Hello. Is there some widely accepted, cheap, common, multi-sourced modern drop-in replacement for the CA3240E op-amp, specifically improving speed and linearity while preserving single rail supply, high input and supply voltage (at least 30V wanted for single rail and input difference), FET input, unconditional stability (no compensation cap needed to prevent RF oscillation).
I considered the LT1215 but it's not exactly multisourced, and likely not very future-proof. It also needs that capacitor. I could live with that but the fewer parts and pins the better.
Phil Hobbs wrote in news: snipped-for-privacy@supernews.com:
Not single supply. Turns out LT1208 isn't either, though Digikey's filters found it because they think it is. Ò^O. So far, CA3140 works, and tomorrow I'll try LT1215 which looks like it will be better still. Single supply is critical, partly for headroom on low supply, partly for ease of clipping at ground/negative.
(I searched for rail-to-rail on Digikey too, but that narrows the field way too much. I only need the negative swing to go to ground.)
Phil Hobbs wrote in news: snipped-for-privacy@supernews.com:
I got something wrong too. LT1215 isn't FET input. (And ADA4627 isn't single rail, won't go to ground). LT1215 works great though, I tested it today. But if there are IC's more widely available, more futureproof, please tell me (I assume I'm not the only person trying to avoid being left in the lurch by rarity, high cost, or sudden early obsolescence).
The CA3140 still amazes my by being capable enough to do fairly good square waves at upwards of 100 KHz, but I think there must surely be an IC of drop- in caompatibility that can do them at 1 MHz by now. Something this basic seems oddly hard to find, Picking standards used to be a LOT easier!
When the 3140 came out, it was a big leap ahead. The average op-amp was the 741 which had pretty limited specs. There were better ones, but they cost more and were harder to get if you were a hobbyist, and some of them required some work to use properly.
The 3140 came along and it was a bit thing, really high input impedance, I cant' even remember the details. But about the same time, TI came out with their TL070 type devices, which were also a decent leap forward over the 741. The 3140 was not the same sort of device as the 070, but in a general way, they were.
Which means that you need to be specific. Lots of op-amps have come (and sometimes gone) in the time since the CA3140 came along. Some of what was luring about it back then is no longer so unique. Higher input impedance devices are out there, that TL070 for instance. Maybe not as high impedance as the 3140 (memory says it used a mosfet on the inputs, while the TL070 used jfets), but for many purposes the difference didn't matter. So if you wanted to make a sample and hold, the 3140 might still have a lead, but if you just wanted to use higher value resistors, likely many devices are suitable. The 741 had a gain-bandwidth of 1MHz, which looked good at first, and then when you needed gain the reality set in. Bumping it up to 10MHz, and that is where things went if I recall properly when the 3140 and the TI op-amp came along, actually was a big improvement at the time, likely still is for a lot of applications.
No, I have no idea what's available these days.
And if you're talking about higher frequency oscillators, invert the problem. Trying to stretch an audio device into low RF can be a problem, but there are plenty of devices that operate at a higher frequency that suddenly makes 100KHz or 1MHz low frequency. If you need a good square wave, start off with an actual RF oscillator, even an RC oscillator built around logic gates, and suddenly the limitations of the device start disappearing (until you push higher in frequency). Need a more "perfect" square wave? Then kick the oscillator higher in frequency and add a flip flop to divide it down, getting half the frequency but a better square wave.
Michael Black wrote in news: snipped-for-privacy@darkstar.example.net:
I think the TL070 and 080 and relations are nice, but not single rail. Many books aimed at hobbyists seemed to prefer the CA3140, maybe because in so many different situations, it just worked. I like the LF412 for same reason, but it's not as widely useful as CA3240.
Curious. CA3140 being unique, yet so generally useful. Maybe that's what's gone now, unique now seems to mean fit for a very specific task, where selection is impossible without very detailed knowledge of context. I like the CA3140 precisely because it so often works when that is missing. :) Probably also why books aimed at hobbyists use it too. Maybe the LT1215 could have earned a rep like that but not now, not when it's like picking the right oak in a forest.
3140 is jfet, btw.. Turns out I don't need the high resistance input, my circuit now seems better (faster) with lower resistance and more power. I'm using differential amps for arbitrary wave shapes up to at least a MHz, with gains equal to or less than unity. LT1215 isn't stable for this, so I'd like an amp that is, but otherwise equally good or better, and single rail. I will keep hunting... :) But any pointers to narrow it down will be nice.
Ok. Thanks though, intersting post to read and reply to. I've used the Digikey filters at times to help me search, I think this is how I found LT1215, but as they show me current parts, I have no way to know which ones have stamina. :) I've never placed a bet other than a lottery ticket, but I'm beginning to have a deep respect for those punters who really know the form book.
Not oscillators, but I like those ideas, I have a cheap signal generator I bought from China that has excellent easy dialing of any frequency from
0.01Hz to 5MHz, but some dodgy linearity and noise. A few digital gates can help with that, and I was especially considering a 4046 PLL to get me a clean triangle wave as well as square. (Triangle + comparator might do good pulse width changes too, the generator's inbuilt triangle is horrible). One thing I did once is make a crystal oscillator from NOT gates and a ripple counter, to get a 50 Hz time base for a clock module. Was my first PCB design...
Inverting the problem is a good idea too, I haven't looked at RF specific amps yet because I need good DC performance and single rail supply, but usability ao 1 MHz. It seems that fast audio or general purpose, even a very fast precision amp (like OPA2277 but faster) might get me there, if it used a single supply.
One way I can change the question radically, is what indispensible giants among opaps stride the great horizons these days? The ones everu audio hobbyist might know, IC's that seem immortal. CA3140 is a good candidate, but only good to 200 KHz or so for my current task. I'm aiming for designs that use standard IC's if possible, to guarantee that almost anyone wanting to build one can get the parts easily. If I knew about ten of the most current 'reached-for' op-amps for generally amaxing people with what versatile power they have, I can start testing the most likely of those, and take it from there. Catalogs like Maplin used to make thos a very easy task, even publishing comparison tables. It was like a distillation some engineers had made, and the choices were fairly good so far as I know, but that was at least 15 years ago, they might pick some different amps now...
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.