Author Topic: EEVblog 1464 - TOP 5 Jellybean Comparators  (Read 3151 times)

0 Members and 1 Guest are viewing this topic.

Offline EEVblog

  • Administrator
  • *****
  • Posts: 34541
  • Country: au
    • EEVblog
EEVblog 1464 - TOP 5 Jellybean Comparators
« on: March 24, 2022, 09:27:35 pm »
The TOP 5 jellybean comparators, plus a bonus and special snowflake choice.



00:00 - Jellybean Comparators
01:20 - Traps for young players using Opamps as comparators
04:43 - Is the old School LM311 still THE jellybean?
11:17 - TS391 Small single comparator
16:33 - LM393/LM2903 Dual comparator
21:27 - LM339/LM2901 Quad comparator
22:52 - LMV331/LMV393/LMV339 Low voltage jellybean comparator
31:50 - TS3021 Fast precision rail-rail comparator
34:45 - TSM102 Special snowflake kitchen sink Opamp/Comparator/Reference
 

Online PartialDischarge

  • Super Contributor
  • ***
  • Posts: 1353
  • Country: 00
Re: EEVblog 1464 - TOP 5 Jellybean Comparators
« Reply #1 on: March 24, 2022, 09:41:09 pm »
I like the TSX393 from ST, only 5uA, but it comes in double package...
 

Online PartialDischarge

  • Super Contributor
  • ***
  • Posts: 1353
  • Country: 00
Re: EEVblog 1464 - TOP 5 Jellybean Comparators
« Reply #2 on: March 24, 2022, 09:58:08 pm »
As you mention high speed comparators(<5ns) are expensive and not widely available. From the design viewpoint I don’t really understand it, having a nonlinear output stage supposedly should make things easier and cheaper
 

Online Kleinstein

  • Super Contributor
  • ***
  • Posts: 11439
  • Country: de
Re: EEVblog 1464 - TOP 5 Jellybean Comparators
« Reply #3 on: March 24, 2022, 10:39:13 pm »
The LM311 is different from the LM393 / LM339: it is faster and takes more supply current and has the output transistor with a emitter terminal.
The TS391 / TL331 is kind of the equivalent single to the LM393 , though less common and not yellybean anymore in the strict sense.

The output of the LM311 is nice to have something like a +-15 V supply and than get the result relative to ground. Many other OPs have ouput switch to the negative supply.

High speed comparators are expensive because for high performance they need different processes (e.g. fast PNPs). A somewhat tricky combination is high speed and a large possible input differential.
There are quite a few fast comparators available, but most of them low voltage. A moderately priced is the TL3016.
Many high speed comparators also have extra functions, like latch, enable or similar. These extra functions make a direct exchange tricky. High speed may also no longer work with open collector and even mormal TTL may be on the slow side.
 
The following users thanked this post: EEVblog

Online Avelino Sampaio

  • Regular Contributor
  • *
  • Posts: 141
  • Country: br
Re: EEVblog 1464 - TOP 5 Jellybean Comparators
« Reply #4 on: March 25, 2022, 03:34:21 pm »
The LM319 join my list. It has twice the speed (80ns) of the LM311 (165ns) and costs just 1$ (SMD) and 3$ (DIP)
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 19610
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: EEVblog 1464 - TOP 5 Jellybean Comparators
« Reply #5 on: March 25, 2022, 04:32:39 pm »
Partial to MCP6562 myself.

As I recall, there wasn't much in the space between ancient LM393alikes (~ 200ns, 300uA) and the screaming-fast ones (<8ns), with the MCP and TS3021 being some of the few examples (~ 50ns, 100uA).  Hmm, I forget offhand what the fast ones typically draw for Icc.

Another cheat: LVDS receivers are often good comparators.  Sure they're only guaranteed for 100mV input offset, but that's just to scare you off.  Typical isn't going to be precision, but that can be acceptable in an auto-zero application, delta-sigma converter say.  Beware of Vicm; some only work in a narrow range (hardly any good for their intended digital purpose even!).  Want to say this is more of an issue with ASICs and FPGAs than dedicated receiver chips, but yeah, just to be sure.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline ormaaj

  • Contributor
  • Posts: 20
  • Country: us
Re: EEVblog 1464 - TOP 5 Jellybean Comparators
« Reply #6 on: March 25, 2022, 05:49:34 pm »
AD8561 is pretty close to a drop in upgrade to the LT1016 for about the same (high) price fwiw. You'd opt for something else if you don't absolutely need to replace a big ole PDIP though.
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 14632
  • Country: us
  • DavidH
Re: EEVblog 1464 - TOP 5 Jellybean Comparators
« Reply #7 on: March 25, 2022, 07:14:46 pm »
The LM393 and LM339 are the companion comparators to the LM358 and LM324 single supply operational amplifiers.  The LM392 is a 358 operational amplifier and 393 comparator in a dual operational amplifier package.

The 391 is suppose to be a single 393/339.

The comparator response time has nothing to do with output rise and fall time.  (1) A slow comparator like the 393/339 will have no problems driving fast logic.

High speed comparators are expensive because for high performance they need different processes (e.g. fast PNPs). A somewhat tricky combination is high speed and a large possible input differential.

The old 10 nanosecond NE529 and 5 nanosecond AM685 comparators are all NPN.  These were alternate sourced and might have been considered jellybean parts (for high speed comparators) before the LT1016 came out.  Any ECL differential line receiver is an all NPN fast comparator.

The LT1016 was a major improvement over older high speed comparators because it was stable in its linear range making it much easier to apply.  That strobe pin is not just there for logic strobing; it also allows the comparison to be made at a time when the comparator inputs are not crossing their linear range to prevent the comparator from oscillating.

As I recall, there wasn't much in the space between ancient LM393alikes (~ 200ns, 300uA) and the screaming-fast ones (<8ns), with the MCP and TS3021 being some of the few examples (~ 50ns, 100uA).  Hmm, I forget offhand what the fast ones typically draw for Icc.

I think just about everybody who second sourced the 311 back in the day had some 40 to 60 nanosecond comparators of their own, but they were never as popular as the slower and faster parts.  The comparator function was just too esoteric.  Faster ones like the NE529 and AM685 were more likely to be alternate sourced because they were so much more useful.

(1) Note that comparators have output *and* input slew rate limits.  The output slew rate produces the rise and fall time but the input slew rate limit can cause problems with high speed input signals.  See page 10 of the LT1011 datasheet which is an improved 311.
 

Offline ormaaj

  • Contributor
  • Posts: 20
  • Country: us
Re: EEVblog 1464 - TOP 5 Jellybean Comparators
« Reply #8 on: March 25, 2022, 08:34:12 pm »
Right LT1011 is another nice but expensive one :)
 

Offline thm_w

  • Super Contributor
  • ***
  • Posts: 3942
  • Country: ca
  • Non-expert
Re: EEVblog 1464 - TOP 5 Jellybean Comparators
« Reply #9 on: March 25, 2022, 11:02:18 pm »
Now you mention how many 339 are in stock at digikey, had to check some other parts:

Resistor - 0402 10k - 57 million
Diode - S1MB-13-F DIODE GEN PURP 1KV 1A SMB - 2 million
Microcontroller - PSOC CY8C20534-12PVXIT - 244k
Comparator - NCX2200GW, 125 IC COMPARATOR RRIO SINGLE - 150k

Resistors I get, but, who needs that many PSOCs?
 

Offline scopeman

  • Regular Contributor
  • *
  • Posts: 248
  • Country: us
Re: EEVblog 1464 - TOP 5 Jellybean Comparators
« Reply #10 on: April 02, 2022, 04:08:37 am »
Not Jellybean but I really like the LMC7215. Really low power and push pull.

Sam
W3OHM
W3OHM
 

Offline Doctorandus_P

  • Super Contributor
  • ***
  • Posts: 2302
  • Country: nl
Re: EEVblog 1464 - TOP 5 Jellybean Comparators
« Reply #11 on: May 14, 2022, 09:39:32 pm »
@eevblog
I think you made a little error here.

This page:
https://www.eevblog.com/2022/03/31/eevblog-1464-top-5-jellybean-comparators/

seems to link to the video for #1465 about inductor measurement with DMM's.
 
The following users thanked this post: EEVblog

Offline EEVblog

  • Administrator
  • *****
  • Posts: 34541
  • Country: au
    • EEVblog
Re: EEVblog 1464 - TOP 5 Jellybean Comparators
« Reply #12 on: May 18, 2022, 01:12:57 pm »
@eevblog
I think you made a little error here.

This page:
https://www.eevblog.com/2022/03/31/eevblog-1464-top-5-jellybean-comparators/

seems to link to the video for #1465 about inductor measurement with DMM's.

Oops, fixed, thanks.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf