EEVblog
Вставка
- Опубліковано 10 лис 2024
- Part 2 in designing a better uCurrent.
A datasheet shootout between the Maxim MAX4239 and the TI OPA189
A deeper dive into the datasheet specs from the previous video, looking at gain bandwidth product, noise, supply current and voltages, offset voltages histograms, input bias currents, recovery and settling time, and slew rates.
www.ti.com/lit...
datasheets.max...
Cascading opamps tutorial: • EEVblog #572 - Cascadi...
Opamp Noise tutorial: • EEVblog #528 - Opamp I...
Forum: www.eevblog.co...
#Opamp #Datasheet #Tutorial
EEVblog Main Web Site: www.eevblog.com
The 2nd EEVblog Channel: / eevblog2
Support the EEVblog through Patreon!
/ eevblog
AliExpress Affiliate: s.click.aliexpr...
Buy anything through that link and Dave gets a commission at no cost to you.
Donate With Bitcoin & Other Crypto Currencies!
www.eevblog.co...
T-Shirts: teespring.com/s...
Dave, I don't know if you noticed but the TI part showing 3.0uV at 25C was listed as being at +-18V supply, while the Maxim part showed 2.0uV at 25C with a 5.5V supply. Isn't this actually a tighter tolerance for the given supply voltage from TI? Does the max offset scale up with increasing voltage? If it does, then linearly scaling the TI voltage vs max offset to 5.5V supply yields a 0.98uV maximum offset, more than twice as good as the Maxim part. I'm sure the value doesn't scale linearly, if it does scale with voltage, but I'm guessing you've got at least an equivalent spec between the two for the maximum offset voltage.
I'm also curious to find out if the 70pA vs 6pA input curent actually makes any difference when you sample and compare the two op amps.
Great video. It's nice that you documented what seems to be a great method for comparison of op amps. Thanks!
It shouldn't scale up, as offset is an input parameter. But it might change slightly with supply voltage though, that will depend upon the architecture and process node stuff. But I don't believe you can make any sort of extrapolation like that. I have never actually tried to measure that particular aspect though.
@@EEVblog Test the 189 offset at 5V supply voltage?
If you think about a standard npn diff any mismatch in the pull up resistor will show up as an offset and the offset itself will get worst with higher supply voltages, devices have finite output impedance -> higher supply voltage results in higher current which will result in a large voltage drop across the resistors and make the offset worst. Well at least for this simple case.
@studer1679 I came to search this from comments. Good dave has ack it
The answer is in the figure 7-7 in the datasheet. The supply voltage does influence it slightly, it is about 0.75 uV over whole supply range, 1.5 uV max. So still very good.
Seriously can't wait, I love it when you cover this sort of "deepdive" stuff...cheers.
You mean cheese 😉
But when head is on bottom it seems more naturally. Because speaking head without a body scares a little.
@Volodymyr Zakolodyazhny, agreed
Awesome videos! I am doing a PhD in experimental Physics and I do a lot of electronic designs for my group and I can tell you that I have learnt most of it from your videos over the years :)
Awesome Dave. This is incredibly useful for me. I am designing a current sensor actually and these are super relavent. I am measureing switch current, which is not as low as microamps but it is very high di/dt slew rate.
Floating Dave is the best idea of 2020
LOL
All hail the benevolent floating above Dave from down under.
But it's "Floating Dave Head" Mk II because it's Flipped. :)
it would be even better if it moved at different positions along the video
@@msana4420 I'm wondering why his head isn't upside down for me even though he's in Australia.
The input current of cmos op-amps is primarily due to the input ESD protection diodes, not gate current. The 189 has a much higher supply voltage, thus higher leakage current,
I have a design video (optical oscilloscope probe) use the dual 2189
The 189 is a chopper so at room temp it’s dominated by the switching current. But yeah the ESD leakage will dominate as the temperature increases. Roughly doubling every 10 degrees C.
Hey Dave!!!
I’m playing catch-up with your UA-cam videos. Your videos are really great for us semi-pro’s that are trying to relearn electronics.
I had to google, “Bob’s Ur Uncle” along with other phrases. No worries now!
Thumbs up on the mailbag and multimeter tear down videos.
I started watching EEVBLOG myself when getting back into electronic design a couple of years ago. It was the best thing I did to help get myself up to speed after a nearly 30 year hiatus. Dave is one of the best electronics tutors/mentors on UA-cam. Thumbs WAY UP to EEVBlog.
Thank you for that specific content!!! Unfortunately some people don't interesting, but I see it 2-5 times!!! I'm very appreciate such a adventure into a datasheets!
I really liked the comparison via the datasheets - I'd love to see more of this.
Great! Really looking forward to the follow up...
I wish I was your student. The enthusiasm is what I like here. Thanks Dave
Floating Dave is a nice new experience for viewers. Please don't use it too often, though.
The noise figures mostly correlate with the bias currents, i.e. low noise high bias and vice versa. That was already the case with the old choppers from Intersil, ICL7650 and ICL7652.
Less noisy devices use FETs with bigger silicon area, usually.
See also explanation in AoE about the design of the hp34420A nV DMM, and its ultra large input FETs, which normally have several nA of bias current..
To answer the question posed by a previous video, I’ve been seeing a lot more midroll ads, and you can tell which ones are placed by the UA-cam algorithm because they perfectly interrupt the video at the worst time.
There should not see any midroll ads in this video, it is disabled.
@@EEVblog I'm glad you turned them off. I've seen some videos recently that were just over 40 minutes long with 8, yes 8, annoying and unrelated ad breaks jammed into them by UA-cam. I'm going to give those guys a little while to begin disabling them, but if they don't I may be unsubscribing their channels.
I like the floating Dave concept. Very creative...
Now I want a floating Dave on an UFO.
It greatly saddened me when your head descended from the heavens to the pits of Tartarus. 16:16
Looking forward to future video comparing bench performance of chips.
Floating dave is the best dave :) I really like the idea of the flipped picture!
Btw I saw a Delorean on a sunday joyride yesterday.... beautiful car :D
Another exciting video (not a joke). Looking forward to new uCurrent (series)
Floating Dave head is iconic!
I just want a Dave Head flying around the screen 😂
I'm sure that's scriptable...
@@EEVblog I want to see a davehead screen saver bounce around on an oscope.
Like a screensaver Dave head 🤣🤣
I have no idea why anybody would thumbs down this video.
Serial haters.
Thanks for the video, great explanation!
Definitely Dave head up the top where possible.
Thank you flying spaghetti monster for bringing back the floating Dave head.
Floating Dave is very important. Visual is just as important as audio.
I'm definitely a fan of this reversed 'floating head' setup; it's much more engaging. Perhaps wearing a green shirt to key out your shoulders would be even better? I worry it'd look a little too weird, but maybe worth trying out methinks!
I suspect it would look too weird.
@@EEVblog You should do it just for the lolz
Just one thing I pick up maybe you overlooked. The offset of opa189 is lower at lower supply voltage. This is very critical in the comparison. Maxim one can't work at higher voltages.
The "Zero-drift" is of course just a marketing name, they do actually have some drift (and yes, it was specified in the TI's datasheet), but of course much less than without the chopper;)
waiting for part 3 :)
I would say OP189 is better. It is better to look at the distribution, than the typical / max values, because various manufacturers do define these things differently.
Waiting for this video
Dave, I definitely prefer it down the bottom.
I think I speak for all of us when I say we like our Down Under Dave.
From my POV ,he's actually Up Slightly Further Dave.
I was expecting to see a Linear datasheet and was surprised instead
I have Dave's head set up as peppers ghost to correct the back to front Dave. "Floating Dave is spooky at a distance"
There is one other parameter that wasn't mentioned. That is the product life. When was it released and for how long will it still be available. If you are making a product with a planned 10 year life cycle you don't want to select a part that the manufacturer will stop making in 2 years.
Dave you should pull a pull-down resistor on your image, you seem to be floating at the top of my screen!
Dave, did you have a video explaining all these criterias in the datasheet? e.g. CMRR and other stuff?
First DaveCAD and now DaveHED, what next.
Thanks
Dave should follow a different relatively simple function each time, starting with Sin(t)
Take great care when the spec of a chip is in pA input bias/offset. Those current levels are VERY hard in an actual circuit implementation. Especially on tiny packages like SOT-23, where the clpackage surface path is very small supply to inputs can have significant surface leakages from supply or ground connections with real world moisture and dirt. In perfect clean controlled lab test maybe 1pA but real world I doubt it. Just look up the setup required when measuring such current levels. Reference for instance Keithly Electrometer guides for such low current measuring.
I regard the TI data for OPA189 to be quite realistic and regard the other 1pA figure a very theoretical value under ideal/perfect conditions.
It specifically states how difficult it is to test it as the input bias is dominated by other factors beyond the opamp itself.
@@rich1051414 I agree they state that, my point is that the data given is highly impractical to achieve and the Ti data is more believable to consider for actual functional circuits. The very low pA values depend on too many other sources to be of any practical use. To actually measure in that range uses tall glass supports or air routed paths for example. Nothing like that is possible for a chip in SOT23 package. I believe to select on those parameters will allow wrong expected performance and rejecting supperior parts.
Maybe TI is specing it at something they can test and guarantee unlike Maxim's hand waving.
Why is bandwidth so import for a shunt amplifier?
Dave says “of course bandwidth is one of the important criteria of a shunt amplifier”
I think of it as amplifying a DC (or very near DC) voltage.
I'd be interested to know this too
steveh4044, after I wrote my question I thought about it more.. and maybe it is important for tracking large and sudden changes in current, if you are also measuring the voltage output from a uCurrent for example, on a scope. So you don’t get an incorrect impression about the turn on/off characteristics or other events, or to see current spikes accurately for example.
Ah, I just realised.. a current shunt in a lab power supply (needed for current limiting the output), would need a very high bandwidth current shunt amplifier to respond immediately to sudden current draws (like shorts).
Other people can explain better and more fully perhaps, but I think I answered it correctly...
@@davadoff Thanks, that makes sense.
Dave I have slightly irrelevant question, what pdf viewer do you use? I use sumatra only complaint about it is that it doesn't support multi-window mode
Just Adobe. Drawboard PDF for this stuff as I can write on it.
Don't understand the ADC comparison of % in the recovery time... How does ADC precision compare to recovery time?
how about a floating Dave that traverses the screen and bounces off the edges?
If you wore a matching green T Shirt, then the floating head would look a lot better.
I was just about to build some low noise ucurrents, crazy!
Here is a stupid question;
Why can't you use a FET transistor stage(s) to amplify to millivolts, then use a cheap-ass OP AMP with the right bandwidth for the job, for the output stage?
Remember, If I knew, I wouldn't ask!
I like Dave at the corner and text big, I can't read the datasheet without going full screen!
25:34 is he going to talk about price, is he? (I used these TI parts before and they aren't giving this sort of performance away free) ;)
💖💖💖
How does the Analog Devices ADA4523-1 compare?
I guess it's best to list the pros and cons in order of importance.
Nice Video!
Could you do something to the green screen reflection? This greenish feel is a little bit odd.
Yep, poor lighting in the current cubicle
I love floating Dave head, he should move around on animated chicken legs every now and then for some variety, or something more amusing to you. There's enough geeks in your audience to help make that happen!
Would you guys use some of these for audio? OPA2134 , ne5532 are very popular, how about these?
Opa189/2189 can be ok. Max has too low maximum supply voltage.
G'day Dave, just wondering if you could do a similar video on how to find a modern equivalent for Schottky Barrier Diode 5KQ30 - Nihon Component, having a hell of a time in finding a replacement.
Floating floats my boat :)
dave head works where ever you cover the least data
Vote for Dave head at the bottom. (Bottom right corner, non flipped).
What about price and availability
floaty dave is the bees knees mate
i'll take Floating Dave Head anywhere on the screen frankly, doesn't matter to me.
I prefer without a floating Dave over the datasheet.
Dave needs to flip up side down under
well my goto precision part would be the lt2052 but anyways
20:37 Sure... At Unity Gain.
Hai this is dalwinder s sidhu nice to watch you on youtube on op amps can you please discover that there is phase error in op amp if both terminals of op amp are fed with same signsl please find out is it true or false thanks
Floating dave's head is like Dick smith's head :D
I feel like I'm auditing tax returns with some datasheets
Dave head is great lol
Top floating Dave head is a nice feature :)
So does that mean your head is now on backwards in our pic, Dave?
:-/
Not backwards, inside out.
Why do we even need dave head to begin with ?
Hmm... isn't reversed Floating Dave® actually Evil Floating Dave®? It is Floating Daves® polar opposite and normal Floating Dave® is not evil..
I prefer Dave head wherever it's least obtrusive.
But all else being equal floating is better.
could you design uCurrent PLATINUM? MAYBE..?
Don't know what to call it, Platinum did cross my mind. Suggestions invited.
@@EEVblog gold uCurrent update!!
@@EEVblog How about a nanoCurrent? Or uCurrent+.
Floating mirrored Dave head FTW!
Can we get a floating Dave head that bounces around the screen and like the old DVD screensaver?
ua-cam.com/video/5mGuCdlCcNM/v-deo.html
Floating head in the centre seems odd .. bottom left or right please.
Nah I like it in the center personally
I like this novelty, middle bottom works best, aesthetically speaking. I like floating Dave but not head-on-door Dave. The only way I can see it working on the top, is hanging upside down like a bat, but that’s my hang up not yours! Change it up, keep the logic fuzzy and analog!
#FloatingPointDave
hello, my television has got a problem. when i change channels before normal sound there is a voice pıt from one speaker of two ? thank you......
Off Topic: Can you Expose(?) EMROD:
*NZ to trial world-first commercial long range, wireless power transmission*
newatlas.com/energy/long-range-wireless-power-transmission-new-zealand-emrod/
"Emrod currently has a working prototype of its device, but will build another for Powerco, with plans to deliver by October, then spend several months in lab testing before moving to a field trial. The prototype device will be capable of delivering "only a few kilowatts" of power, but can easily be scaled up. "We can use the exact same technology to transmit 100 times more power over much longer distances," said Emrod founder and serial entrepreneur Greg Kushnir. "Wireless systems using Emrod technology can transmit any amount of power current wired solutions transmit." "
*Greg Kushnir of EMROD:*
ua-cam.com/video/4IvimvUVxkI/v-deo.html
Hello thrte am shane i ask about catalog am starting over i lost everything in apartment fair 4.24. This year. I do it for hobby went. Two School on it I really like watching on UA-cam
How people suppose to make there money back if they spend $6000 on a profesional oscilloscope?
Huge investment for a beginner to make cash back or make a living fixing junk they find in a dumpster
What does that have to do with this video?
And who said you need a $6000 oscilloscope?
Clearly this is not related to the current video, but who expects to make their money back on the expense. Main reason to spend such amounts or even far more is some performance aspects of the 'Professional' equipment. Say for example 20GS/s and over 4GHz bandwidth. Some just need the higher spec performance or can not do the job required. Even companies need to spend over $200K if they must have 80GHz bandwidth because they work in such frequency domains. Actual payback is not in actual $ terms but simply being able to get the required work done. If you simply want to earn a few dollars to pay for equipment you can not afford professional level gear and stick to 'maybe' measuring results.