Help with LM78xx & 79xx series voltage regulators

Semiconductor Integrated Circuits (ICs)
, ,
1

I’ve been repairing a number of receivers lately, but I’ve had the same issue happen with a few of them whenever I use the “replacement part” form DigiKey. I’ll use the current receiver I was working on, because it was working 98%, every input, the display, knobs, etc all working.

However the audio was a bit off. I first tested it by plugging in some headphones and while there was sound, at low volumes it was crackling/distorting. (What type of crackling? I’m not sure if you’ve heard older stereos when you powered them off they would continue to play music for a few seconds (as the capacitors discharged) and as it drained the audio would begin to distort before it was gone).
Well this Pioneer Elite 82thxi was doing that *but while on, I mean I could hear the music fine but there was obvious distortion, sounding like it was just not quite getting enough power. Also another indicator was at low volumes this distortion was worse and even cut out the left earbud, but turning the volume up the louder it went the more both L/R came in clearer. So I thought I need to check the power, SADLY this unit was not designed well enough that you could test the voltages with the top off. So I had to dissemble the entire unit. I did this and got to the board (and of course it’s connected with two 1 inch long cables each with 20pins) so I still could not test it in circuit - I had to remove it. I see the typical “JRC 78m05” and “JRC 78m12” and “79m12” etc series voltage regulators. So I popped one out which was a 78m12 and I test it putting a 20v DC PSU it was outputting 11v instead of 12v, then I tested the others and many of them were not outputting the correct voltages. So knowing that the “current replacements” on DigiKey are the LM78xx LM79xx series I went ahead and ordered the replacements from DigiKey. FWhen I got them I installed them right away (checking my soldering was perfect with no shorts/brdges) it was good. Then I put the unit all back together and I double and tripple checked everything was plugged in and in the right place, then hooked up 2 speakers and also the same headphones.
When I powered it on the relay clicked like it was about to boot up, but the display did not come on… and I thought hmm odd. I pressed power it clicked off, I pressed power again and it repeated the same thing the screen did not light up (seemed like the power on self test failed so it did not fully boot). I then checked the fuses to find 3 of 5 were now blown.

This is REALLY strange to me because it this entire receiver was working properly-ish prior to me replacing these voltage regulators. AND everything was plugged in correctly so I wonder why would the fuses blow like that?

I’ve had this same issue with other receivers I repaired recently where the fuse blows after I replace the voltage regulators. And suddenly it dawns on me… the original are JRC labeled voltage regulators (such as the) “78m05, 78m12 and 79m05 etc” are all black housing units (including the backs). Where as the Digikey replacements/equivalents are replaced with a metal back.

I thought I wonder if this different heat dissipater could not be the problem? I looked the data sheets up and it turns out the metal backs are connected to “base” on the 78xx series, and the “collector” on the 79xx series. AND all these receivers I installs these in that have “blown a fuse” ALL have these transistors mounted to the chassis for heat dissipation (aka common ground) meaning the backs of these voltage regulators are now connected to the chassis ground… Also the “voltage regulator board” is isolated from the ground/chassis. As the PCB itself is does NOT have any “metal ground pads” where it mounts to the chassis. So the PCB itself is isolated from “grounding to the chassis”… so I’m really thinking the entire “voltage board” including the voltage regulators are meant to be isolated from the chassis ground (especially since the original JRC’s do not have a metal back). (My thinking here is that the chassis ground is used by the amp/audio portion and thus the voltage for power for the screen LED’s etc needs to be on it’s own isolated ground? Because it’s not meant to be shared on the same ground as the audio portion which is connected to the RCA grounds etc??)

So maybe I keep installing these and the receivers are designed to use ground isolated back versions of these voltage regulators… there for this might explain why I keep blowing fuses!?

If this is the case then what a dumb “replacement” part if it’s not actually compatible, and also does not warn you “hey we’ve changed the design so you might want to isolate the back pads”

*Also as a bonus how do you test a 79xx negative voltage regulator? It seems when I feed them + voltage they heat up and fail? (and yes I know the pins are swapped from the 78xx)

1 Like
2

I meant to attach this image! (Kind of lame this forum times out edits/updates)

78M12A vs LM7812
78M12A vs LM78121920×984 144 KB

1 Like
3

Looks like the Japanese Radio Company made a version with an isolated thermal connection instead of the original TO-220 LM78xx/79xx standard of a conductive thermal connection.

To work around this you can use a TO-220 mica insulator between a standard 78xx and the heatsink.

Used to see these used a lot in 70s/80s audio gear but usually on transistors because the gear I repaired had discrete power supplies with far better specs than the lowly 78xx/79xx series can achieve.

I just noticed that ST is selling it both ways, here’s the insulated TO-220 package, much cheaper than a mica insulator.

You test them by feeding them negative voltage not positive, see example circuit/test circuit in data sheets.
See section 4/page 6

3 Likes
4

Well thanks for that! These ARE the ones I should have ordered! (kind of wasted my $ buying wrong cross reference part #) I had searched for the “L78xxA” and was recommended the LM series as it’s replacement.

Excellent, I found ALL of the 78/79’s that I needed in this insulated package and added them to my cart! Thanks again! (HOPEFULL the circuit protection in these two receivers preventing anything else from blowing!)

Also testing the 79xx with negative voltage is what I was assuming I needed (reading the specs) but I had seen a few YT videos supplying positive voltage to test them (and yet when I did this, all that did was blow them). I’ll have to look a simple negative voltage circuit for testing these in the future!

5

Be careful it’s not the LM vs L that makes the difference, the L7912CV has a non-insulated metal tab.

The parts are all 79xx series. Depending on the manufacturer, the prefix can be, L, LM, NCP, NJM, KA, SG, UA, MC, and probably others.

You need to check the manufacturers data sheet to be certain you are buying the package variation you need. Historically and possibly still, in addition to the common TO-220 package they have also been made in TO-3, TO-39, TO-257, and a variety of surface mount packages too.

6

Right.
However the original JRC’s are not LM79xx but L79mXX… (as seen in photo I provided)
Also when looked for the data sheets previously for the “79mXXA” and “78mXXA” I only found data sheets for “CP” not “A” so it’s kind of hard to compare specs on paper however these look more like the originals, so all I can do is assume since I was not able to find the exact data sheets on the original #'s

7

An “M” version used to be available in all the prefix versions, it stood for having a 0.5A instead of the original 1.0A rating (now standard units often extend to 1.5A).

With parts from ST, a “P” suffix means insulated tab and a “V” suffix means non-insulated tab. There is usually no direct equivalent lettering scheme for package style between manufacturers.

To summarize standard 78xx/79xx series linear regulator part numbering:

  • 8 = positive and 9 = negative
  • An “M” in between the 78/79 and the “xx” indicates it’s a lower power version
    (was less expensive when introduced, circuits can usually use the 1.0A version instead)
  • Prefix can be any combination of 2 to 3 letters depending on the manufacturer
  • LM was the prefix from the inventors of the part, National Semiconductor
  • The suffix characters can have an enormous amount of variation, are manufacturer specific, and can designate many other parameters including package, temperature range, etc.
8

Thank you for your inputs. Sadly without the original data sheets I am not fully able to find out the original specs. (Also the receiver manuals don’t give you specs on these just part #'s)

Could I “push” you towards answering one last quick question? (I am trying to get an order in with Digikey so it goes out Monday morning)
But I am not 100% sure which these are:

1 Like