Can anyone identify this mystery IC
A little more... https://imgur.com/f4noWuk We have this on display under a microscope at VCF Mid-Atlantic museum. It is strangely constructed to my eye. no lead frame or bond wires. Its build ON the leadframe. Its symmetrical, appears to be a dual something. A mixed signal IC design collegue said: ---- That lead-frame is really weird. It seems to be some sort of flip-chip, but the leads don’t even line up from one side of the chip to the other! Never seen anything like it. The only other thing I would GUESS is that this is some sort of analog IC, probably bipolar. ----
I would say quad or hex, rather than dual. You can see it by finding those six sets of transistors, three on each side, each with the same sets of leads brought out to pins. Hex inverter, a-la 7404 but mil-spec? I agree that the asymmetrical package is very unusual. The density is so low that we should be able to suss out its function given a bit of study. Already the Vdd/Vss pins are identifiable at the center top and bottom. -Dave On 8/21/22 11:23, Douglas Crawford via vcf-midatlantic wrote:
A little more... https://imgur.com/f4noWuk
We have this on display under a microscope at VCF Mid-Atlantic museum.
It is strangely constructed to my eye. no lead frame or bond wires. Its build ON the leadframe. Its symmetrical, appears to be a dual something.
A mixed signal IC design collegue said: ---- That lead-frame is really weird. It seems to be some sort of flip-chip, but the leads don’t even line up from one side of the chip to the other! Never seen anything like it. The only other thing I would GUESS is that this is some sort of analog IC, probably bipolar. ----
-- Dave McGuire, AK4HZ New Kensington, PA
These are the transistors? https://imgur.com/aLpqVpE I got one additional response from another designer: "Looks like some kind of gate array with only the metalization layer being clearly visible. I worked on alot of these in the late 70's. I have no explanation for the bonding pad structures." On 8/21/2022 11:27 AM, Dave McGuire via vcf-midatlantic wrote:
I would say quad or hex, rather than dual. You can see it by finding those six sets of transistors, three on each side, each with the same sets of leads brought out to pins. Hex inverter, a-la 7404 but mil-spec?
I agree that the asymmetrical package is very unusual.
The density is so low that we should be able to suss out its function given a bit of study. Already the Vdd/Vss pins are identifiable at the center top and bottom.
-Dave
On 8/21/22 11:23, Douglas Crawford via vcf-midatlantic wrote:
A little more... https://imgur.com/f4noWuk
We have this on display under a microscope at VCF Mid-Atlantic museum.
It is strangely constructed to my eye. no lead frame or bond wires. Its build ON the leadframe. Its symmetrical, appears to be a dual something.
A mixed signal IC design collegue said: ---- That lead-frame is really weird. It seems to be some sort of flip-chip, but the leads don’t even line up from one side of the chip to the other! Never seen anything like it. The only other thing I would GUESS is that this is some sort of analog IC, probably bipolar. ----
On 8/21/22 12:39, Douglas Crawford via vcf-midatlantic wrote:
These are the transistors? https://imgur.com/aLpqVpE
I'd need to look more closely, but those appear to be clusters of transistors, six of them. They are likely gates.
I got one additional response from another designer: "Looks like some kind of gate array with only the metalization layer being clearly visible. I worked on alot of these in the late 70's. I have no explanation for the bonding pad structures."
Agreed; this has either had its passivation layer removed, or it had never been applied. I suspect the latter, as it would've been removed with acid, and there isn't so much as a trace of debris from that. -Dave -- Dave McGuire, AK4HZ New Kensington, PA
Ah, now we are getting somewhere. We can only guess then, from the remaining metalization layer correct? Educated guess? On 8/21/2022 12:43 PM, Dave McGuire via vcf-midatlantic wrote:
On 8/21/22 12:39, Douglas Crawford via vcf-midatlantic wrote:
These are the transistors? https://imgur.com/aLpqVpE
I'd need to look more closely, but those appear to be clusters of transistors, six of them. They are likely gates.
I got one additional response from another designer: "Looks like some kind of gate array with only the metalization layer being clearly visible. I worked on alot of these in the late 70's. I have no explanation for the bonding pad structures."
Agreed; this has either had its passivation layer removed, or it had never been applied. I suspect the latter, as it would've been removed with acid, and there isn't so much as a trace of debris from that.
-Dave
Yes, but we should be able to determine everything, or almost everything we need to know from that layer. We'd need better and higher-resolution photos. -Dave On 8/21/22 12:48, Douglas Crawford via vcf-midatlantic wrote:
Ah, now we are getting somewhere. We can only guess then, from the remaining metalization layer correct? Educated guess?
On 8/21/2022 12:43 PM, Dave McGuire via vcf-midatlantic wrote:
On 8/21/22 12:39, Douglas Crawford via vcf-midatlantic wrote:
These are the transistors? https://imgur.com/aLpqVpE
I'd need to look more closely, but those appear to be clusters of transistors, six of them. They are likely gates.
I got one additional response from another designer: "Looks like some kind of gate array with only the metalization layer being clearly visible. I worked on alot of these in the late 70's. I have no explanation for the bonding pad structures."
Agreed; this has either had its passivation layer removed, or it had never been applied. I suspect the latter, as it would've been removed with acid, and there isn't so much as a trace of debris from that.
-Dave
-- Dave McGuire, AK4HZ New Kensington, PA
Very good. Better images coming soon. We'll take it offline unless someone chimes in. DC On 8/21/2022 12:50 PM, Dave McGuire via vcf-midatlantic wrote:
Yes, but we should be able to determine everything, or almost everything we need to know from that layer. We'd need better and higher-resolution photos.
-Dave
On 8/21/22 12:48, Douglas Crawford via vcf-midatlantic wrote:
Ah, now we are getting somewhere. We can only guess then, from the remaining metalization layer correct? Educated guess?
On 8/21/2022 12:43 PM, Dave McGuire via vcf-midatlantic wrote:
On 8/21/22 12:39, Douglas Crawford via vcf-midatlantic wrote:
These are the transistors? https://imgur.com/aLpqVpE
I'd need to look more closely, but those appear to be clusters of transistors, six of them. They are likely gates.
I got one additional response from another designer: "Looks like some kind of gate array with only the metalization layer being clearly visible. I worked on alot of these in the late 70's. I have no explanation for the bonding pad structures."
Agreed; this has either had its passivation layer removed, or it had never been applied. I suspect the latter, as it would've been removed with acid, and there isn't so much as a trace of debris from that.
-Dave
please don't this is interesting On Sun, Aug 21, 2022, 12:53 PM Douglas Crawford via vcf-midatlantic < vcf-midatlantic@lists.vcfed.org> wrote:
Very good. Better images coming soon. We'll take it offline unless someone chimes in.
DC
On 8/21/2022 12:50 PM, Dave McGuire via vcf-midatlantic wrote:
Yes, but we should be able to determine everything, or almost everything we need to know from that layer. We'd need better and higher-resolution photos.
-Dave
On 8/21/22 12:48, Douglas Crawford via vcf-midatlantic wrote:
Ah, now we are getting somewhere. We can only guess then, from the remaining metalization layer correct? Educated guess?
On 8/21/2022 12:43 PM, Dave McGuire via vcf-midatlantic wrote:
On 8/21/22 12:39, Douglas Crawford via vcf-midatlantic wrote:
These are the transistors? https://imgur.com/aLpqVpE
I'd need to look more closely, but those appear to be clusters of transistors, six of them. They are likely gates.
I got one additional response from another designer: "Looks like some kind of gate array with only the metalization layer being clearly visible. I worked on alot of these in the late 70's. I have no explanation for the bonding pad structures."
Agreed; this has either had its passivation layer removed, or it had never been applied. I suspect the latter, as it would've been removed with acid, and there isn't so much as a trace of debris from that.
-Dave
One more image, accounting for the pins. Maybe blue are caps. Purple no idea at all. https://imgur.com/a/XlpS9dE On 8/21/2022 12:50 PM, Dave McGuire via vcf-midatlantic wrote:
Yes, but we should be able to determine everything, or almost everything we need to know from that layer. We'd need better and higher-resolution photos.
-Dave
On 8/21/22 12:48, Douglas Crawford via vcf-midatlantic wrote:
Ah, now we are getting somewhere. We can only guess then, from the remaining metalization layer correct? Educated guess?
On 8/21/2022 12:43 PM, Dave McGuire via vcf-midatlantic wrote:
On 8/21/22 12:39, Douglas Crawford via vcf-midatlantic wrote:
These are the transistors? https://imgur.com/aLpqVpE
I'd need to look more closely, but those appear to be clusters of transistors, six of them. They are likely gates.
I got one additional response from another designer: "Looks like some kind of gate array with only the metalization layer being clearly visible. I worked on alot of these in the late 70's. I have no explanation for the bonding pad structures."
Agreed; this has either had its passivation layer removed, or it had never been applied. I suspect the latter, as it would've been removed with acid, and there isn't so much as a trace of debris from that.
-Dave
They are changing their tune, Dave :) One says: " Metalization is hiding all the lower levels. Looks like a device with just 6 transistors." Another responds: "So it would be a *transistor* array, not a gate array then. If so, then I’m confused as to why it would need so many pins if there are only 6 transistors. Unless it was a prototyping part where all the transistors were individually made accessible." (I'm positing with them # pins is because possibly they're gates) On 8/21/2022 12:50 PM, Dave McGuire via vcf-midatlantic wrote:
Yes, but we should be able to determine everything, or almost everything we need to know from that layer. We'd need better and higher-resolution photos.
-Dave
On 8/21/22 12:48, Douglas Crawford via vcf-midatlantic wrote:
Ah, now we are getting somewhere. We can only guess then, from the remaining metalization layer correct? Educated guess?
On 8/21/2022 12:43 PM, Dave McGuire via vcf-midatlantic wrote:
On 8/21/22 12:39, Douglas Crawford via vcf-midatlantic wrote:
These are the transistors? https://imgur.com/aLpqVpE
I'd need to look more closely, but those appear to be clusters of transistors, six of them. They are likely gates.
I got one additional response from another designer: "Looks like some kind of gate array with only the metalization layer being clearly visible. I worked on alot of these in the late 70's. I have no explanation for the bonding pad structures."
Agreed; this has either had its passivation layer removed, or it had never been applied. I suspect the latter, as it would've been removed with acid, and there isn't so much as a trace of debris from that.
-Dave
I think those are actually six small groups of transistors, not six transistors. Possibly inverters; these can be built in a CMOS chip with two transistors, which would usually be physically adjacent. -Dave On 8/21/22 15:35, Douglas Crawford via vcf-midatlantic wrote:
They are changing their tune, Dave :)
One says: " Metalization is hiding all the lower levels. Looks like a device with just 6 transistors." Another responds: "So it would be a *transistor* array, not a gate array then. If so, then I’m confused as to why it would need so many pins if there are only 6 transistors. Unless it was a prototyping part where all the transistors were individually made accessible."
(I'm positing with them # pins is because possibly they're gates)
On 8/21/2022 12:50 PM, Dave McGuire via vcf-midatlantic wrote:
Yes, but we should be able to determine everything, or almost everything we need to know from that layer. We'd need better and higher-resolution photos.
-Dave
On 8/21/22 12:48, Douglas Crawford via vcf-midatlantic wrote:
Ah, now we are getting somewhere. We can only guess then, from the remaining metalization layer correct? Educated guess?
On 8/21/2022 12:43 PM, Dave McGuire via vcf-midatlantic wrote:
On 8/21/22 12:39, Douglas Crawford via vcf-midatlantic wrote:
These are the transistors? https://imgur.com/aLpqVpE
I'd need to look more closely, but those appear to be clusters of transistors, six of them. They are likely gates.
I got one additional response from another designer: "Looks like some kind of gate array with only the metalization layer being clearly visible. I worked on alot of these in the late 70's. I have no explanation for the bonding pad structures."
Agreed; this has either had its passivation layer removed, or it had never been applied. I suspect the latter, as it would've been removed with acid, and there isn't so much as a trace of debris from that.
-Dave
-- Dave McGuire, AK4HZ New Kensington, PA
Ah, so the red lines are not b/c/e? That's what I was first thinking. Here me trying to make some sense of it https://i.imgur.com/gOZujMR.jpg Red lines inputs? Could the blue things be capacitors? Maybe purple outputs? From this view... OR gates? At least I have some agreement now that several learned eyes see transistors. It might be enough for our purposes to just have a good guess of how many transistors there are, and be able to identify them. What they are doing is a less important detail for our purposes at this point in the exhibit. So how many do to think you see in each group? We may never know for sure, but I'll probably run with your number as an estimate. On 8/21/2022 3:43 PM, Dave McGuire via vcf-midatlantic wrote:
I think those are actually six small groups of transistors, not six transistors. Possibly inverters; these can be built in a CMOS chip with two transistors, which would usually be physically adjacent.
-Dave
On 8/21/22 15:35, Douglas Crawford via vcf-midatlantic wrote:
They are changing their tune, Dave :)
One says: " Metalization is hiding all the lower levels. Looks like a device with just 6 transistors." Another responds: "So it would be a *transistor* array, not a gate array then. If so, then I’m confused as to why it would need so many pins if there are only 6 transistors. Unless it was a prototyping part where all the transistors were individually made accessible."
(I'm positing with them # pins is because possibly they're gates)
On 8/21/2022 12:50 PM, Dave McGuire via vcf-midatlantic wrote:
Yes, but we should be able to determine everything, or almost everything we need to know from that layer. We'd need better and higher-resolution photos.
-Dave
On 8/21/22 12:48, Douglas Crawford via vcf-midatlantic wrote:
Ah, now we are getting somewhere. We can only guess then, from the remaining metalization layer correct? Educated guess?
On 8/21/2022 12:43 PM, Dave McGuire via vcf-midatlantic wrote:
On 8/21/22 12:39, Douglas Crawford via vcf-midatlantic wrote:
These are the transistors? https://imgur.com/aLpqVpE
I'd need to look more closely, but those appear to be clusters of transistors, six of them. They are likely gates.
I got one additional response from another designer: "Looks like some kind of gate array with only the metalization layer being clearly visible. I worked on alot of these in the late 70's. I have no explanation for the bonding pad structures."
Agreed; this has either had its passivation layer removed, or it had never been applied. I suspect the latter, as it would've been removed with acid, and there isn't so much as a trace of debris from that.
-Dave
On 8/21/22 17:07, Douglas Crawford via vcf-midatlantic wrote:
Ah, so the red lines are not b/c/e? That's what I was first thinking. Here me trying to make some sense of it https://i.imgur.com/gOZujMR.jpg
Red lines inputs? Could the blue things be capacitors?
I do not think they're capacitors. Remember, this is an IC; think three-dimensionally. All we can see from that picture is the upper metallization layer, where connections to silicon are made. There are several layers below that.
Maybe purple outputs? From this view... OR gates?
Possible.
At least I have some agreement now that several learned eyes see transistors. It might be enough for our purposes to just have a good guess of how many transistors there are, and be able to identify them. What they are doing is a less important detail for our purposes at this point in the exhibit.
Yes, there are definitely transistors there.
So how many do to think you see in each group? We may never know for sure, but I'll probably run with your number as an estimate.
Well, as I said, we'd need a higher-resolution image with better lighting. We need to be able to see a layer or two below that. I'm guessing there are two or three. -Dave -- Dave McGuire, AK4HZ New Kensington, PA
Pattern indicates 2 dimensional thinking. Sent from: My extremely complicated, hand held electronic device.
On Aug 21, 2022, at 5:32 PM, Dave McGuire via vcf-midatlantic <vcf-midatlantic@lists.vcfed.org> wrote:
On 8/21/22 17:07, Douglas Crawford via vcf-midatlantic wrote:
Ah, so the red lines are not b/c/e? That's what I was first thinking. Here me trying to make some sense of it https://i.imgur.com/gOZujMR.jpg Red lines inputs? Could the blue things be capacitors?
I do not think they're capacitors. Remember, this is an IC; think three-dimensionally. All we can see from that picture is the upper metallization layer, where connections to silicon are made. There are several layers below that.
Maybe purple outputs? From this view... OR gates?
Possible.
At least I have some agreement now that several learned eyes see transistors. It might be enough for our purposes to just have a good guess of how many transistors there are, and be able to identify them. What they are doing is a less important detail for our purposes at this point in the exhibit.
Yes, there are definitely transistors there.
So how many do to think you see in each group? We may never know for sure, but I'll probably run with your number as an estimate.
Well, as I said, we'd need a higher-resolution image with better lighting. We need to be able to see a layer or two below that. I'm guessing there are two or three.
-Dave
-- Dave McGuire, AK4HZ New Kensington, PA
Perfect. 2 or 3 is a close enough for now. That's 12 to 18 transistors, estimated range. Calling for someone that can image a die! Thanks Dave. On 8/21/2022 5:32 PM, Dave McGuire via vcf-midatlantic wrote:
On 8/21/22 17:07, Douglas Crawford via vcf-midatlantic wrote:
Ah, so the red lines are not b/c/e? That's what I was first thinking. Here me trying to make some sense of it https://i.imgur.com/gOZujMR.jpg
Red lines inputs? Could the blue things be capacitors?
I do not think they're capacitors. Remember, this is an IC; think three-dimensionally. All we can see from that picture is the upper metallization layer, where connections to silicon are made. There are several layers below that.
Maybe purple outputs? From this view... OR gates?
Possible.
At least I have some agreement now that several learned eyes see transistors. It might be enough for our purposes to just have a good guess of how many transistors there are, and be able to identify them. What they are doing is a less important detail for our purposes at this point in the exhibit.
Yes, there are definitely transistors there.
So how many do to think you see in each group? We may never know for sure, but I'll probably run with your number as an estimate.
Well, as I said, we'd need a higher-resolution image with better lighting. We need to be able to see a layer or two below that. I'm guessing there are two or three.
-Dave
I can image the die. -Dave On 8/21/22 18:11, Douglas Crawford via vcf-midatlantic wrote:
Perfect. 2 or 3 is a close enough for now. That's 12 to 18 transistors, estimated range.
Calling for someone that can image a die!
Thanks Dave.
On 8/21/2022 5:32 PM, Dave McGuire via vcf-midatlantic wrote:
On 8/21/22 17:07, Douglas Crawford via vcf-midatlantic wrote:
Ah, so the red lines are not b/c/e? That's what I was first thinking. Here me trying to make some sense of it https://i.imgur.com/gOZujMR.jpg
Red lines inputs? Could the blue things be capacitors?
I do not think they're capacitors. Remember, this is an IC; think three-dimensionally. All we can see from that picture is the upper metallization layer, where connections to silicon are made. There are several layers below that.
Maybe purple outputs? From this view... OR gates?
Possible.
At least I have some agreement now that several learned eyes see transistors. It might be enough for our purposes to just have a good guess of how many transistors there are, and be able to identify them. What they are doing is a less important detail for our purposes at this point in the exhibit.
Yes, there are definitely transistors there.
So how many do to think you see in each group? We may never know for sure, but I'll probably run with your number as an estimate.
Well, as I said, we'd need a higher-resolution image with better lighting. We need to be able to see a layer or two below that. I'm guessing there are two or three.
-Dave
-- Dave McGuire, AK4HZ New Kensington, PA
If no one local offers, then OK! On 8/21/2022 6:43 PM, Dave McGuire via vcf-midatlantic wrote:
I can image the die.
-Dave
On 8/21/22 18:11, Douglas Crawford via vcf-midatlantic wrote:
Perfect. 2 or 3 is a close enough for now. That's 12 to 18 transistors, estimated range.
Calling for someone that can image a die!
Thanks Dave.
On 8/21/2022 5:32 PM, Dave McGuire via vcf-midatlantic wrote:
On 8/21/22 17:07, Douglas Crawford via vcf-midatlantic wrote:
Ah, so the red lines are not b/c/e? That's what I was first thinking. Here me trying to make some sense of it https://i.imgur.com/gOZujMR.jpg
Red lines inputs? Could the blue things be capacitors?
I do not think they're capacitors. Remember, this is an IC; think three-dimensionally. All we can see from that picture is the upper metallization layer, where connections to silicon are made. There are several layers below that.
Maybe purple outputs? From this view... OR gates?
Possible.
At least I have some agreement now that several learned eyes see transistors. It might be enough for our purposes to just have a good guess of how many transistors there are, and be able to identify them. What they are doing is a less important detail for our purposes at this point in the exhibit.
Yes, there are definitely transistors there.
So how many do to think you see in each group? We may never know for sure, but I'll probably run with your number as an estimate.
Well, as I said, we'd need a higher-resolution image with better lighting. We need to be able to see a layer or two below that. I'm guessing there are two or three.
-Dave
Mystery allegedly solved. New testimony from Corey says this was a Bell Labs demonstration vehicle. Not necessarily intended to ever be a real IC. On 8/21/2022 6:43 PM, Dave McGuire via vcf-midatlantic wrote:
I can image the die.
-Dave
On 8/21/22 18:11, Douglas Crawford via vcf-midatlantic wrote:
Perfect. 2 or 3 is a close enough for now. That's 12 to 18 transistors, estimated range.
Calling for someone that can image a die!
Thanks Dave.
On 8/21/2022 5:32 PM, Dave McGuire via vcf-midatlantic wrote:
On 8/21/22 17:07, Douglas Crawford via vcf-midatlantic wrote:
Ah, so the red lines are not b/c/e? That's what I was first thinking. Here me trying to make some sense of it https://i.imgur.com/gOZujMR.jpg
Red lines inputs? Could the blue things be capacitors?
I do not think they're capacitors. Remember, this is an IC; think three-dimensionally. All we can see from that picture is the upper metallization layer, where connections to silicon are made. There are several layers below that.
Maybe purple outputs? From this view... OR gates?
Possible.
At least I have some agreement now that several learned eyes see transistors. It might be enough for our purposes to just have a good guess of how many transistors there are, and be able to identify them. What they are doing is a less important detail for our purposes at this point in the exhibit.
Yes, there are definitely transistors there.
So how many do to think you see in each group? We may never know for sure, but I'll probably run with your number as an estimate.
Well, as I said, we'd need a higher-resolution image with better lighting. We need to be able to see a layer or two below that. I'm guessing there are two or three.
-Dave
participants (4)
-
Christian Liendo -
Dave McGuire -
Douglas Crawford -
Sentrytv