On a sunny day (Sat, 29 Jan 2022 00:27:22 -0000 (UTC)) it happened Arnie Dwyer
<spamme@not.com> wrote in <XnsAE2DC5EADE5E9idtokenpost@144.76.35.252>:
>Sid 03 <sidwelle@gmail.com> wrote:
>
>> Decade counter:
>> I am looking for a decade counter, I found some on-line like the 4017
>> and 74HC[T]390. But I am now sure any of those are what I want. At my
>> previous job we use to have quad decade counters in one chip available.
>> That has been a few years ago and not sure where to look now. What I
>> want to be able to do is tie at least two of them end to end and get a
>> divider of 10 and 100. Maybe the terminology is wrong and I should be
>> looking for a divider ?
>>
>> Any help is appreciated.
>> Thanks
>
>Tom Van Baak published a list of PIC counters that give various ratios,
>such as 1e7 (10MHz to 1Hz) with jitter under 2ps. Here is a list with
>source code:
>
>picDIV -- Single Chip Frequency Divider
>http://www.leapsecond.com/pic/picdiv.htm
>
>PIC divider jitter measurement
>http://www.leapsecond.com/pic/jitter/
Sure, using Mircochip PICs much is possible
here my frequency counter in an RS232 connector, powered from the RS232 DTR::
http://panteltje.com/panteltje/pic/freq_pic/
But then he needs a PIC programmer etc...
I am sure an ebay search will get you many cheap frequency counters too.
Reply by Arnie Dwyer●January 28, 20222022-01-28
Sid 03 <sidwelle@gmail.com> wrote:
> Decade counter:
> I am looking for a decade counter, I found some on-line like the 4017
> and 74HC[T]390. But I am now sure any of those are what I want. At my
> previous job we use to have quad decade counters in one chip available.
> That has been a few years ago and not sure where to look now. What I
> want to be able to do is tie at least two of them end to end and get a
> divider of 10 and 100. Maybe the terminology is wrong and I should be
> looking for a divider ?
>
> Any help is appreciated.
> Thanks
On Tue, 25 Jan 2022 18:31:47 -0800 (PST), Sid 03 <sidwelle@gmail.com>
wrote:
>On Tuesday, January 25, 2022 at 7:45:11 PM UTC-6, bill....@ieee.org wrote:
>> On Wednesday, January 26, 2022 at 3:51:44 AM UTC+11, sidw...@gmail.com wrote:
>> > Decade counter:
>> > I am looking for a decade counter, I found some on-line like the 4017 and 74HC[T]390.
>> > But I am now sure any of those are what I want. At my previous job we use to have quad decade counters in one chip available. That has been a few years ago and not sure where to look now.
>> > What I want to be able to do is tie at least two of them end to end and get a divider of 10 and 100.
>> > Maybe the terminology is wrong and I should be looking for a divider ?
>> >
>> > Any help is appreciated.
>> The market for long counters in a single package probably went away when programmable logic chips came along. Something out of the Xilinx Coolrunner range could probably be programmed to do you job without needing much supply current.
>>
>> https://www.xilinx.com/products/silicon-devices/cpld/coolrunner-ii.html
>>
>> I've used an ICT PA7024 electrically erasable programmable logic array to do that kind of job, but that was nearly thirty years ago - the charm of the PA7024 was that was a drop-in replacement for the 22V10 part, but appreciably more powerful (if not all that powerful).
>>
>> --
>> Bill Sloman, Sydney
>
>I posted here because I don't know much about the counters and wanted some advice.
>This is just a project to try and measure the revolutions of a pump.
>I see alot of these on-line as "CD74HC390E".
>If someone could help me decode the prefix 'CD' and suffix 'E' that would be a big help as well.
How are you sensing pump rotation? Be aware that these IC counters
are far faster than any pump, and if you don't have some kind of low
pass filter followed by a Schmitt trigger, the IC counters will run
far faster than any pump ever could - you will be measuring edge noise
and mechanical chatter, not rotation.
Joe Gwinn
Reply by Dimiter_Popoff●January 26, 20222022-01-26
On 1/26/2022 17:55, Phil Hobbs wrote:
> Jasen Betts wrote:
>> On 2022-01-25, Sid 03 <sidwelle@gmail.com> wrote:
>>> Decade counter:
>>
>>> I am looking for a decade counter, I found some on-line like the
>>> 4017 and 74HC[T]390. But I am now sure any of those are what I want.
>>> At my previous job we use to have quad decade counters in one chip
>>> available.
>>
>> I seem to remember a triple or quad with muliplexed BCD output, but
>> IIRC they stopped making them last century.
>
> Interesil used to make frequency counter chips like that, and AD
> actually still sells the ICM7217 four-digit decade counter. Of course
> it's $21 and has muxed 7-segment outputs, but hey, it's a quad decade
> counter. ;)
>
> Cheers
>
> Phil Hobbs
>
I remembered an 8 bit counter with a writable latch so one could make
it divide by any 8 bit value but it took looking at an old library
disk to recall its name, it was 74LS592... I must have used it as a
baud rate divider at some point or something, 30+ years ago. I knew
I had used it but can't remember for what purpose and on which board.
Looks like it never got made as HC/HCT though, probably no longer
available. But it must have been nice, writable via an 8 bit bus
in a DIP16 (I have drawn the part I found only in dip16, looks like
I never used it in so-16).
Reply by Phil Hobbs●January 26, 20222022-01-26
Jasen Betts wrote:
> On 2022-01-25, Sid 03 <sidwelle@gmail.com> wrote:
>> Decade counter:
>
>> I am looking for a decade counter, I found some on-line like the
>> 4017 and 74HC[T]390. But I am now sure any of those are what I want.
>> At my previous job we use to have quad decade counters in one chip available.
>
> I seem to remember a triple or quad with muliplexed BCD output, but
> IIRC they stopped making them last century.
Interesil used to make frequency counter chips like that, and AD
actually still sells the ICM7217 four-digit decade counter. Of course
it's $21 and has muxed 7-segment outputs, but hey, it's a quad decade
counter. ;)
Cheers
Phil Hobbs
--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510
http://electrooptical.nethttp://hobbs-eo.com
Reply by Anthony William Sloman●January 26, 20222022-01-26
On Wednesday, January 26, 2022 at 2:56:55 PM UTC+11, Don Y wrote:
> On 1/25/2022 7:31 PM, Sid 03 wrote:
> > I posted here because I don't know much about the counters and wanted some advice.
>
> > This is just a project to try and measure the revolutions of a pump.
>
> So, it's a dead slow input signal.
Which probably means that you need to put it through a Schmitt trigger input stage. Slowly changing voltages have a nasty habit of producing a cluster of edges as the input goes through the switching voltage. Schmitt triggers can't prevent this, but they do make it less likely.
> Are you interested in getting "nice" (human readable) numbers? Or,
> just driving the output frequency down?
>
> I.e., unless the pump emits "X*100 pulses per revolution" (in which case,
> dividing by 100 would give you "X revolutions"), you may, instead, want
> to pick a divisor that directly gives you a number that is easier to
> "consume".
>
> E.g., if you are wanting "RPM" and wanted to update your "data" every second,
> you'd be looking to reduce:
> N pulses per revolution / 60
> So, if the pump produced 100 pulses per revolution and you observed 100 pulses
> *in* that second, you would know that the pump was rotating at 60 RPM. If
> your frequency divisor divided by (100 / 60), you would directly see that
> result.
>
> Or, if one revolution moved M units of liquid (?), you could normalize
> your output to directly yield units per minute, hour, etc. by an appropriate
> choice of frequency divisor.
>
> If you don't care about "nice units", then you can divide by anything that
> drives the output frequency low enough to be directly observable (with
> whatever you have "watching" this output). In which case, you can pick
> any old ripple counter and use it.
> > I see alot of these on-line as "CD74HC390E".
> > If someone could help me decode the prefix 'CD' and suffix 'E' that would be a big help as well.
> The prefix is chosen by the vendor of the part. CD was RCA, SN was TI, MC was
> Motogorilla, MM for Nat Semi, etc. (note vendors can choose multiple different
> prefixes... *hopefully* without conflict with other vendors!).
>
> The suffix often indicates a set of operating conditions -- temperature range,
> accuracy, supply voltage tolerances, etc.
In this case it's about packaging, and nothing else.
> Sometimes, there may also be additional suffixes -- like "dash numbers" to
> indicate speed ranges.
>
> The "HC" embedded in the part number often indicates a logic family (high speed
> CMOS... as the original CMOS parts were typically pretty pokey).
HCT was faster than basic CMOS, but HCT specifically said that it was TTL compatible. HC wasn't. ACT is faster still and also TTL compatible.
> The numeric portion of the part number defines the actual functionality
> of the part. "In general" (ha!), a xx74yy###zzz from any manufacturer
> will be the same basic part fabricated in different technologies, speed
> grades, etc.
>
> Note that a manufacturer need not follow this "rule". They are free to
> offer THEIR parts under whatever numbering scheme *they* develop. So,
> you might find a particular *functionality* in MECL III with an entirely
> different part number than 100K ECL, etc. A vendor has control over the
> part numbers he offers. And, can even change the part numbers of
> existing products to suit his fancy (e.g., the NS16032 magically became
> the NS32016, overnight!)
>
> [There have been some annoying deviations from this "rule" -- most notably
> the early 27xx EPROMs where it was crucial to specify a vendor AND a
> part number instead of just a "generic" part number]
Don Y knows more about software than hardware, which has made this post less useful than it might have been.
--
Bill Sloman, Sydney
Reply by Jasen Betts●January 26, 20222022-01-26
On 2022-01-25, Sid 03 <sidwelle@gmail.com> wrote:
> Decade counter:
> I am looking for a decade counter, I found some on-line like the
> 4017 and 74HC[T]390. But I am now sure any of those are what I want.
> At my previous job we use to have quad decade counters in one chip available.
I seem to remember a triple or quad with muliplexed BCD output, but
IIRC they stopped making them last century.
> That has been a few years ago and not sure where to look now.
> What I want to be able to do is tie at least two of them end to end
> and get a divider of 10 and 100.
4518 is a dual synchronous counter with individual BCD outputs,
the "Q2" or "Q3" outputs will give divide-by 10 but not at 50% duty cycle.
4518 variants are still available from TI
--
Jasen.
Reply by Don Y●January 25, 20222022-01-25
On 1/25/2022 7:31 PM, Sid 03 wrote:
> I posted here because I don't know much about the counters and wanted some advice.
> This is just a project to try and measure the revolutions of a pump.
So, it's a dead slow input signal.
Are you interested in getting "nice" (human readable) numbers? Or,
just driving the output frequency down?
I.e., unless the pump emits "X*100 pulses per revolution" (in which case,
dividing by 100 would give you "X revolutions"), you may, instead, want
to pick a divisor that directly gives you a number that is easier to
"consume".
E.g., if you are wanting "RPM" and wanted to update your "data" every second,
you'd be looking to reduce:
N pulses per revolution / 60
So, if the pump produced 100 pulses per revolution and you observed 100 pulses
*in* that second, you would know that the pump was rotating at 60 RPM. If
your frequency divisor divided by (100 / 60), you would directly see that
result.
Or, if one revolution moved M units of liquid (?), you could normalize
your output to directly yield units per minute, hour, etc. by an appropriate
choice of frequency divisor.
If you don't care about "nice units", then you can divide by anything that
drives the output frequency low enough to be directly observable (with
whatever you have "watching" this output). In which case, you can pick
any old ripple counter and use it.
> I see alot of these on-line as "CD74HC390E".
> If someone could help me decode the prefix 'CD' and suffix 'E' that would be a big help as well.
The prefix is chosen by the vendor of the part. CD was RCA, SN was TI, MC was
Motogorilla, MM for Nat Semi, etc. (note vendors can choose multiple different
prefixes... *hopefully* without conflict with other vendors!).
The suffix often indicates a set of operating conditions -- temperature range,
accuracy, supply voltage tolerances, etc.
Sometimes, there may also be additional suffixes -- like "dash numbers" to
indicate speed ranges.
The "HC" embedded in the part number often indicates a logic family (high speed
CMOS... as the original CMOS parts were typically pretty pokey).
The numeric portion of the part number defines the actual functionality
of the part. "In general" (ha!), a xx74yy###zzz from any manufacturer
will be the same basic part fabricated in different technologies, speed
grades, etc.
Note that a manufacturer need not follow this "rule". They are free to
offer THEIR parts under whatever numbering scheme *they* develop. So,
you might find a particular *functionality* in MECL III with an entirely
different part number than 100K ECL, etc. A vendor has control over the
part numbers he offers. And, can even change the part numbers of
existing products to suit his fancy (e.g., the NS16032 magically became
the NS32016, overnight!)
[There have been some annoying deviations from this "rule" -- most notably
the early 27xx EPROMs where it was crucial to specify a vendor AND a
part number instead of just a "generic" part number]
Reply by Rick C●January 25, 20222022-01-25
On Tuesday, January 25, 2022 at 10:31:51 PM UTC-4, sidw...@gmail.com wrote:
> On Tuesday, January 25, 2022 at 7:45:11 PM UTC-6, bill....@ieee.org wrote:
> > On Wednesday, January 26, 2022 at 3:51:44 AM UTC+11, sidw...@gmail.com wrote:
> > > Decade counter:
> > > I am looking for a decade counter, I found some on-line like the 4017 and 74HC[T]390.
> > > But I am now sure any of those are what I want. At my previous job we use to have quad decade counters in one chip available. That has been a few years ago and not sure where to look now.
> > > What I want to be able to do is tie at least two of them end to end and get a divider of 10 and 100.
> > > Maybe the terminology is wrong and I should be looking for a divider ?
> > >
> > > Any help is appreciated.
> > The market for long counters in a single package probably went away when programmable logic chips came along. Something out of the Xilinx Coolrunner range could probably be programmed to do you job without needing much supply current.
> >
> > https://www.xilinx.com/products/silicon-devices/cpld/coolrunner-ii.html
> >
> > I've used an ICT PA7024 electrically erasable programmable logic array to do that kind of job, but that was nearly thirty years ago - the charm of the PA7024 was that was a drop-in replacement for the 22V10 part, but appreciably more powerful (if not all that powerful).
> >
> > --
> > Bill Sloman, Sydney
> I posted here because I don't know much about the counters and wanted some advice.
> This is just a project to try and measure the revolutions of a pump.
> I see alot of these on-line as "CD74HC390E".
> If someone could help me decode the prefix 'CD' and suffix 'E' that would be a big help as well.
The prefix is not a code, rather simply a maker's mark, if you will. If a company designs a chip they give it a number usually with a letter or two or three at the start. If someone licenses the design they use the exact same part number. If they instead make a similar part with the same functionality, they often use the same base number, but append their own letters. The letters at the end are not consistent across all semiconductors, but often are for a maker or at least a line of parts from a maker. 'E' often means a high degree of static resistance. Or it may simply be a revision letter although 'E' is getting up there. It could also be an indicator of environment aspects such as being free of various harmful compounds/elements. It can also be the package designation.
The data sheet will tell you nothing about the prefix, but may elude something about the suffix. I have found that many times makers use some odd designation details that they do not explain in black and white in the data sheet. I've often had to call to find out some details of a part number. You would think they might consider it important to make it easy to figure out what part number to use to order the durn things!
--
Rick C.
- Get 1,000 miles of free Supercharging
- Tesla referral code - https://ts.la/richard11209
Reply by Anthony William Sloman●January 25, 20222022-01-25
On Wednesday, January 26, 2022 at 1:31:51 PM UTC+11, sidw...@gmail.com wrote:
> On Tuesday, January 25, 2022 at 7:45:11 PM UTC-6, bill....@ieee.org wrote:
> > On Wednesday, January 26, 2022 at 3:51:44 AM UTC+11, sidw...@gmail.com wrote:
> > > Decade counter:
> > > I am looking for a decade counter, I found some on-line like the 4017 and 74HC[T]390.
> > > But I am now sure any of those are what I want. At my previous job we use to have quad decade counters in one chip available. That has been a few years ago and not sure where to look now.
> > > What I want to be able to do is tie at least two of them end to end and get a divider of 10 and 100.
> > > Maybe the terminology is wrong and I should be looking for a divider ?
> > >
> > > Any help is appreciated.
> > The market for long counters in a single package probably went away when programmable logic chips came along. Something out of the Xilinx Coolrunner range could probably be programmed to do you job without needing much supply current.
> >
> > https://www.xilinx.com/products/silicon-devices/cpld/coolrunner-ii.html
> >
> > I've used an ICT PA7024 electrically erasable programmable logic array to do that kind of job, but that was nearly thirty years ago - the charm of the PA7024 was that was a drop-in replacement for the 22V10 part, but appreciably more powerful (if not all that powerful).
>
> I posted here because I don't know much about the counters and wanted some advice.
> This is just a project to try and measure the revolutions of a pump.
> I see a lot of these on-line as "CD74HC390E".
>
> If someone could help me decode the prefix 'CD' and suffix 'E' that would be a big help as well.