevented from creating a similar, or even nearly identical, design in the future.

The company has various educational products. You can agree not to divulged relevant details of those specific products. But say a product has a circuit you're familiar with, an amplifier, etc., that's part of your engineering toolkit, it cannot be magically included with vague language.

What you think it implies is, of course, irrelevant!

As written, it can easily be interpreted to mean that plans/designs/circuits you develop after leaving cannot be shared with your new clients. (Ridiculous, maybe not enforceable).

Ditto Teachspin property, but /that/ isn't ridiculous.

In the UK you could amend that letter to be something to your liking, sign and date the amendment, and return that.

Or, as I once did w.r.t. a patent application, edit the PDF to insert a word such as "reasonable" in the relevant point.

Restraint of trade.