11 Feb
2014
11 Feb
'14
7:49 a.m.
The RICM needs to repair a S603 triple pulse amplifier flip-chip. The
failed diode is listed as a DEC D-664. The DEC enthusiast cross-reference
lists show it as a 1N3606. These diodes are nearly impossible to find. Can
someone recommend a substitution for the 1N3606 diode?
--
Michael Thompson
11 Feb
11 Feb
9:05 a.m.
On 02/11/2014 07:49 AM, Michael Thompson wrote:
The RICM needs to repair a S603 triple pulse amplifier
flip-chip. The
failed diode is listed as a DEC D-664. The DEC enthusiast cross-reference
lists show it as a 1N3606. These diodes are nearly impossible to find. Can
someone recommend a substitution for the 1N3606 diode?
A quick lookup suggests that 1n914 or 1n4148 are suitable choices.
These are
very common fast switching doides.
Allison
9:49 a.m.
From: allison: Tuesday, February 11, 2014 6:05 AM
On 02/11/2014 07:49 AM, Michael Thompson wrote:
My understanding of the difference between a D662 and a D664 is that
the D664 is supposed to be a fast switching diode with a low forward
voltage drop. The D662 is a similar diode, but requires a known, fairly
high forward voltage drop. (I'd recommend select 1N914 or 1N4148
for the D662, but only some manufacturers have forward voltage drops
of 0.75V.) Unfortunately, forward voltage drop doesn't seem to be a
thing that part numbers of modern switching diodes routinely control.
The D664, however, wants the forward drop as low as possible. Some
1N914 and 1N4148 will fit the bill, but others won't. If you are looking
for a general purpose substitution (rather than a specific application),
the 1N4154 should work well. (When I studied the issue, I didn't find
any manufacturers of 1N4154 which had high forward voltage drops.)
Hope that helps,
Vince
The RICM needs to repair a S603 triple pulse
amplifier flip-chip. The
failed diode is listed as a DEC D-664. The DEC enthusiast cross-reference
lists show it as a 1N3606. These diodes are nearly impossible to find. Can
someone recommend a substitution for the 1N3606 diode?
A quick lookup suggests that 1n914 or 1n4148 are suitable choices.
These are
very common fast switching doides. 
12:17 p.m.
On Feb 11, 2014, at 9:49 AM, Vincent Slyngstad <vrs at msn.com> wrote:
From: allison: Tuesday, February 11, 2014 6:05 AM
The bright side is that they're common as dirt and about as cheap, so if
you're willing to invest some time and effort getting intimate with a
multimeter, you can buy a thousand of them and bin them for voltage drop
yourself. 1N914/4148 (to my knowledge, they're basically identical) is
usually specified for "I need a diode and don't expect to blow it up",
so the characteristics aren't usually all that tightly controlled.
On 02/11/2014 07:49 AM, Michael Thompson wrote:
My understanding of the difference between a D662 and a D664 is that the D664 is supposed
to be a fast switching diode with a low forward voltage drop. The D662 is a similar
diode, but requires a known, fairly high forward voltage drop. (I'd recommend select
1N914 or 1N4148 for the D662, but only some manufacturers have forward voltage drops
of 0.75V.) Unfortunately, forward voltage drop doesn't seem to be a thing that part
numbers of modern switching diodes routinely control. The RICM needs to repair a S603 triple pulse
amplifier flip-chip. The
failed diode is listed as a DEC D-664. The DEC enthusiast cross-reference
lists show it as a 1N3606. These diodes are nearly impossible to find. Can
someone recommend a substitution for the 1N3606 diode?
A quick lookup suggests that 1n914 or 1n4148 are suitable choices. These are
very common fast switching doides. The D664, however, wants the forward drop as low as
possible. Some 1N914 and 1N4148 will fit the bill, but others won't. If you are
looking for a general purpose substitution (rather than a specific application), the
1N4154 should work well. (When I studied the issue, I didn't find any manufacturers
of 1N4154 which had high forward voltage drops.)
If you're looking for low forward voltage drops, you could also try a
Schottky diode, which also tend to be fairly fast. Some may be hard
to find in through-hole, but depending on how worried you are about
visual authenticity, you could wire up some surface-mount packages
either with actual wires or with a tiny PCB jig.
- Dave
12:35 p.m.
On Feb 11, 2014, at 12:17 PM, David Riley <fraveydank at gmail.com> wrote:
On Feb 11, 2014, at 9:49 AM, Vincent Slyngstad <vrs at msn.com> wrote:
...
I?m a bit puzzled by this ?low voltage drop? requirement. Yes, Schottky will do that,
though I didn?t think those were in common use back then. Or, for sufficiently old
designs, germanium diodes.
The D664, however, wants the forward drop as low
as possible. Some 1N914 and 1N4148 will fit the bill, but others won't. If you are
looking for a general purpose substitution (rather than a specific application), the
1N4154 should work well. (When I studied the issue, I didn't find any manufacturers
of 1N4154 which had high forward voltage drops.)
If you're looking for low forward voltage drops, you could also try a
Schottky diode, which also tend to be fairly fast. Some may be hard
to find in through-hole, but depending on how worried you are about
visual authenticity, you could wire up some surface-mount packages
either with actual wires or with a tiny PCB jig. From what I remember of long ago physics and EE, the
forward drop of a junction diode is a function of the material and does not change
significantly from device to device. So if you?re looking at a silicon junction diode (as
the 1N914 is), the forward drop should be pretty much the same for all of them.
paul
2:15 p.m.
From: Paul Koning: Tuesday, February 11, 2014 9:35 AM
Yes, the D664 is used in much the same way today's designer would
use a Schottky diode. Presumably because Schottky diodes were not
in common use at the time.
On Feb 11, 2014, at 12:17 PM, David Riley
<fraveydank at gmail.com> wrote:
> On Feb 11, 2014, at 9:49 AM, Vincent Slyngstad <vrs at msn.com> wrote:
>
> ...
>> The D664, however, wants the forward drop as low as possible.
If you're
looking for low forward voltage drops, you could also try a
Schottky diode, which also tend to be fairly fast. Some may be hard
to find in through-hole, but depending on how worried you are about
visual authenticity, you could wire up some surface-mount packages
either with actual wires or with a tiny PCB jig.
I?m a bit puzzled by this ?low voltage drop? requirement. Yes, Schottky
will do that, though I didn?t think those were in common use back then. From what I remember of long ago physics and EE, the
forward drop
of a junction diode is a function of the material and does not change
significantly from device to device. So if you?re looking at a silicon
junction diode (as the 1N914 is), the forward drop should be pretty
much the same for all of them.
From a given manufacturer with a given manufacturing
process/plant,
yes. But it isn't considered an important property of a
switching diode.
A different manufacturer, or even a different factory of the same one,
the diode's internal geometry may be different. Some Fairchild 1N914,
for instance, look like 1N4448, and some look like 1N4148. I suspect
they run both processes, and just sell whichever as the 1N914.
Presumably, if you cared (these days), you'd have specified a Schottky
diode. I haven't decided what to think about blanket substitution of
Schottky diodes in the vintage circuits, though.
Vince
12:54 p.m.
From: David Riley: Tuesday, February 11, 2014 9:17 AM
On Feb 11, 2014, at 9:49 AM, Vincent Slyngstad <vrs
at msn.com> wrote:
The only problem with that plan is that a given manufacturer, and
especially a given lot, will likely have nearly identical characteristics.
One could, however potentially discover that some lots of Fairchild
1N914 are more like D664, while Vishay are always more like D662,
etc.
Vince
of 0.75V.) Unfortunately, forward voltage drop
doesn't seem to be a
thing that part numbers of modern switching diodes routinely control.
The bright side is that they're common as dirt and about as cheap, so if
you're willing to invest some time and effort getting intimate with a
multimeter, you can buy a thousand of them and bin them for voltage drop
yourself. 1N914/4148 (to my knowledge, they're basically identical) is
usually specified for "I need a diode and don't expect to blow it up",
so the characteristics aren't usually all that tightly controlled.
6:43 p.m.
On 02/11/2014 09:49 AM, Vincent Slyngstad wrote:
From: allison: Tuesday, February 11, 2014 6:05 AM
Most 1n914/4148s I've used don't see .75V till you near the max
current. Usually then are more
like .6-.62. If you need real low forward drop the Schottky 1n5711 are
around .25-.30 depending
on current.
On 02/11/2014 07:49 AM, Michael Thompson wrote:
My understanding of the difference between a D662 and a D664 is that
the D664 is supposed to be a fast switching diode with a low forward
voltage drop. The D662 is a similar diode, but requires a known,
fairly high forward voltage drop. (I'd recommend select 1N914 or
1N4148 for the D662, but only some manufacturers have forward voltage
drops
of 0.75V.) Unfortunately, forward voltage drop doesn't seem to be a
thing that part numbers of modern switching diodes routinely control.
The RICM needs to repair a S603 triple pulse
amplifier flip-chip. The
failed diode is listed as a DEC D-664. The DEC enthusiast
cross-reference
lists show it as a 1N3606. These diodes are nearly impossible to
find. Can
someone recommend a substitution for the 1N3606 diode?
A quick lookup suggests that 1n914 or 1n4148 are suitable choices.
These are
very common fast switching doides. The D664, however, wants the forward drop as low as
possible. Some
1N914 and 1N4148 will fit the bill, but others won't. If you are
looking for a general purpose substitution (rather than a specific
application), the 1N4154 should work well. (When I studied the issue,
I didn't find any manufacturers of 1N4154 which had high forward
voltage drops.)
Generally the voltage drop MAY affect the noise margin at the inter
flip-chip level. However there
is already margin there to addow for temperature and other things.
I suspect a working diode (not open or shorted) will make it go. Keep
in mind fast by then standard is
slower then 1n914.
Allison
12:26 p.m.
On Tue, Feb 11, 2014 at 07:49:45AM -0500, Michael Thompson wrote:
The RICM needs to repair a S603 triple pulse amplifier
flip-chip. The
failed diode is listed as a DEC D-664. The DEC enthusiast cross-reference
lists show it as a 1N3606. These diodes are nearly impossible to find. Can
someone recommend a substitution for the 1N3606 diode?
My list says sub: 1N3606 Newsletter said 1N3604, 1N914
http://www.pdp8online.com/pdp8cgi/query_docs/query.pl?Search=cross
*Newsletter
http://archive.computerhistory.org/resources/access/text/2012/03/102654910-…
page 100 of pdf.
3822
days inactive
3822
days old
8 comments
6 participants
participants (6)
-
ajp166＠verizon.net -
djg＠pdp8online.com -
fraveydank＠gmail.com -
michael.99.thompson＠gmail.com -
paulkoning＠comcast.net -
vrs＠msn.com