Old parts that blow the doors off newer ones

In the "Replacement for 741" thread, this came up, which seems like it deserves its own thread.

> Yeah, todays opamps 'blow the doors off' any older stuff. > I don't know comparators as well... oh there was an LT part.. > LT1016, I think. That's a nice snappy comparator. (not cheap) > > Oh to the OP as far as parts not in DIP's. Your best best is to buy > some of these little pcb's that do soic to dip. A buck or so each on > Digikey I think. > > George H. >

One glaring exception that I'm having my face rubbed in at the moment is the surpassing excellence of the late lamented LF357. It has very very low input capacitance, lowish noise for 1976 or whenever it came out, and 20 MHz of bandwidth.

None of the >30V op amps I've tried has been able to touch it as a low-level TIA for use with a bootstrapped PD and common-base stage, which is one of the examples from my Front Ends chapter (the very last thing to wrap up before shipping the book off).

The circuit uses a 301k // 0.4 pF feedback network with a BFU520A common-base stage and an MPSA18 bootstrap. (There are MPSA18 equivalents that you can still get.) The problem is the Johnson noise of the 301k resistor, differentiated by the op amp's input capacitance.

It works beautifully with LF357s from my stash, but fails miserably with: OPA140, OPA172, OPA604A, TLE2081, LT1122,

all of which ought to have worked if their datasheet Cin specs bore any relation to reality. (Spoiler: they don't.)

Friday I'll try it with reduced supply voltage and an ADA4817-1. It really frosts my shorts to give away 10 dB of dynamic range (+-5 vs

+-15), even while paying 10x more for the amplifier.

I'll update this when I have more data. In the mean time,

Got any more examples of (electronic) stuff you used to do but can't anymore on account of obsolete parts? You can (sort of) get 5 GHz PNPs again, so that one is (sort of) fixed, but there must be lots of others.

Cheers

Phil Hobbs

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

What's Cin on the LF357? (and there's always two Cin numbers on a spec sheet... I always choose the larger.)

I like the opa192, but I've never 'checked' the Cin spec.

1.6 pF differential 6.4 pF common mode
Reply to
George Herold

About 2-3 pF, based on the corner frequency of the noise.

Those datasheet numbers are pretty well worthless--they appear to be made on unpowered devices. JL and I have both tried to get the semis to tell us how that measurement is done, but they never do. (Probably the apps guys have no idea.)

I'm going to have to build a C_in measuring gizmo.

Cheers

Phil Hobbs

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

Mesfets, like CLY2.

Avago SOT-89 phemts, of course.

Most anything SOT-89!

BFG25

Clamp-amps. Nobody makes them any more.

Basically all the NEC fast parts are gone

Who makes the fast PNPs?

Shockingly, the MC68332 is still available.

Reply to
John Larkin

Huh.. but look mine have two digits.. more precision that has to mean better. :^)

Unpowered. That makes sense. (when I wanted to measure the C of an led I just stuck it unpowered onto the lcr bridge.)

Hey, here's an idea, measure the noise corner freq, with a big noisy feed back R. :^)

Does comparing the noise (from say 1 meg R) with two opamp configurations help? A follower and/or a TIA, ground the resistor or the opamp.

GH

Reply to
George Herold

Renesas has some 12-GHz class pHEMTs. I have some but haven't tested them yet.

NTE2403. They might have bought 20 reels of BFT92s and relabelled them. ;)

They're much more expensive, of course, but if that means it's economic to produce them, I'm glad to have the design space back. (We have plenty for our own use, but that isn't much help in the licensing area.)

I'm mostly asking about what design possibilities have disappeared as the parts go obsolete. My 70 dB, 100 MHz, single stage iso amp (MRF966 and one resistor) is another one I miss. I could still build one, of course, but it would take probably 6 or 8 parts instead of 2.

Cheers

Phil Hobbs

(proud owner of 2 reels of BFT92s among other nice things)

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

Not declared obsolete yet but long lead times reported on this product:

Reply to
bitrex

The Avago phemts would switch an amp really fast with 0.8 volts of gate drive, namely an EL or EP gate. Nice for small laser drivers.

The EPC GaN fets can do the same or better, but need volts of gate swing, which is a nuisance to do fast.

Reply to
John Larkin

deserves

o buy

ach on

s the

y low input

of

level TIA

he

ore

n-base

you can still

erentiated by

h:

relation

It really

n while

ymore on

, so that one

Ahh the LF357s. In my case the National LF356. I lucked out and got some pre production engineering samples ca 1973 thanks to Dennis Bohn. Was doing s ome low noise audio stuff.

Reply to
gray_wolf

Yeah, we went round on that last time it came up. There's something weird about the process, so they mostly build to order.

Cheers

Phil Hobbs

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

The AD8066 (8065 single, I use the dual) is expensive but may be worth a shot. +/-12V (can't think of another with that spec) and another weirdness is the input current over common mode voltage (looks like they have plain N jfets at the inputs and as the gate-source opens they are no longer FETs, they do characterize some work in that range but I have never gone there in my designs).

Dimiter

------------------------------------------------------ Dimiter Popoff, TGI

formatting link

------------------------------------------------------

formatting link

Reply to
Dimiter_Popoff

The 356 is still in full production. It's too slow for this job, and also seems to have more input capacitance for some reason.

Cheers

Phil Hobbs

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

Re obsolete parts: I miss film caps, more hi tech, the death of CDs means the 785nm laser diode (from Sanyo) is gone. Nice narrow laser and could be temperature tuned to the Rb lines.

George H.

Reply to
George Herold

LM394, CA3080, long gone , so how do you make log sweepable (3 decade range) analogue function generators now ?

MK

Reply to
Michael Kellett

The LM13700 isn't going anywhere anytime soon, and the CA3086 is still available from Chinese suppliers:

Reply to
bitrex

Or if you need better matching than the CA3086:

Reply to
bitrex

MAT14. Same die as the MAT04, but with most of the testing omitted. Of course you can get a nice MCU and 16-bit DAC for the same price. ;)

For noise cancellers I'm also using a trick: making the die temperatures of BFP650 diff pairs track by servoing V_CE of one side so that the instantaneous dissipations are the same. (Its Early voltage is big enough that you can do that.)

As somebody pointed out, you can do it with OTAs as well. The LT1228 is pretty cool for a lot of uses if you don't care that it's noisy.

Cheers

Phil Hobbs

Reply to
Phil Hobbs

Wish someone would de-cap one and find out what the output stage actually is. One of life's little mysteries...

Reply to
bitrex

Send it to the Zeptobars guy.

Cheers

Phil Hobbs

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

Magnetic Amplifiers. Good small square-loop ferrites are unobtainium nowadays. 14mm OD is the smallest I can get.

Best regards, Piotr

Reply to
Piotr Wyderski

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.