Beware STK4278L lot# E422

[[email protected]]

Changed one twice and it was pretty obviously the fault because the
original was intermittent. Both of the replacements from lot # E422
caused horizontal nonlinearity in the left side.

It was impossible to converge, kinda wish I hadn't tried.

It's not clipping, the DC controls move the raster evenly over the
whole screen, nor is there any significant DC offset. I believe the
problem is an insufficient slew rate (high frequency response), so some
may be better than others. Some may converge but I'd rather have an in
spec part.

This happened in a Sony EXR series projection, they and similar Sonys
are the only sets I've ever seen these ICs in, if there are others the
ICs might work, but I dont see how. It's obvious they're not going to
work in these. So don't touch those dials and don't bother
troubleshooting the waveform generator.

Last time this happened to me it was nowhere near as severe and with a
scant ½" of extra overscan the set looked good. This one would've
needed over a full inch, I can't do that.

I thought it worth mentioning because there are alot of these sets
around and most have strong CRTs. What's more, I don't think these sets
get more than ½ a dozen different problems, H tuning cap/Hout and H
regulator, the occasional flyback, diodes in the SMPS/2SC4582. The only
really bad thing I've found is those %&*#$ SRS modules and the set is
good enough I think to use an external sound system. I've only seen one
bad CRT in them ever. When they first came out I thought they were a
bitch to work on, but more recent designs have caused be to do an about
face.

JURB

 

[Mark D. Zacharias]

Changed one twice and it was pretty obviously the fault because the
original was intermittent. Both of the replacements from lot # E422
caused horizontal nonlinearity in the left side.

It was impossible to converge, kinda wish I hadn't tried.

It's not clipping, the DC controls move the raster evenly over the
whole screen, nor is there any significant DC offset. I believe the
problem is an insufficient slew rate (high frequency response), so some
may be better than others. Some may converge but I'd rather have an in
spec part.

This happened in a Sony EXR series projection, they and similar Sonys
are the only sets I've ever seen these ICs in, if there are others the
ICs might work, but I dont see how. It's obvious they're not going to
work in these. So don't touch those dials and don't bother
troubleshooting the waveform generator.

Last time this happened to me it was nowhere near as severe and with a
scant ½" of extra overscan the set looked good. This one would've
needed over a full inch, I can't do that.

I thought it worth mentioning because there are alot of these sets
around and most have strong CRTs. What's more, I don't think these sets
get more than ½ a dozen different problems, H tuning cap/Hout and H
regulator, the occasional flyback, diodes in the SMPS/2SC4582. The only
really bad thing I've found is those %&*#$ SRS modules and the set is
good enough I think to use an external sound system. I've only seen one
bad CRT in them ever. When they first came out I thought they were a
bitch to work on, but more recent designs have caused be to do an about
face.

JURB


Maybe counterfeits? Who are you buying them from?

Mark Z.

 

[[email protected]]

They came from a local supplier because we wanted them fast. We have
had problems with some of their parts, but not all.

I don't suppose their going to reveal their sources, but if anyone on
this NG runs into some E422s that could point the way.

In a way I have trouble understanding counterfeit ICs, somebody goes to
all the trouble to make the series of masks, get everything and the
process going and has a run of ICs that meet all the JANTX specs except
slew rate ? I just don't think this likely, but that doesn't mean it's
impossible.

I think it's more likely that the actual manufacturer nade them, found
them under spec and sold them for scrap, then the scrapper, using some
testing equipment tests them, but can't do it under actual operating
conditions, deems the IC good and then sells it.

To have this crappy slew rate or the wierd distortion in these ICs
which actually do work I'll bet the substrate is good, and I'll lay
money on that they bought or produced some crappy PNP output devices.
Whenever I ever saw this happen it was on the left side of the screen.
In silicon, PNP devices are harder to produce, I don't remember the
reason right now but I think it has to do with the dopants. The
opposite was true of germanium devices.

Yes indeed the ouptputs are semi-discreet in these ICs, once you get
one that's really stuck to the heatsink and the cover pops off as you
hammer it off of there you'll see what I mean.

If I were to make really counterfeit ICs there would be nothing in
there. I would sell IC packages without a chip inside, if someone
really wanted to make a bunch of money fast why go through all the
bullshit. Billy Joe told Bobby Sue to go on take the money and run.
Doing this along with the hierachy of distributors etc., one could make
a nice piece of change before TSHTF and be gone. The difference is that
these things work, almost.

A whole different possibility is that the real manufacturers are
deliberately polluting the market. I wouldn't put it past them, not
after seeing their products.

JURB

 

[Jason D.]

I think it's more likely that the actual manufacturer nade them, found

them under spec and sold them for scrap, then the scrapper, using some
testing equipment tests them, but can't do it under actual operating
conditions, deems the IC good and then sells it.

To have this crappy slew rate or the wierd distortion in these ICs
which actually do work I'll bet the substrate is good, and I'll lay
money on that they bought or produced some crappy PNP output devices.
Whenever I ever saw this happen it was on the left side of the screen.
In silicon, PNP devices are harder to produce, I don't remember the
reason right now but I think it has to do with the dopants. The
opposite was true of germanium devices.

Yes indeed the ouptputs are semi-discreet in these ICs, once you get
one that's really stuck to the heatsink and the cover pops off as you
hammer it off of there you'll see what I mean.

JURB

Jurb is right, the STK is easy to break open and everything is potted
in clear soft silicone, and the amp dies on crunk of copper cubes
covered with either clear silicone or black hard potting.

Not that hard to reverse engineer by probing and taking apart a bunch
and seperate hybrid board in pieces to get precise measurements on
certain printed on resistor strips.

Same with older switching STR ICs (metal baseplate) like used in
Zenith chassis.

Still, I still really think the design of package is very poor
thermal-wise. When one bolts down a STK IC (two screws), the IC bows,
middle part gets thick layer of silicone heatsink compound plus the
plate is aluminum instead of copper. I'm talking about STK 730's, STK
372's, etc. I have to replace many in samsung units. Plus samsung
also cheapen out using STK 372's in place of convergence rings to keep
stress levels less.

Sigh.

Cheers, Wizard

 

[Mark D. Zacharias]

Jurb is right, the STK is easy to break open and everything is potted
in clear soft silicone, and the amp dies on crunk of copper cubes
covered with either clear silicone or black hard potting.

Not that hard to reverse engineer by probing and taking apart a bunch
and seperate hybrid board in pieces to get precise measurements on
certain printed on resistor strips.

Same with older switching STR ICs (metal baseplate) like used in
Zenith chassis.

Still, I still really think the design of package is very poor
thermal-wise. When one bolts down a STK IC (two screws), the IC bows,
middle part gets thick layer of silicone heatsink compound plus the
plate is aluminum instead of copper. I'm talking about STK 730's, STK
372's, etc. I have to replace many in samsung units. Plus samsung
also cheapen out using STK 372's in place of convergence rings to keep
stress levels less.

Sigh.

Cheers, Wizard

We used to buy generic STK's from Andrews Electronics. Some of them were so
poorly made the legs fell off right there in the parts bag.

The transistors I bought from them were "2SC3182 - 2SA1265" and "2SC3281 -
2SA1302" .

All were counterfeits in SK bags but with "Toshiba" printed on the parts.
Only thing is, Toshiba had discontinued them long ago. When the parts failed
under even a light load, we busted them open. The interior finish quality
was inferior, and the semiconductor dies were only 2.5mm instead of 5mm
which the originals were.

Nowadays we only buy semi's from the original manufacturer or B & D
Enterprises, who is a distributor for Sanken, etc and whom we trust.

Mark Z.

 

[John-Del]

Mark D. Zacharias wrote:
rd

We used to buy generic STK's from Andrews Electronics. Some of them were so
poorly made the legs fell off right there in the parts bag.

I had the same problem with semis (power devices) from Tritronics.
Pure garbage. I buy most of my semis in the original packages direct
from Panasonic, Sony, Toshiba, Hitachi, Mitsubishi, etc. They do cost
a lot more, but NOTHING is more expensive than a re-check. When I
repair something, I don't ever want to see it again, at least not for
the same problem....

I also buy semis from Electronica in Denver. I talked to the owner who
swears he buys directly from the OEMs. In three years dealing with
them, I haven't had one failure. I heard from others that B&D
electronics also has good devices, but I have no experience with them.

John

 

[Mark D. Zacharias]

Mark D. Zacharias wrote:
rd


I had the same problem with semis (power devices) from Tritronics.
Pure garbage. I buy most of my semis in the original packages direct
from Panasonic, Sony, Toshiba, Hitachi, Mitsubishi, etc. They do cost
a lot more, but NOTHING is more expensive than a re-check. When I
repair something, I don't ever want to see it again, at least not for
the same problem....

I also buy semis from Electronica in Denver. I talked to the owner who
swears he buys directly from the OEMs. In three years dealing with
them, I haven't had one failure. I heard from others that B&D
electronics also has good devices, but I have no experience with them.

John

B & D are great people, and the product is first-rate. I'll keep Electronica
in mind.

Mark Z.