Logic seems to be made for ultra small SMD's , or is it the other way around. Low voltages with low currents in mm2 packages with tiny solder island on the bottom .
Amazing ! How can they solder this reliable to a PCB ? Unbelievable !
If only these manufacturers could name their packages the same way.
It's a jungle out there in names and sufixes !
Amazing ! How can they solder this reliable to a PCB ? Unbelievable !
If only these manufacturers could name their packages the same way.
It's a jungle out there in names and sufixes !
Thanks AIM65 . It's just a milled test PCB so it didn't matter that much but I had fun making it.
Simple set up : 4 oscillators with LVC17 buffers for nice rise and fall slopes . HC and LVC D-FF to divide the higher freq because my DMM could only go to 10 Mhz according to its specs , but it could measure 20 Mhz.
A HC393 for further division .
Then the HC an LVC gates to measure current draw at different frequencies and at different Vcc's , with signal in 1 or the 2 inputs ( or 3 for LVC10) . Threshold voltage & max current draw at the thresholds , that is why 2 of the trimmer are for. 2 regulators with trimmers for 1,25 to 3,6V for the LVC's and 125 Mhz oscillator. 2,5 to 5,6 V for HC and the LVC1G17's . Current measurement over a 1 ohm R , DMM measures down to 10uV , so 1mV = 1 mA , easy.
Testing takes a lot of time and I didn't want to keep my old computer powered on , so I wrote it all on paper.
I tried to measure the HC's and LVC's the same way to directly compare , but like you read above with Cpd's (that don't make sense) , there are differences between simple gates of course and for the same freq / supply voltage , Texas Instr 's LVC gates comsume a little less than TI's HC gates despite LVC's much higher speed and output current (lower Rdson of the output FET's).
If you mean 74LS04 ... and read the above , you will know I have no interest in TTL logic . LS = low power schottky TTL. I have only 1 TTL IC I bought more than 30 years ago : 74F74 to compare to 74AC11074 , the 74AC won ...
Fun fact , or not : The old 4069UB with 32,768 Khz X-tal <10 uA op 5 V.
10 Mhz X-tal : 1,12 mA with 2 of it's brothers as buffer on 5 V , quits at about 4,35 V . Pretty good for the old CMOS4000.
20Mhz x-tal LVC00 without buffer on output : 2,16 mA on 3,3V.
I like the 1 gate Schmitt-trigger oscillators , but with TI's 74HC132 : 1 khz with 10nF x 100K = 800uA at 5V. Same but 1K x 10nF = about 100Khz at 2,02 mA , 5V. So the 32Khz cheap X-tal with the 4069UB at 10 uA , clearly wins. For ST oscillators Nexperia's consume more . One thing NXP is not better at.
Simple set up : 4 oscillators with LVC17 buffers for nice rise and fall slopes . HC and LVC D-FF to divide the higher freq because my DMM could only go to 10 Mhz according to its specs , but it could measure 20 Mhz.
A HC393 for further division .Then the HC an LVC gates to measure current draw at different frequencies and at different Vcc's , with signal in 1 or the 2 inputs ( or 3 for LVC10) . Threshold voltage & max current draw at the thresholds , that is why 2 of the trimmer are for. 2 regulators with trimmers for 1,25 to 3,6V for the LVC's and 125 Mhz oscillator. 2,5 to 5,6 V for HC and the LVC1G17's . Current measurement over a 1 ohm R , DMM measures down to 10uV , so 1mV = 1 mA , easy.
Testing takes a lot of time and I didn't want to keep my old computer powered on , so I wrote it all on paper.
I tried to measure the HC's and LVC's the same way to directly compare , but like you read above with Cpd's (that don't make sense) , there are differences between simple gates of course and for the same freq / supply voltage , Texas Instr 's LVC gates comsume a little less than TI's HC gates despite LVC's much higher speed and output current (lower Rdson of the output FET's).
If you mean 74LS04 ... and read the above , you will know I have no interest in TTL logic . LS = low power schottky TTL. I have only 1 TTL IC I bought more than 30 years ago : 74F74 to compare to 74AC11074 , the 74AC won ...
Fun fact , or not : The old 4069UB with 32,768 Khz X-tal <10 uA op 5 V.
10 Mhz X-tal : 1,12 mA with 2 of it's brothers as buffer on 5 V , quits at about 4,35 V . Pretty good for the old CMOS4000.
20Mhz x-tal LVC00 without buffer on output : 2,16 mA on 3,3V.
I like the 1 gate Schmitt-trigger oscillators , but with TI's 74HC132 : 1 khz with 10nF x 100K = 800uA at 5V. Same but 1K x 10nF = about 100Khz at 2,02 mA , 5V. So the 32Khz cheap X-tal with the 4069UB at 10 uA , clearly wins. For ST oscillators Nexperia's consume more . One thing NXP is not better at.
Last edited:
Ok, I understand; it could be interesting to gather results in XL spreadsheet.
There are so many families today, with very close characteristics, that it become difficult to make a choice and then to get the parts available from your usual distributor. Very often, I choose the type of logic function I need, then goes to the distributors sites to check if I found something which suits my needs with a reasonable price. Unless specific needs I often fall in LVC.
Chris
There are so many families today, with very close characteristics, that it become difficult to make a choice and then to get the parts available from your usual distributor. Very often, I choose the type of logic function I need, then goes to the distributors sites to check if I found something which suits my needs with a reasonable price. Unless specific needs I often fall in LVC.
Chris
^ Wel you have to test because when they come with these unreliable numbers it's a gamble on what you buy. Of course this is CMOS logic , not uV offsets and nV noise of opamps . Logic has wide tolerances , and most won't give a toss whether it consumes a mA more or less , or a couple of ns more or less.
Still if you buy at a distributor where prices are sort of the same you try to buy the best one , but you'll never know for sure until you test them for yourself.
For LVC without real ST you really need fast slopes on the input .
Still if you buy at a distributor where prices are sort of the same you try to buy the best one , but you'll never know for sure until you test them for yourself.
For LVC without real ST you really need fast slopes on the input .
^ Your preference of course in Pdil and 0,412 $ and 25 pF . I'll have the NXP AHC14 0,277 $ in SOIC with less Cpd and about the same speed or buy at TI : AHC14 0,155 $ > out of stock , I wonder why , so the 0,166$.
Even faster : 74LVC1G14 also 5 V 0,066 $ in SC-70 but 25pF at 5V .
74AHC14 is definite competition for the 74HCS04 or 14 (at 0142 $) which is just a bit slower , AC14 not so much.
Even faster : 74LVC1G14 also 5 V 0,066 $ in SC-70 but 25pF at 5V .
74AHC14 is definite competition for the 74HCS04 or 14 (at 0142 $) which is just a bit slower , AC14 not so much.
Last edited:
I made 2 errors here , sorry . I just retested it to be sure.
32,768Khz on 2,5 V is 10 uA , not 5 V. On 5V it is only 120uA so still better than the 1 gate ST . When I ever get ST LVC's or the ultra low power AUP ST inverter I will test them for lower current draw and push it to its extreme to see what the highest oscillator freq is that you can get with it .
2nd error : remarkably the 10Mhz Xtal oscillator with 4069UB does not quit at 4,35 V , but goes on even below the 3 V which is the minimum voltage for the 4000 series.
It quits at 2,6 V , at 2,7 V it draws 390 uA at 5 V but when you short the input of the inverter you need to go over 3 V to start it again. At 3V it takes 530 uA for 10 Mhz and that is great of an old and slower 4069UB.
I looked into that , but I'm still in wonderful XP
, with office 2003 that I barely ever use. Excel probably isn't compatible with the new ones , I know I can't open the new ones with excel 2003.Heading is also A B C... you can't name columns ... I never use excel.
I use Works database v7 which you can't open if you don't have Works , which MS killed off a while ago.
And is kind of a mess : values for different frequencies , at different supply voltages, sometimes different voltages for the input frequency , 1 or both (or even 3 ) inputs used of a gate and measured.
...wonderful XP
Save, As... CSV is pretty nearly universal. Won't preserve multiple pages or graphs but real sheet-whackers don't do such things. eXcel warns you about lost features but it is just bragging.
This forum won't attach a CSV, but tack a .TXT on the end, and tell downloaders to strip it off.
There is a whole pile of Service Packs, labeled MSO 2007 but applicable to 2003, which lets old programs open/save most new MS formats.
There is also Open/Libre/Free-Office.
^I made a test excel but there is no CSV option to ''save as" .
Your attached file only shows 4 numbers in a text doc.
I had to take a course in windows office around 2003 and it was boring as hell. Computers should be fun . Office software isn't . (of course trouble shooting isn't fun , but a necessity)
It's going to be tough to change all my Works Databases into something new when I have to buy a new computer hopefully with linux.
Your attached file only shows 4 numbers in a text doc.
I had to take a course in windows office around 2003 and it was boring as hell. Computers should be fun . Office software isn't . (of course trouble shooting isn't fun , but a necessity)
It's going to be tough to change all my Works Databases into something new when I have to buy a new computer hopefully with linux.
I've heard that before. MSO's stupid Save As window, even on a 1500px high screen, only shows a few of the dozens of possible save formats. That is why it has a scroll bar.
With screen-grabs of MSO and O/S version.
Attachments
Apparently not much interest or people that have used HCS on an audio forum so I
reached out to the TI forum to get some answers before ordering.
You need to have a business email address to post there , but because of a glitch or
fluke , I got in anyway. It's a repeat of some of the things I already wrote here .
SN74HCS00: 74HCS gates' Cpd - Logic forum - Logic - TI E2E support forums
I've got a decent but not a crystal clear answer .
I waited to post it here because I wanted to post some follow up questions to clarify but couldn't get back in.
So I contacted them , and they were nice and helpful but wouldn't post my
questions. I understand of course , I shouldn't have been able posted the first one
anyway.
Yes datasheets have errors , check out this one :
https://www.ti.com/lit/ds/symlink/sn74hc02.pdf?ts=1622187160742 page 7
Cpd 4,5 pF ...this was revised dec 2020 !
I get that datasheet checker must be a very boring job , unless you're really interested in the components .
TI has some big pdf's about other families like AHC , AUC ,.. but only SCEA69A
about HCS . (that I can find)
reached out to the TI forum to get some answers before ordering.
You need to have a business email address to post there , but because of a glitch or
fluke , I got in anyway.
It's a repeat of some of the things I already wrote here . SN74HCS00: 74HCS gates' Cpd - Logic forum - Logic - TI E2E support forums
I've got a decent but not a crystal clear answer .
I waited to post it here because I wanted to post some follow up questions to clarify but couldn't get back in.
So I contacted them , and they were nice and helpful but wouldn't post my
questions. I understand of course , I shouldn't have been able posted the first one
anyway.
Yes datasheets have errors , check out this one :
https://www.ti.com/lit/ds/symlink/sn74hc02.pdf?ts=1622187160742 page 7
Cpd 4,5 pF ...this was revised dec 2020 !
I get that datasheet checker must be a very boring job , unless you're really interested in the components .
TI has some big pdf's about other families like AHC , AUC ,.. but only SCEA69A
about HCS . (that I can find)
Constant Cpd from 2V to 6V supply is very unlikelyYes datasheets have errors , check out this one :
https://www.ti.com/lit/ds/symlink/sn74hc02.pdf?ts=1622187160742 page 7
Cpd 4,5 pF ...this was revised dec 2020 !
I get that datasheet checker must be a very boring job , unless you're really interested in the components .
TI has some big pdf's about other families like AHC , AUC ,.. but only SCEA69A
about HCS . (that I can find)
^ Sure , but 4.5pF at 2V to 9.5pF at 6V for HCS ?
Very few datasheets give Cpd at different voltages .
It was one of my follow up questions I didn't get to post .
The TI guy states :
"For the HCS family, it's actually not strange for all of the gates to be10 pF because of how we define the spec and how this family of devices work. Technically the Cpd varies from about 4.5pF at 2V to 9.5pF at 6V, however we only put one value in the datasheet, and it's only a typical value.Our systems engineer chose to go with 10pF because Cpd is traditionally specified at the maximum supply voltage".
Look at D-FF's HCS74 datasheet pg 5 https://www.ti.com/lit/ds/symlink/sn74hcs74.pdf?ts=1623634175191https://www.ti.com/lit/ds/symlink/s...h%3Dsearch-everything%26usecase%3Dpartmatches
it says 2 (yes two) to 6 V typical Cpd = 10 pF. Checking some other families , when (rarely) they show Cpd for different voltages it doesn't vary that much . Look at 74AUP1G00 https://www.ti.com/lit/ds/symlink/sn74aup1g00.pdf pg 9 : Cpd all 4 pF ... And some others too but never so much as 4.5pF at 2V to 9.5 at 6 V .
Yes, yes I know "typical value" ...
Another thing he says when I asked about the big difference in Cpd for LVC74's regular and 1G, 2G packages :
"The first release of the SN74LVC74A datasheet was in 1993, while the SN74LVC1G74 came out in October 2009. Technology changes rapidly -- even though these two are specified for similar supply voltages, they are not built on the same process node"
So technology became much worse for the 74LVC1G74 and 2G74 ( and at a much higher price) , looking at their Cpd....or again datasheet error ?
Very few datasheets give Cpd at different voltages .
It was one of my follow up questions I didn't get to post .
The TI guy states :
"For the HCS family, it's actually not strange for all of the gates to be10 pF because of how we define the spec and how this family of devices work. Technically the Cpd varies from about 4.5pF at 2V to 9.5pF at 6V, however we only put one value in the datasheet, and it's only a typical value.Our systems engineer chose to go with 10pF because Cpd is traditionally specified at the maximum supply voltage".
Look at D-FF's HCS74 datasheet pg 5 https://www.ti.com/lit/ds/symlink/sn74hcs74.pdf?ts=1623634175191https://www.ti.com/lit/ds/symlink/s...h%3Dsearch-everything%26usecase%3Dpartmatches
it says 2 (yes two) to 6 V typical Cpd = 10 pF. Checking some other families , when (rarely) they show Cpd for different voltages it doesn't vary that much . Look at 74AUP1G00 https://www.ti.com/lit/ds/symlink/sn74aup1g00.pdf pg 9 : Cpd all 4 pF ... And some others too but never so much as 4.5pF at 2V to 9.5 at 6 V .
Yes, yes I know "typical value" ...
Another thing he says when I asked about the big difference in Cpd for LVC74's regular and 1G, 2G packages :
"The first release of the SN74LVC74A datasheet was in 1993, while the SN74LVC1G74 came out in October 2009. Technology changes rapidly -- even though these two are specified for similar supply voltages, they are not built on the same process node"
So technology became much worse for the 74LVC1G74 and 2G74 ( and at a much higher price) , looking at their Cpd....or again datasheet error ?
Ok , one year later ! Apparently not much interest in T.I's 74HCS , or no-one tried it. But I wanted to get some closure on this topic and finally bought and tested some of them (74HCS132 , 74HCS32, 74HCS14, 74HCS74 ).
Short version (so you don't have to read the rest) : They are all that T.I promised : Renewed HCmos with Schmitt trigger inputs with higher speed at lower current consumption. Their Cpd is much closer than the wildly swinging Cpd of HCmos.
As suspected in opening post the ST thresholds are closer together than HC or AHC.
I will not be buying anymore HC's IF there is a HCS version. Complete succes. Weird that 2 years after their introduction, no other manufacturer is making them , or maybe under different name ?
Short version (so you don't have to read the rest) : They are all that T.I promised : Renewed HCmos with Schmitt trigger inputs with higher speed at lower current consumption. Their Cpd is much closer than the wildly swinging Cpd of HCmos.
As suspected in opening post the ST thresholds are closer together than HC or AHC.
I will not be buying anymore HC's IF there is a HCS version. Complete succes. Weird that 2 years after their introduction, no other manufacturer is making them , or maybe under different name ?