Help with component ID

[patricks]

Hi everyone,

Hoping for some help with IDs on some components, please. I'm hunting for some NPN transistors in some old scavenged junk, and not having any luck googling the codes on these.

The items in question are labeled
Metal can things labelled 586-024 7750 - I think these might be Fairchild General purpose single op-amps

Metal can things labelled 014-364 106 - I think these are military spec PNP transistors

TI201 (TO-92 package) - are these j201 jfets from TI? If so yay (incidentally, is active still using thru-hole jfets given the lack of availability, or have people moved to using op-amps instead?). Otherwise the 014698 on the other side might mean they're the same component as below

Motorola things in a TO-92 package labelled 014 698 - I think these might be equivalent to 2N5088 NPN transistors but I'm not sure

Any help greatly appreciated 
Cheers, Pat








 

[Rob Strand]

Always tricky with those customer part numbers.

The board seems like 1978.   Some parts 1977.

Metal can things labelled 586-024 7750 - I think these might be Fairchild General purpose single op-amps

Agreed. 
I think the part number is an Ampex part number cross-referencing to a uA709C.
(I saw a pic going back to 1973 which is consistent with uA709C.)

https://ia904506.us.archive.org/19/items/Ampex-AVR-1-Preliminary-Manual-Vol-4/Ampex%20AVR-1%20Preliminary%20Manual%20Vol.%204.pdf

Metal can things labelled 014-364 106 - I think these are military spec PNP transistors

I can only add the manufacturer is TI.

After some digging through that doc I linked, it's also an Ampex part.  2N3072.

TI201 (TO-92 package) - are these j201 jfets from TI? If so yay (incidentally, is active still using thru-hole jfets given the lack of availability, or have people moved to using op-amps instead?). Otherwise the 014698 on the other side might mean they're the same component as below

I thought TI stopped making small transistors by 1977.

The other marking on the part is 014698 which is the same as the Motorola part.

Motorola things in a TO-92 package labelled 014 698 - I think these might be equivalent to 2N5088 NPN transistors but I'm not sure

014-698 is another Ampex part.
Still haven't found a list
014-698  appears in this cross-reference,
http://www.uneeda-audio.com/ampex/xistr.htm

2N3117
2N3565
2N5088

I think these are linked because of the gain.
2N3565 is Fairchild process 10, but is 2N5088?
2N3117 seems to be a Motorola device.

This Ampex doc gives the 014-698 as 2N3565,
https://ampex.sufficiently.com/files/Ampex_AG-440C/6e_electronics.pdf

[patricks]

Fantastic, thanks so much Rob!

70s = Vintage mojo parts, woo hoo! 

Thanks for confirming, and the links are great. The transistors will be handy, there are at least ten in the boards.

The opamp is interesting, I'll try sticking it in the socket in a couple of circuits (rat, orange squeezer) and see how it sounds.

Once again, really appreciate the help

[Rob Strand]

All usable parts that's for sure.   (Got lucky with that Ampex link.  The opamp led me there.)

The opamp is interesting, I'll try sticking it in the socket in a couple of circuits (rat, orange squeezer) and see how it sounds.

The opamp is old in the eyes of both audio and instrumentation people.   However there's no reason it can be put to good use in a pedal.   IMHO the uA709 is unexplored territory.   

This is one of the earlier (and good) datasheets,
https://www.alldatasheet.com/datasheet-pdf/pdf/54849/FAIRCHILD/UA709.html

Notice you need the compensation cap(s) and resistor.   If you dig around you might find a few different values in different circuits.

[patricks]

Thanks again!
I hadn't found a datasheet with the pin layout for the metal can package yet. I noticed the need for a compensation cap, I'll experiment with values and see what happens 

[patricks]

I'm quite happy with the haul, in the end! I rescued:
Transistors:
2 x BC108 NPn hFE 359 and 439

014-698 NPN transistors (2N3117, 2N3565, 2N5088 equivalent)

• 18 x TI branded; 10 with hFE 181-292, 8 with hFE 399-439
• 5 x Motorola branded, with different pinout (EBC vs CBE) and hFE 669-639

2 x Motorola branded 014-653 (NPN, 2N3904 equivalent) hFE 143 and 174

4 x MPS3638A (PNP, general purpose) hFE 102-190

1 x 014-364 (PNP, general purpose, mil spec, metal can) hFE 98

Op-amps:
2 x 586-024 single op-amps, ua709 equivalent

1 x U5B XXX-393 (part of the label had rubbed off); I've tracked this down as a single op-amp, either also a ua709 equivalent or possibly a ua741 equivalent. My vote's on ua709 since the other two are these equivalents.

Diodes:
31 x 1N914 equivalent
3 x 1N4004

Some good gain ranges there for all the usual applications.
I had a hard time understanding the compensation cap arrangement for the ua709 equivalent op-amps, then realised that in the datasheet the transient response test circuit has a cap and a resistor from pin 1 to 8 (both labelled compensation in), plus another cap from pin 5 (labelled compensation out) to pin 6/output.
The values for the compensation caps and resistors are quite different to the LM308 compensation cap so at least I've learned in advance it's not a drop-in replacement in the rat

[Rob Strand]

1 x U5B XXX-393 (part of the label had rubbed off); I've tracked this down as a single op-amp, either also a ua709 equivalent or possibly a ua741 equivalent. My vote's on ua709 since the other two are these equivalents.

How sure are you of the -393?   
The majority of Ampex IC parts numbes start with code 586- but I can't find a 586-393.
There's a few other prefixes used for ICs like 806- and 809-.

Around the -393 region we get.
[See section on Fairchild]

Linear:
586-321   uA723C, F723
586-335   LM306, LM306H, LM306M
586-415   MC857P, MC857
586-416   MC839P, MC839

Logic:
586-326   SN7404
586-336   DM8800
586-435   F9308, (uL9308?)
586-441   DM8830

Could be anything.  Perhaps a good idea to confirm the power pins by tracing the PCB
and which pins have any compensation caps, if any.  If you are really keen trace the
circuit around it.

I had a hard time understanding the compensation cap arrangement for the ua709 equivalent op-amps, then realised that in the datasheet the transient response test circuit has a cap and a resistor from pin 1 to 8 (both labelled compensation in), plus another cap from pin 5 (labelled compensation out) to pin 6/output.
The values for the compensation caps and resistors are quite different to the LM308 compensation cap so at least I've learned in advance it's not a drop-in replacement in the rat

The LM709C is definitely different to other opamps.  It's probably the earliest jellybean opamp.

You will need to dig up some old circuits and applications notes for tips.   A good deal of the time you might be thinking: Why those values?   To be honest it's so long since I've looked at the LM709C I'd be thinking the same thing.  FWIW there's two compensation networks.  One around the output stage and one  back in the "preamp".   IIRC the one around the output is sometimes optional and the one around the preamp is more like normal compensation.

Another approach would be to put the LM709C in a RAT circuit and tinker with the compensation values to see if you could match the waveform.

Here's another datasheet:
https://www.javanelec.com/CustomAjax/GetAppDocument/cadc4e0a-d28d-417a-b083-aa3c400dfa2d?type=1&inlineName=True
Just to be clear:  U5x...393 is a Fairchild part number.  Not an Ampex number.  That being the case
we don't expect a 586- prefix.

However, ...  I can't find U5A ... 393 in the Fairchild documents.

http://www.bitsavers.org/components/fairchild/_dataBooks/1971_Fairchild_Linear_Integrated_Circuits_Data_Catalog.pdf

But on page 10 there's a lot of different parts with numbers,
U5x...393
U6A..393

One standing out is
MC1458P U6A7747393
Interesting:

"the μA709 was introduced in 1965"  - National Semiconductor, AN-241

"... For example, the old UA709 amplifier had some
good features , and some characteristics
that were not so good. But the UA709's low
noise was never mentioned on its datasheet.
So, to this day, people ask me , "Where can I
find a low-noise op-amp with  lots of
gain-bandwidth product?"
"And I have t o explain that the UA709 really does have good
low noise voltage,  lower than all  the 741s
or the BIFETs.  So, in some applications , the
23-year-old UA709  is still the best  IC f o r
the job.  But  the datasheet kept that a
secret ."  - Bob Pease ~ 1990

Bob Pease also mentions the uA709 is easy to blow up
not protected against ESD.

Small expose on the LM709 by Bob Pease in National Semiconductor AN-1485.

[PRR]

> uA709 is easy to blow up

I'll confirm that. Altho one(?) data sheet said '709 would survive brief shorts on the output, I never figured how brief 'brief' was. Had a large and costly RCA broadcast mix console that was all '709s. The least probe slip and it was off to Radio Shack (Tandy) to get another '709. Sure was lower Voltage hiss than the '301 that came next.

And as Pease says: IF you were as smart as Widlar you could figure a compensation faster than most of the chips in the next decade. But ordinary mortals were best served by squinting the compensation graph for their closed loop gain and taking those values seriously. As well as best-practice supply bypassing.

[patricks]

How sure are you of the -393?   
The majority of Ampex IC parts numbes start with code 586- but I can't find a 586-393.
There's a few other prefixes used for ICs like 806- and 809-.

Around the -393 region we get.
[See section on Fairchild]

Linear:
586-321   uA723C, F723
586-335   LM306, LM306H, LM306M
586-415   MC857P, MC857
586-416   MC839P, MC839

Logic:
586-326   SN7404
586-336   DM8800
586-435   F9308, (uL9308?)
586-441   DM8830

Could be anything.  Perhaps a good idea to confirm the power pins by tracing the PCB
and which pins have any compensation caps, if any.  If you are really keen trace the
circuit around it.

Here's another datasheet:
https://www.javanelec.com/CustomAjax/GetAppDocument/cadc4e0a-d28d-417a-b083-aa3c400dfa2d?type=1&inlineName=True
Just to be clear:  U5x...393 is a Fairchild part number.  Not an Ampex number.  That being the case
we don't expect a 586- prefix.

However, ...  I can't find U5A ... 393 in the Fairchild documents.

http://www.bitsavers.org/components/fairchild/_dataBooks/1971_Fairchild_Linear_Integrated_Circuits_Data_Catalog.pdf

But on page 10 there's a lot of different parts with numbers,
U5x...393
U6A..393

One standing out is
MC1458P U6A7747393

Only a little bit sure 
I came across the same Fairchild Linear Integrated Circuits Data Catalog, the ua709 equivalence is a best guess based on page 25 listing the ua709 with a part number of U5B7709393 and the fact that the other two are ua709 equivalents. But like you said, there are a long list of options in the catalogue.
This document has a Fairchild to Motorola conversion table, and options for U5B XXXX 393 include
ua702
ua709
ua710
ua741
ua748

I'll never really know without some extensive testing, and even then I might risk blowing it up or it may be a dud.
I think I'll mount it in resin with an LED under it as an ornament so that it can cast its mojo glow over my work space 

You will need to dig up some old circuits and applications notes for tips.   A good deal of the time you might be thinking: Why those values?   To be honest it's so long since I've looked at the LM709C I'd be thinking the same thing.  FWIW there's two compensation networks.  One around the output stage and one  back in the "preamp".   IIRC the one around the output is sometimes optional and the one around the preamp is more like normal compensation.

Another approach would be to put the LM709C in a RAT circuit and tinker with the compensation values to see if you could match the waveform.

I've managed to find a couple of application notes, so I've got a starting point for cap and resistor values, and one of the 709s in the board I rescued it from has a cap-and-resistor across pins 1 and 8 so I've got a real-world reference to compare against too.
I've also got a modified rat circuit on the breadboard at the moment, so I'll tinker with that

Interesting:

"the μA709 was introduced in 1965"  - National Semiconductor, AN-241

"... For example, the old UA709 amplifier had some
good features , and some characteristics
that were not so good. But the UA709's low
noise was never mentioned on its datasheet.
So, to this day, people ask me , "Where can I
find a low-noise op-amp with  lots of
gain-bandwidth product?"
"And I have t o explain that the UA709 really does have good
low noise voltage,  lower than all  the 741s
or the BIFETs.  So, in some applications , the
23-year-old UA709  is still the best  IC f o r
the job.  But  the datasheet kept that a
secret ."  - Bob Pease ~ 1990

Bob Pease also mentions the uA709 is easy to blow up
not protected against ESD.

Small expose on the LM709 by Bob Pease in National Semiconductor AN-1485.

> uA709 is easy to blow up

I'll confirm that. Altho one(?) data sheet said '709 would survive brief shorts on the output, I never figured how brief 'brief' was. Had a large and costly RCA broadcast mix console that was all '709s. The least probe slip and it was off to Radio Shack (Tandy) to get another '709. Sure was lower Voltage hiss than the '301 that came next.

And as Pease says: IF you were as smart as Widlar you could figure a compensation faster than most of the chips in the next decade. But ordinary mortals were best served by squinting the compensation graph for their closed loop gain and taking those values seriously. As well as best-practice supply bypassing.

That's very cool, thanks! In my searching I came across the Op Amps for Everyone design reference, there's some history of op-amps at the beginning, and apparently one major military manufacturer published a paper titled something like (The 12 Pearl Harbour conditions of the ua709" (I can't find the original publication) 
Given that they're so easy to blow up and these boxes were found on a scrap pile outside a studio they may all be duds, but it'll be fun testing anyway

[Rob Strand]

I came across the same Fairchild Linear Integrated Circuits Data Catalog, the ua709 equivalence is a best guess based on page 25 listing the ua709 with a part number of U5B7709393 and the fact that the other two are ua709 equivalents. But like you said, there are a long list of options in the catalogue.
This document has a Fairchild to Motorola conversion table, and options for U5B XXXX 393 include
ua702
ua709
ua710
ua741
ua748

I'll never really know without some extensive testing, and even then I might risk blowing it up or it may be a dud.
I think I'll mount it in resin with an LED under it as an ornament so that it can cast its mojo glow over my work space

I didn't expect Motorola to provide a reverse cross-reference since the Fairchild part numbers are non-standard ones!

You could definitely narrow things down from the PCB. In particular the presence of the compensation caps.  Maybe the power supply pins.

The LM710/uA710 has a 10 pin dip package but there's also a 8-pin can.  Not sure if the pin out for the can matches an LM741.

The LM702 might not come in 8 pin packages.

So that would leave only the compensation parts on the PCB to help identify the part.

I kept a MC68000 CPU as a souvenir.  They are massive DIP parts.  Someone could easily think they are fake:

http://www.retro.co.za/68000/marvin/CPU.jpg

[patricks]

Those are good points, thanks. I'll cross reference the pinouts and see whether I can narrow it down.

That thing's enormous! I think the raspberry pi pico units that I've got in drawer somewhere are about the same size. I can see why people might think it's fake 

[patricks]

I traced out the connections to the unknown chip as best I could today, it seems to be either a 709, 741, or 740, uncompensated.

I think I'll freeze it in carbonite and bask in its mojo