Author Topic: The beginning of SMT, old equipment  (Read 4979 times)

0 Members and 1 Guest are viewing this topic.

Offline joeqsmithTopic starter

  • Super Contributor
  • ***
  • Posts: 11630
  • Country: us
The beginning of SMT, old equipment
« on: August 15, 2020, 01:08:12 pm »
I had seen parts surface mounted from Japan in the early 70s but I was not personally involved with it in until the 80s. 

We had a small benchtop pick and place machine that had a dedicated PC with it.  I think it used audio tapes like the early hobby computers.  It was made by a major Japanese company.  Maybe Sanyo.  I tried to find a picture of it to show.


Offline gooligumelec

  • Contributor
  • Posts: 34
  • Country: au
    • Gooligum Electronics
Re: The beginning of SMT, old equipment
« Reply #1 on: August 19, 2020, 07:17:41 am »
I've been sorting through (tossing out most of) my old hobby electronics magazines.  I was surprised in a 1981 or '82 (no later than that) Electronics Today to see a project where they specified some SMT parts because they needed to minimise lead inductance.
So if it was starting to hit the hobby sphere by the early '80s, it must have existed in industry in the '70s.

David Meiklejohn
www.gooligum.com.au
 

Offline greenpossum

  • Frequent Contributor
  • **
  • Posts: 408
  • Country: au
Re: The beginning of SMT, old equipment
« Reply #2 on: August 19, 2020, 08:22:40 am »
I remember data sheets from the TTL (standard, not LS), and maybe even the DTL period that offered flatpack packages. So they go back many decades.

Here we are, a DTL chip from 1966 with flat package option.

« Last Edit: August 19, 2020, 08:29:47 am by greenpossum »
 

Offline capt bullshot

  • Super Contributor
  • ***
  • Posts: 3033
  • Country: de
    • Mostly useless stuff, but nice to have: wunderkis.de
Re: The beginning of SMT, old equipment
« Reply #3 on: August 19, 2020, 08:41:43 am »
I remember data sheets from the TTL (standard, not LS), and maybe even the DTL period that offered flatpack packages. So they go back many decades.

Here we are, a DTL chip from 1966 with flat package option.

Seen them too. Often with shiny gold plated leads, so I imagined they were typically used in hi-rel aerospace, spacecraft or military applications.
« Last Edit: August 19, 2020, 08:45:23 am by capt bullshot »
Safety devices hinder evolution
 

Online exe

  • Supporter
  • ****
  • Posts: 2559
  • Country: nl
  • self-educated hobbyist
Re: The beginning of SMT, old equipment
« Reply #4 on: August 19, 2020, 10:28:02 am »
I remember seeing brand new old Chinese/Soviet chips in junk stores with late 70s date code in surface mount ceramic packages which resemble what a modern SOP looks like with metric pitch, but the pins are flat instead of gull wing.

Ah, yes, remember them too! I've seen ICs in white ceramic substrate and gold leads. Presumably, those were for military/aerospace applications.

Some more examples can be seen here: https://xlom.ru/vidy-metalloloma/radiolom-raznovidnosti-opisanie .
 

Offline joeqsmithTopic starter

  • Super Contributor
  • ***
  • Posts: 11630
  • Country: us
Re: The beginning of SMT, old equipment
« Reply #5 on: August 19, 2020, 11:39:28 am »
From my old computer from the early 70s.  This was the only board made in Japan.  The rest were all made in the USA using TH parts.   

Offline free_electron

  • Super Contributor
  • ***
  • Posts: 8515
  • Country: us
    • SiliconValleyGarage
Re: The beginning of SMT, old equipment
« Reply #6 on: August 19, 2020, 11:59:45 am »
but , people ! the original IC's WERE surface mounted !
The apollo flight computer was all SMT
All the early Cray machines ( Cray -i ) was all SMT
The very first RTL gates from Texas instruments came in 10 pin ceramic SMT package. That was 1 year after the invention of the IC !

Surface mount is as old as the integrated circuit.
Professional Electron Wrangler.
Any comments, or points of view expressed, are my own and not endorsed , induced or compensated by my employer(s).
 
The following users thanked this post: Alex Eisenhut, tooki

Online exe

  • Supporter
  • ****
  • Posts: 2559
  • Country: nl
  • self-educated hobbyist
Re: The beginning of SMT, old equipment
« Reply #7 on: August 19, 2020, 12:38:20 pm »
Why we had TH components at all? For manual assembly?
 

Offline SilverSolder

  • Super Contributor
  • ***
  • Posts: 6126
  • Country: 00
Re: The beginning of SMT, old equipment
« Reply #8 on: August 19, 2020, 02:20:07 pm »
Why we had TH components at all? For manual assembly?

I would guess that what we today think of as through-hole components, were once considered point-to-point components with no PCB in sight...

 
The following users thanked this post: exe

Offline greenpossum

  • Frequent Contributor
  • **
  • Posts: 408
  • Country: au
Re: The beginning of SMT, old equipment
« Reply #9 on: August 19, 2020, 02:35:58 pm »
Why we had TH components at all? For manual assembly?

Once upon a time there were no boards and holes, unless you count the holes in the terminal strips.
 

Offline MK14

  • Super Contributor
  • ***
  • Posts: 4527
  • Country: gb
Re: The beginning of SMT, old equipment
« Reply #10 on: August 19, 2020, 02:39:15 pm »
I remember data sheets from the TTL (standard, not LS), and maybe even the DTL period that offered flatpack packages. So they go back many decades.

Here we are, a DTL chip from 1966 with flat package option.

It seems that the early/original 1961 and onwards, TTL devices. Were especially for military applications. Hence the (what we would call surface mount), what they called, 14 lead flat-pack, packages.

This is from 1966, although it was originally from 1963, when it was that manufacturer.

http://www.bitsavers.org/components/sylvania/SM2927_Sylvania_Universal_High_Level_Logic_May66.pdf

The following link, explains the early TTL, was before Texas Instruments (around 1964), industry standard versions. Apparently, before the 74 series, there was actually a 64 series. (Military can be the 54 series).

https://en.wikipedia.org/wiki/7400-series_integrated_circuits

Quote
History

Texas Instruments SN5451 in the original flat package
Although the 7400 series was the first de facto industry standard TTL logic family (i.e. second-sourced by several semiconductor companies), there were earlier TTL logic families such as:

Sylvania Universal High-level Logic in 1963[8][9][10]
Motorola MC4000 MTTL[11][12][13]
National Semiconductor DM8000[14][better source needed]
Fairchild 9300 series[15][16]
Signetics 8200 and 8T00[13][15][17]
The 7400 quad NAND gate was the first product in the series, introduced by Texas Instruments in a military grade metal flat package (5400W)

EDIT:
Here we go. Link seems to say, first flat packs, invented in 1962.

https://en.wikipedia.org/wiki/Flatpack_(electronics)

Quote
History

A Flatpack integrated circuit on a circuit board, among TO-5 packages
The original flatpack was invented by Y. Tao in 1962 while he was working for Texas Instruments to improve heat dissipation
« Last Edit: August 19, 2020, 02:48:41 pm by MK14 »
 
The following users thanked this post: tooki

Offline vk6zgo

  • Super Contributor
  • ***
  • Posts: 7563
  • Country: au
Re: The beginning of SMT, old equipment
« Reply #11 on: August 19, 2020, 11:40:11 pm »
Even prior to the general use of ICs, surface mount capacitors were used in wideband RF amplifiers.
They were commonly called "chip capacitors" in those days------- the modern ones look exactly the same, & are used in the same application!
 
The following users thanked this post: MK14

Offline joeqsmithTopic starter

  • Super Contributor
  • ***
  • Posts: 11630
  • Country: us
Re: The beginning of SMT, old equipment
« Reply #12 on: August 20, 2020, 01:26:21 am »
but , people ! the original IC's WERE surface mounted !
The apollo flight computer was all SMT
All the early Cray machines ( Cray -i ) was all SMT
The very first RTL gates from Texas instruments came in 10 pin ceramic SMT package. That was 1 year after the invention of the IC !

Surface mount is as old as the integrated circuit.

But Free_electron!!! I don't believe there was a single post about WHEN ICs were first surface mounted!!

I was actually more curious about how people were processing them.   That first P&P machine we ever had looked like it was made by Fisher Price for a 9 year old.   It was slow enough that we could assemble boards faster by hand.   

Why we had TH components at all? For manual assembly?

We were very slow to adopt SMT mainly because of reliability.  TH parts had a bit more give to them.  I remember damaging some of the early boards with the potting compound.  At one point we were using conformal coating to give the potting some cushion.   10 or so years later we ran into problems again with potting with some QFPs.  The thermal expansion was enough to fracture the joints.   I think back to those days compared with the parts we place today.  It's amazing how things have changed.     

Offline free_electron

  • Super Contributor
  • ***
  • Posts: 8515
  • Country: us
    • SiliconValleyGarage
Re: The beginning of SMT, old equipment
« Reply #13 on: August 20, 2020, 11:59:06 am »
Why we had TH components at all? For manual assembly?

I would guess that what we today think of as through-hole components, were once considered point-to-point components with no PCB in sight...
correct. resistors caps and liek were soldered to eyelets.
when circuit boards came along , bend the leads and poke them in holes.
vacuum tubes got pins for sockets.
transistors came along when circuit boards were already common so it made sense to give em leads like capacitors and resistors wo we can poke them in holes.
when integrated circuits came into existence stuff had advanced far enough that surface mounting was possible. it does not make sense to make a microchip and then stuff it in a package that occupies more space than its implementation with discretes would do...

Now, there is another reason the first IC's were ceramic flat-pack.
- hermetical sealing: the epoxy compound we all now know did not exist ( the compound existed , but not the method of packaging ic's in it ) . ceramic side brazed packages already existed. They could be leak tested , so perfect for the fragile integrated circuit.
- cooling : the first ic's were power hungry and ran hot. pack a bunch together and you have a thermal issue... the ceramic material is a good heat conductor. put the ic on a board and press a heatsink to the package.

That's what Cray did. They did not want pins through a board as every pin occupies a potential routing channel. Many connections could be made without using a single via so the gain in routing density was very large. you could pack the devices closer to each other increasing the assembly density. And that was the entire goal : miniaturisation.
Professional Electron Wrangler.
Any comments, or points of view expressed, are my own and not endorsed , induced or compensated by my employer(s).
 
The following users thanked this post: SilverSolder

Offline jmelson

  • Super Contributor
  • ***
  • Posts: 2758
  • Country: us
Re: The beginning of SMT, old equipment
« Reply #14 on: August 21, 2020, 03:42:03 am »
I had seen parts surface mounted from Japan in the early 70s but I was not personally involved with it in until the 80s. 

We had a small benchtop pick and place machine that had a dedicated PC with it.  I think it used audio tapes like the early hobby computers.  It was made by a major Japanese company.  Maybe Sanyo.  I tried to find a picture of it to show.
I have a computer that was built in 1965 for the X15 project (Chuck Yeager and all, breaking the sound barrier) that is entirely built with surface mount chips.
That is the Honeywell Alert.

Jon
 
The following users thanked this post: joeqsmith, MK14

Offline rrinker

  • Super Contributor
  • ***
  • Posts: 2046
  • Country: us
Re: The beginning of SMT, old equipment
« Reply #15 on: August 21, 2020, 03:29:29 pm »

Now, there is another reason the first IC's were ceramic flat-pack.
- hermetical sealing: the epoxy compound we all now know did not exist ( the compound existed , but not the method of packaging ic's in it ) . ceramic side brazed packages already existed. They could be leak tested , so perfect for the fragile integrated circuit.
- cooling : the first ic's were power hungry and ran hot. pack a bunch together and you have a thermal issue... the ceramic material is a good heat conductor. put the ic on a board and press a heatsink to the package.

That's what Cray did. They did not want pins through a board as every pin occupies a potential routing channel. Many connections could be made without using a single via so the gain in routing density was very large. you could pack the devices closer to each other increasing the assembly density. And that was the entire goal : miniaturisation.

 If you look at the Sylvania datasheet MK14 posted, they specifically mention the sealing as well as the thermal transfer characteristics for the flat pack in the section on the packaging near the bottom. The more things change, the more they stay the same.
 

Offline joeqsmithTopic starter

  • Super Contributor
  • ***
  • Posts: 11630
  • Country: us
Re: The beginning of SMT, old equipment
« Reply #16 on: August 21, 2020, 06:32:40 pm »
I have a computer that was built in 1965 for the X15 project (Chuck Yeager and all, breaking the sound barrier) that is entirely built with surface mount chips.
That is the Honeywell Alert.

Jon
would you please post a few pictures of it?  I'm a bit of a vintage computer buff.

Offline jmelson

  • Super Contributor
  • ***
  • Posts: 2758
  • Country: us
Re: The beginning of SMT, old equipment
« Reply #17 on: August 22, 2020, 01:18:18 am »
would you please post a few pictures of it?  I'm a bit of a vintage computer buff.
Yes, I really need to do this!  I'll see if I find time this weekend.

Jon
 
The following users thanked this post: joeqsmith

Offline jmelson

  • Super Contributor
  • ***
  • Posts: 2758
  • Country: us
Re: The beginning of SMT, old equipment
« Reply #18 on: March 15, 2021, 12:30:08 am »
would you please post a few pictures of it?  I'm a bit of a vintage computer buff.
OK, it's been a long time, but here are some pics.
Note this is serial # 5.  I believe this computer came from Emerson Electric, and was used on the AN/ASQ-96 project, which apparently
was for locating the position of enemy radars.  But, the Honeywell Alert computer was developed for the X-15 project, and is mentioned in Chuck Yeager's book as the "mini-honey".  It is a 24-bit computer, and was one of the first built with integrated circuits.  Although it shares this honor with the Apollo Guidance Computer, which was started about a year earlier, it was delivered and flew on the X-15 first.

The computer fits in a 1/2 ATR rack, and the memory module fits in the same size housing next to it (not pictured).
It draws 25 A at 5 V DC, and fits on 6 modules that have two boards laminated together with chips on both sides.
Honeywell contracted with Texas Instruments to develop the ICs for this machine, a form of DTL.  Although the computer is connected to the outside world with ordinary D-style connectors. there are no connectors inside!  Each of the 6 modules is connected to the backplane by flexible cables, and cooled by being clamped to aluminum plates between the modules.  When the clamping bolts are removed, the modules can be swung open like a book.

First picture is the connector end of the computer.

Second picture is the label Emerson Electric applied to the case.  3rd picture is the Honeywell label.

4th and 5th pictures are of the modules on each side of the computer, 6th picture is between a couple of the modules.

7th pic is the cooling structure, last pic is another view between the modules.
« Last Edit: March 15, 2021, 12:36:59 am by jmelson »
 
The following users thanked this post: SilverSolder, MK14, alm

Offline jmelson

  • Super Contributor
  • ***
  • Posts: 2758
  • Country: us
Re: The beginning of SMT, old equipment
« Reply #19 on: March 15, 2021, 12:32:10 am »
Hmmm, too many large pics, I guess.
 
The following users thanked this post: SilverSolder, MK14, alm

Offline jmelson

  • Super Contributor
  • ***
  • Posts: 2758
  • Country: us
Re: The beginning of SMT, old equipment
« Reply #20 on: March 15, 2021, 12:33:48 am »
And some more...
 
The following users thanked this post: SilverSolder, MK14, alm, I wanted a rude username

Offline jmelson

  • Super Contributor
  • ***
  • Posts: 2758
  • Country: us
Re: The beginning of SMT, old equipment
« Reply #21 on: March 15, 2021, 12:34:30 am »
And the last couple.
 
The following users thanked this post: SilverSolder, MK14, alm

Offline rsjsouza

  • Super Contributor
  • ***
  • Posts: 5980
  • Country: us
  • Eternally curious
    • Vbe - vídeo blog eletrônico
Re: The beginning of SMT, old equipment
« Reply #22 on: March 15, 2021, 01:41:07 am »
Jnelson, are you Major Nelson, an airforce pilot married to a genie? :D

At any rate, thank you for the beautiful pictures. The method of opening the boards with no connectors reminded me of my Casio AS-A - some pictures at my site

Not as early, but the 1970's calculators quickly migrated to surface mount to reduce costs and miniaturization.
Vbe - vídeo blog eletrônico http://videos.vbeletronico.com

Oh, the "whys" of the datasheets... The information is there not to be an axiomatic truth, but instead each speck of data must be slowly inhaled while carefully performing a deep search inside oneself to find the true metaphysical sense...
 

Offline jmelson

  • Super Contributor
  • ***
  • Posts: 2758
  • Country: us
Re: The beginning of SMT, old equipment
« Reply #23 on: March 15, 2021, 01:50:55 am »
Jnelson, are you Major Nelson, an airforce pilot married to a genie? :D
Hardly, but this artifact is as close as I'll ever get to Chuck Yeager, etc.

Jon
 

Online Benta

  • Super Contributor
  • ***
  • Posts: 5839
  • Country: de
Re: The beginning of SMT, old equipment
« Reply #24 on: March 15, 2021, 01:53:34 pm »
@Jon, those pictures are just way too cool! Thanks.

IIRC, those military ceramic/gold flatpacks weren't soldered, but welded to the PCB (which is probably also ceramic). Can you confirm?

 

Offline jmelson

  • Super Contributor
  • ***
  • Posts: 2758
  • Country: us
Re: The beginning of SMT, old equipment
« Reply #25 on: March 15, 2021, 03:22:44 pm »
@Jon, those pictures are just way too cool! Thanks.

IIRC, those military ceramic/gold flatpacks weren't soldered, but welded to the PCB (which is probably also ceramic). Can you confirm?
I had a program loading/test console that came with the computer where the chips were ultrasonically welded to pins on the boards.
Then, the backs of the boards had something like wire-wrap wire ultrasonically welded to the pins.

But, this Alert computer is NOT built that way.  The chips are, as far as I can tell, HAND soldered to the glass-epoxy boards.
Some of the boards in the middle of the machine have LOTS of patch wires and glued-on chips on them.

The circuit modules have no traces on the outside of the modules.  All power and signal traces are buried inside the boards.
There is a massively thick motherboard that must have lots of layers to interconnect everything.

Also, as far as I can tell, the computer still works.  The first time I powered it up, I verified I could "jam in" 1's or 0's to most of the registers, but couldn't get it to clock.  The major registers have set and reset pins of the FFs brought out to connectors for this.  Later, I found you needed to supply -6 V for the clock generator.  Then, I could jam instructions into the memory read connector and the memory address would count up an various rates.  The multiply is a lot slower than add.  That is as far as I got with it.

Jon
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf