12
\$\begingroup\$

Another question recently asked about the MC7805, and linked to its the datasheet for the generic MC7800-series of regulators1.

When I read it, I noticed Figure 1, Representative Schematic Diagram, and there is a construction that I can't understand, which I have highlighted in the following image:

enter image description here

As you can see, there are a bunch of resistors, all of them bypassed! I know that the diagrams in these datasheet are only representative, so I suppose my question is:

What is this supposed to represent?

My guess is that it's related to one of the protection modes: current limit, thermal shutdown, and/or safe-area compensation, but I can't understand how - the resistors are apparently completely and permanently bypassed.


1) I originally assumed it was specific for the 7805 variant. Brian Drummond pointed out in a comment to an answer that it is generic, which makes the purpose of the resistors easier to explain.

\$\endgroup\$
  • 1
    \$\begingroup\$ I don't know, so I'll speculate in a comment. They might be laser-zap links, for trimming the input impedance on test. \$\endgroup\$ – Neil_UK Jul 3 '16 at 12:58
  • 3
    \$\begingroup\$ From the fact that the bypassed 18k resistor is on a node marked sense, I think it may be different resistors that are either bypassed or not to achieve other 78xx series parts with a minimal process change (just open up the bypassing). \$\endgroup\$ – Peter Smith Jul 3 '16 at 13:00
  • \$\begingroup\$ @PeterSmith That's interesting, it could explain why they have a number of different but still regular values, like 3k, 6k, 3k, 9k etc. I still wonder why they wanted to include it in the datasheet. \$\endgroup\$ – pipe Jul 3 '16 at 13:04
  • \$\begingroup\$ The diagram said it was representative and specifically the 5V version; the generic 78xx might well have this. Q12 and Q15 appear to be part of the error amplifier, and varying the resistance will adjust the voltage divider; the 5V device appears to have a feedback divider of R10 and R13. Increasing the feedback voltage would increase the output. \$\endgroup\$ – Peter Smith Jul 3 '16 at 13:09
  • \$\begingroup\$ @PeterSmith, yes that's got to be it. Good shout. \$\endgroup\$ – TonyM Jul 3 '16 at 17:56
17
\$\begingroup\$

These resistors are part of the voltage feedback network. Depending on how many of these resistors are in series (meaning: shorted or not) you get a different feedback network.

The feedback network sets the output voltage. So they use this to make the different 7805, 7806, 7812 etc... regulators from the same silicon design.

Basically they're all "born" as 5 V regulators. Then by "(laser ?) zapping" the shorts across the resistors they can be made into regulators for different output voltages.

\$\endgroup\$
  • 2
    \$\begingroup\$ Notice the datasheet is titled "MC7800" and thus not specific to any particular voltage version. Also, the unbypassed R10 and R11 sum to 15K (presumably for 5V). \$\endgroup\$ – Brian Drummond Jul 3 '16 at 13:13
  • \$\begingroup\$ @BrianDrummond Good catch! That explains a lot. \$\endgroup\$ – pipe Jul 3 '16 at 16:52
  • 2
    \$\begingroup\$ "born" as 5 V regulators" - too cute! \$\endgroup\$ – Andrew Morton Jul 3 '16 at 17:14
2
\$\begingroup\$

My guess would be that the resistors are used for the configuration of the device.

Integrated circuits are made in several process steps the last steps are where the metal interconnections are fabricated.

Voltage regulators are made for different voltage levels and the actual value of the feedback network and thereby the output could be changed as one of the final process steps by changing one or two masks.

If all resistors are short-circuited you would have a generic voltage regulator with a certain divider ratio you would get one of the standard voltages.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.