What's new
What's new

Electronic gurus - Why are "hybrids" so hard to find ?

Milacron

Administrator
Staff member
Joined
Dec 15, 2000
Location
SC, USA
I need HC611002G and HC014701G.... this from 1991 vintage CNC control. Unlike even the most obsolete and obscure DIP IC, Google brings up nothing on these part numbers...

not even the usual fake electronic part sites offering to "quote" you in the hopes "you" are a huge company looking for 100,000+ pieces.

I've had this happen before with hybrid IC's...why are these little buggers so hard to find replacements for ? Does someone specialize in them ?


http://en.wikipedia.org/wiki/Hybrid_circuit
 
Milacron
Quote
"Google brings up nothing on these part numbers"

It does now .
1 hit !
Your request for the part in this forum !

Davycrocket
 
Milacron
Quote
"Google brings up nothing on these part numbers"

It does now .
1 hit !
Your request for the part in this forum !
LOL... I've had that happen many times...PM is so large now that Google picks up on things here immediately it seems like.

Also had it happen in similar vein when researching obscure machine models coming up at auctions and all I get is a link back to the same damn auction ! :angry:
 
custom parts

I need HC611002G and HC014701G.... this from 1991 vintage CNC control. Unlike even the most obsolete and obscure DIP IC, Google brings up nothing on these part numbers...

not even the usual fake electronic part sites offering to "quote" you in the hopes "you" are a huge company looking for 100,000+ pieces.

I've had this happen before with hybrid IC's...why are these little buggers so hard to find replacements for ? Does someone specialize in them ?


Hybrid integrated circuit - Wikipedia, the free encyclopedia


In short most(not all, but most) hybrids are custom parts, made to customers drawing for 1 customer.

Some electronics suppliers have their own part number for every part they use, for one of 2 reasons:
1) so you have to buy replacements from them, at a big markup
2) so they can control parts parameters and testing ( IC suppliers change part spec and performance occasionally, big problem when you order a new batch of parts and your design not longer works the same, DAMHIKT)

CarlBoyd
 
Wow are those ever obscure. I was in the electronic parts business and used all my databases and came up empty! Who was the manufacturer, and are you sure it is the chips that failed? HC prefix leads me to believe it is a high speed Schottky device popular in the late 80's. Try calling tech support at one of the large component distributors and you may get someone to take up the challenge. Also post a photo showing the part number and any markings on the chip. Sometimes we could trace them back that way thru the manufacturer or distributor. Lastly the could be "blond chips" a set of wire interconnects to vary the options available on a board. I had to have a bunch made up special for a customer once.

Peter
 
Hybrids are usually custom devices with small production runs, think of a hybrid as a miniature PCB.

They are a PITA many older VFD's use a hybrid driver board which fails occasionally consigning the drive to the parts pile.
 
Milacron,

Keep us informed as to resolution of the problem, I am curious if you:

Trash the electronics and go with something else.
Reverse engineer the HC.
Set the challenge aside, for further research and 'another time'.
Get really lucky and find the components.

What needs the component?

Steve
 
The function of an ancient hybrid may now be replaced by stock chips. It is possible to cut the can open and reverse engineer what the functionality is. I do have contacts in the biz. What part of the control is this in??
 
Not only are they made to one customer's design, they are made to solve a particular problem in a way that only one customer may even want done, and that on one product only. Chances of finding another are about as good as finding a different manufacturer's headstock that will perfectly fit the particular lathe in question.

You will have to find another PWA with the same part on it. Not crazy, since the existence of a hybrid usually means higher volume production.

In many cases you can't tell what it is supposed to do, because there isn't a spec on it outside the files of the control maker, who has no particular incentive to help you with that ancient thing, and probably has a data destruction policy anyhow, or the part manufacturer, who can't tell you due to an NDA, AND has a data destruction policy to boot.

I tried to find info on a particular piece of equipment for a client. Found the company, they knew just what it was, and still make a (rather different) version of it. But their data from the early 1980s was "too hard to find without the original customer's name", so no go, as we didn't know that name.
 
According to magazine articles Field Programmable Gate Array (FPGA) and related devices can provide an "inexpensive" substitute for obsolete hybrids. Assuming that the requisite performance data on the device to be replaced is available and that a FPGA base chip exists with a suitable combination of analogue and digital elements. As I understand it the idea is a program on demand service. Obviously the device is a black box "right inputs and outputs" substitution not identical process so anything depending on careful matching of duty to process characteristics for correct operation is gonna be trouble. Carefully tuned analogue stuff running close to technology limits will be risky.

Inexpensive is a pretty moveable feast and, in this case, probably biased more to the "loosing production worth umpty four thousand a day until the machine is fixed" scenario than to "getting this heap running so I can sell it".

Clive
 
Wow are those ever obscure. I was in the electronic parts business and used all my databases and came up empty! Who was the manufacturer, and are you sure it is the chips that failed?
Not only am I not sure, I suspect most likely the hybrids are fine. I only wanted a source of them "just in case" one of them turns out to be an issue here. I've had hybrid trouble before on other equipment, but previously it was corroded leg/contact related, so I have yet to actually replace a hybrid. But what a cringer to find out that if one ever does go, the only possible fix would be to find another identical board to harvest used ones from. :angry:

It's interesting that some of the hybrids on a "sistership" board appear like a thin absolutely flat sheet of ceramic with 8 contacts... I can't imagine what is "in there" and how they got it in there, the thing is so thin...like guitar pick thin.
 
Old hybrids can be a worry since they often have more exposed wire bonds inside, think 0.5 mil Al wires. If the seals leak this can lead to corrosion of those wires. There is no way to inspect, and as you have discovered it is very difficult to impossible to back engineer one.

The very thin ones might simply be resistor networks.

--Michael
 
Old hybrids can be a worry since they often have more exposed wire bonds inside, think 0.5 mil Al wires. If the seals leak this can lead to corrosion of those wires. There is no way to inspect, and as you have discovered it is very difficult to impossible to back engineer one.

The very thin ones might simply be resistor networks.

--Michael

Or a tamperproof 'dip switch'?

Is it possibly a 'dongle' to lock software to hardware?
 
Yep, hybrids are customer-specific. You're only choice is to cannibilize another unit or reverse-engineer them (really expensive). PALs are a great choice if they're logic-only.
 
FWIW, here's some pictures. The first two pix show a "sisterboard" that is identical except the hybrids have different part numbers from the numbers I listed in the beginning of this thread. Note the "guitar pick thickness" white hybrid is apparently doing the same function as the longer black hybrid in the last photograph. Second photo shows the backside of the white hybrid. The last photograph is of the problematic board. Does the logo on the black hybrid ring any bells ?

IMG_0971.jpg


IMG_0973.jpg



IMG_0977.jpg
 
The Mitsubishi spindle and servo drives used on Mazaks back in the 80's and 90's utilized hybrids. These pesky hybrids are often the cause of failures on these drives. A typical Freqrol spindle drive has 12 of them!

Mitsubishi Electric Automation still has them, but will not sell them to anyone. You have to send the entire board in for repair...and even then they only replace the bad hybrids. Sheesh, after 20 years of service you would think they would replace them all!

But alas, I have collected a supply of spare spindle and servo drives (via my old friend Ebay), primarily for the hybrids.

Greg
 
The white and black 10-pin SIP packages are NOT hybrid ICs. They are resistor networks, commonly used for pull-up or pull-down resistors on logic circuits.

The ones in your photo appear to contain 9 individual 4.7k resistors, with one common lead.
 
My money is on the RN10 being a resistor network maybe 4.7k used as a pullup on inputs. The 4701 on the black one could also be 4700 ohms. 470 + 1 zero. Seen that before.
John

Edit, beaten by a couple of minutes, N2Ixk types faster. :)
 
The white one is a Common Bus 4.7K resistor array (SIP package, single inline pin). The rear view clearly shows the common bus.

two main configurations you'll find;
Common Bus: pin 1 common, pin 2 thru 10 each being a 4.7k resistor.
Isolated resistors: pin 1-2 being a resistor, pin 3-4 being another, etc.
There can also be more complex arrays made.

The black one may also be a 4.7K or a 470 ohm array. A resistance measurement will determine its value, and makeup.

[edit: hazards of opening too many tabs]
 
My money is on the RN10 being a resistor network maybe 4.7k used as a pullup on inputs. The 4701 on the black one could also be 4700 ohms. 470 + 1 zero. Seen that before.
John

Yes, both are resistor networks as pointed out by others. I agree with John on 470 + one zero = 4k7 = 4.7k; this four-band value coding is becoming increasingly common. I think that 'G' is the code for 2% tolerance, though usually the four-band coded resistors are 1% or better.

The number 8923 will be a manufacturing date code - week 23, year 1989. This is consistent with 9009, 9010, 8801 and others visible in the photos, and Don's equipment date of 1991. ICs, transistors and capacitors often carry date codes, which can be useful in dating the circuit boards they are used in. You can easily test the resistor values with a multimeter. The likelihood of their being faulty is very small.

George
 
Last edited:








 
Back
Top