Unknown device "3B" on motor drive boards.

This any use, maybe ?

The section on "HP Diode Coding" looks as though it might contain some relevant numbering to what you have ??

Arfa

On Thu, 15 Sep 2011 15:37:23 -0700, "Robbie Hatley" put finger to keyboard and composed:

If you hadn't identified the component as a dual zener, I would have guessed that it was a PNP transistor with an ECB (123) pinout.

In fact I can't recall seeing a dual diode package where the tab wasn't either the common anode or common cathode. But then my memory isn't the best anymore ...

Can you help us out with a little circuit tracing? The circuitry in that area is typically very simple. In fact, we may be able to work out the zener voltage(s) by knowing the output voltage of the supply plus the values of any feedback resistors in potential dividers, etc.

Could we see a close-up photo?

- Franc Zabkar

--
Please remove one 'i' from my address when replying by email.

HP microwave diodes did not include zeners, as far as I know. Intended for receiver applications, mixers, etc., they don't seem right for

160Vpk applications.

This manual indicates that "3B" could be either a pnp or npn transistor:

The most common item that one sees in a switching supply next to an optoisolator is a TL431 programmable zener (actually a voltage reference/error amplifier combination). Possibly that's what you have, they're multiple-sourced in LOTS of package styles.

Yah, it does look sort of like a transistor, both visually and on meter. However, I have reasons to believe its actually some kind of multiple- zener-diode contraption. More on this below....

I'm guessing pin 2 is common anode, but the two zeners are very different breakdown voltages (like 1V and 5V or something like that), so that the zener from pin 2 to pin 1 conducts both ways on a diode meter, but the zener from pin 2 to pin 3 doesn't.

Many of today's switching power supplies use a transformer switched by a transistor or even directly by a PWM controller IC. The transistor or controller gets feedback from the transformer via an optocoupler (P181 is almost invariably used), with the LED cathode of the optocoupler fed to ground through a back-biased zener diode. Sort of like this appnote for the 2A265 PWM controller:

The zener diode, in practice, is typically pins 2,3 of the mystery "3B" device I'm trying to identify.

For example, one of the circuits this part appears in is the Hyundai N100 motor controller. The power supply board on that looks like this:

Note that if you envision a zener diode in the "3B" device with cathode on pin 3 and anode on pin 2, it becomes much like the appnote I link-to above. (As to what pin 1 is connected to internally, I have no idea yet.)

My camera, even set to "macro", can't give a good image of an SMD. The diagram I linked to in my OP gives a good idea of what it looks like, though:

Oh, and I did find the package name for that shape: its "SO- 89". Here's what the SO-89 package looks like:

(Turns out, the tab and pin 2 are all one piece. Can't see that from top side, but on removing the part, I can see that it's so.)

So now its a matter of finding a SO-89 device with a zener diode from pin 2 to pin 3, and (????) from pin 2 to pin 1.

I've seen this "3B" gizmo on boards from 2 different manufacturers, so I'm assuming its actually a common part in switching power supplies. So I'm surprised I haven't been able to track it down yet.

--
Still puzzled,
Robbie Hatley
Hatley [dot] Software [at] gmail [dot] com

On Fri, 16 Sep 2011 15:49:49 -0700, "Robbie Hatley" put finger to keyboard and composed:

ZD3 in that schematic looks exactly like the TL431 whit3rd alluded to.

A TL431 is effectively a programmable precision zener diode. It has an internal 2.5V reference. The potential divider consisting of R13 and R11 sets the output voltage to ...

Vo = 2.5 x [1K / (1K + 1K)] = +5V

- Franc Zabkar

--
Please remove one 'i' from my address when replying by email.

On Sat, 17 Sep 2011 09:57:33 +1000, Franc Zabkar put finger to keyboard and composed:

Sorry, that should have been ...

Vo = 2.5 x [ (1K + 1K) / 1K ] = +5V

- Franc Zabkar

--
Please remove one 'i' from my address when replying by email.

The legs that conduct both ways and the other one not at all, may indicate it being an Thyristor. SCR/TRIAC etc..

If the value of this two way conduction in down around 100 ohms or less, I would say that is what you have, which would be the Gate-cathode(K) you're seeing. If so, here is some resource you can use on testing thyristors with a DMM.

Just a thought.

Jamie

Hold a hand magnifying lens over the camera lens for an impromptu macro set-up. Take shots at different by 1/4 inch measured spacings for future reference of the exact focus distance.

I lay a circuit board on an old flatbed scanner to get good images, if there are no tall components. If they do, I use my Fuji Finepix S5200 in macro mode.

--
You can't have a sense of humor, if you have no sense.

"Franc Zabkar" wrote in message news: snipped-for-privacy@4ax.com...

I'm with you on that one Franc. They are very common on the TV LCD switchers that I repair quite often. The reason that I had hesitated on this, is that I don't think I have seen one before in that package, with the solder-down tab at the top. However, if there is a version in that package, then I agree that it does seem to be in the right position in the schematic, and the resistors wrapped around it work out very neatly. I have also found that just because it says something like "ZD" as a component identifier, these days, you can't necessarily trust that as being a zener diode. Likewise, you can't always trust symbols on schematics. The designers use CAD packages for preparing PCBs and documentation (when they even bother with that!) and they have a tendency to just use the libraries of parts that come with their package. This means that every now and then, they use a device in their design that has no symbol in the library. Rather than do the proper thing, and either update the library or make a custom symbol, they just use a 'best fit' symbol that's got the right amount of pins and roughly the same function. In this case, "zener" is probably close enough, as the '431 is known as both a programmable zener, and a precision variable voltage source. I have seen this device shown on schematics as both a three pinned rectangle - like a 78xx regulator - and a conventional zener, but with a third connection sticking out of the side at right angles.

Just the other day, I had a case of a mis-drawn device on a schematic. It was in the ignitor circuit for a short-arc discharge lamp. It was shown on the schematic as a thyristor with the gate left unconnected. This made no sense at all in figuring how the circuit worked. When I looked up what the device actually was, it turned out to be a SIDAC - yes, I know, I'd never heard of one either. It is actually a two terminal device, that comes in a variety of packages, including a tabbed thru'-hole three pin version (TO-202), where two pins are joined together. It's correct (manufacturer's) schematic symbol, is a "Z" with a horizontal line through it. Quite similar to a diac. Once this was known, the operation of the circuit became much clearer.

Arfa

Is that schematic out there in wwwland? I was recently trying to find any schematic of the ballast part of such a supply but could only find patent references or nulled-out block diagrams. I was trying to get an idea of the post-trigger start up regime while the lamp is warming up.

Could it be any of these?:

3B- BC856BW Phi SOT323 3B BC856B ITT SOT23 3B BC856BT Phi SOT416 3B FMMT918 Zet SOT23 3B MMBT918 Mot SOT23

On Thu, 15 Sep 2011 15:37:23 -0700, "Robbie Hatley" put finger to keyboard and composed:

FWIW, Hitachi has a HA17432VLTP device with a "3B" marking code in an MPAKV package:

It is designated for "industrial use".

Its pinout is K-A-Ref whereas the pinout of the HA17431VLTP device (marking 3A) is Ref-A-K.

Specifications:

- Franc Zabkar

--
Please remove one 'i' from my address when replying by email.

The manuals are on the manufacturer's website, but that doesn't actually help, as the schematic for the version of ballast that shows the ignitor, is not included in any of the models' manuals. We had to get that one direct from them. I have it printed out, but for some reason, I don't seem to have it saved away anywhere. I have been through the rest of my manual versions that I have got saved, and have found another schematic that has the same ignitor on it, but with the SIDAC (almost) correctly shown as a diac. If you care to mail me direct off-group with a current valid address that you receive and read mail on, I'll send it to you.

I have discovered by being involved in repairing them, that there is actually no great mystery to how the lamp starts up. I too initially thought that it was a carefully controlled start-up procedure, but it's actually not, and seems to rely totally on the characteristics of the lamp itself. Basically, a current-limited AC voltage of something in excess of 100v RMS is applied across the lamp, and at the same time, a series of striking pulses of several kV are superimposed on top of that. This causes the basic arc to strike, and the AC voltage drops to around 25 volts, which seems to be enough to maintain the arc through the gas in the discharge tube. By whatever means - in the case of this particular one by use of a thermal switch heated by a power resistor in the RC network of the pulse generator - the ignition pulses are stopped, leaving just the AC across the lamp. As the lamp starts to warm up, the full plasma discharge starts to form in the arc tube, and as this builds, the impedance of the lamp increases, resulting in the AC voltage across the lamp increasing, until it settles at around 90 to

100 V with the lamp at full operating temperature. This appears to be purely down to the lamp, and nothing to do with the ballast, whose sole job appears to be to limit / control the current in the lamp so that it burns within its rating. The max hot running voltage does tend to drop as the lamp ages, and does vary a little between manufacturers. I have seen voltages ranging between about 80 and 100 volts. I guess it's a bit like a neon bulb where you get a fixed voltage across it of around 90 volts, irrespective of the applied voltage on the back side of the current limiting resistor.

A couple of weeks ago, I had a similar lighting fixture from a different manufacturer on the bench. It used exactly the same short-arc discharge lamp, but had no electronic ballast. Instead, it had a simple passive series inductive ballast, fed directly from the mains. The ignitor for this one was a sealed unit that 'floated' on the lamp side of the choke, so this further reinforced my belief that the electronic ballasts do nothing 'magic' and it's just a case of arcing up the lamp initially, and then limiting the current through it, leaving it to otherwise just 'do its own thing'.

I have a very good pdf on metal halide discharge lamps and how they work, which I could also send you, if you are interested.

Arfa

AH, I think that's it or something very close! Hitachi even specifies that the part is intended to be used in switching power supplies, in conjunction with an optocoupler in the feedback circuit, so it's a very probable match.

The device on the boards I've been working on is SO-89, which Hitachi is calling "UPAK". Any of their "U" versions of the

431 & 432 have this package, with marking codes 4A, 4B, etc. So this is probably the same basic kind of part, but a different manufacturer, using 3B as marking code for the SO-89 version.

Thanks for the tip!

--
Cheers,
Robbie Hatley
Stanton, CA, USA
lonewolf (at) well (dot) com
http://www.well.com/user/lonewolf/

Yep, I think that's probably it! Thanks! (Also see my reply to "Franc Zabkar". I saw his reply first. He mentioned the Hitachi version of 431, 432.)

--
Cheers,
Robbie Hatley
Stanton, CA, USA
lonewolf (at) well (dot) com
http://www.well.com/user/lonewolf/