Re: BU406 vs. BU406D

Matthew Sell wrote:

> (Replies throughout)
>
> At 02:43 PM 3/6/00 -0600, you wrote:
> >Alrighty then, let's talk about this, shall we?
> >
> >NEVER use a 1N4000 series diode in a horizontal output (or in this case a
> high
> >voltage power supply) stage. Usually the diode will overheat and blow
> itself up
> >because it can't turn OFF fast enough to keep from drawing a high reverse
> >current.
>
> Depends on the inductance of the transformer, and the resistance, thus
> determining how quickly the voltage builds when the field collapses. There
> was also no measured rise in temperate above ambient.
>

Partially right. The current is already at maximum through the flyback when the
transistor turns off. Voltage rise time at that point will be determined by the
resonating capacitance and the inductance of the flyback with practically no
effect due to resistance. Fortunately for the 1N4006 the turned on transistor had
it shorted, so it isn't in an on state. Unfortunately for the 1N4006 it is a very
capacitive (as diodes go) part seeing a fast rising REVERSE voltage. It is going
to be conducting current in a reverse mode. This is NOT a good thing.

>
> >
> >When the transistor first turns off the collector will go several hundred
> volts
> >positive (determined by the flyback, resonating capacitors, current
> through the
> >flyback at transistor turn-off time, and added to the power supply voltage to
> >the flyback). At this time the transistor is turned off and the diode
> SHOULD be
> >off, but won't necessarily be if it is a slow part such as a 1N4000 series
> >diode.
>
> >The flyback and the resonating capacitors will resonate at a high
> frequency (a
> >couple of hundred KHz or so), giving you a sine wave that, by itself would
> >slowly decay back to nothing, IF there were nothing to prevent it. But
> instead
> >of the sine wave driving the collector several hundred volts negative (which
> >would break down the transistor by forward biasing the collecor-base junction
> >and reverse biasing the emitter-base junction which will withstand less
> than 10
> >volts without damage) the damper diode turns on. In the case of a BU406D the
> >damper diode is internal. The damper continues to conduct until either the
> >flyback runs out of current OR the transistor is again turned on. In this
> case
> >the transistor turns on first and the whole cycle repeats.
> >
> >The biggest problem using an external damper is finding a place for it.
> Ideally
> >it would be right across the transistor leads as close as possible.
>
> I put mine on the solder side of the board, with the leads as short as
> possible.
>

Good idea, you need to keep the leads short to get the least inductance from the
leads themselves. Any extra inductance will result in excessive voltage across
the diode when it conducts, which is not good for the transistor.

>
> >
> >You say you have about 6 to 8 volts forward on that 1N4006. A BU406
> crosses to
> >an ECG/NTE379 which lists the maximum emitter-base voltage at 9 volts.
>
> Emitter/Base voltage is not the problem. The back-EMF affects the collector
> to emitter junction in this particular configuration. The base of the
> transistor is connected to the emitter through the secondary of transformer
> T1.
>

Emitter/base voltage is EXACTLY the problem. When that diode is supposed to be on
the transistor collector is being driven negative. This turns on the
collector/base junction (something you don't want happening) and almost all of the
voltage is then across a reverse biased base/emiter junction. Although the
collector of a silicon transistor can (in this case, anyway) handle several
hundred volts (and then only when completely turned off), the emiter/base is only
rated for 9 volts (usually 6, this part has higher ratings than most silicon
parts).

And don't count on T1 helping. It has too much inductance to stop a spike, too
much resistance to handle a flyback kickback, and anything that enters T1
secondary is going to be reflected back to its primary where there are parts that
are not rated to handle ANY of this action.

ANY breakdown of an emitter/base junction will degrade the junction. Repeat it
often enough and the part is toast.

>
> And I'd
> >like to know how you were able to measure that 6 to 8 volts. If it were
> with a
> >scope you would have had to have the gain set high enough that the
> positive part
> >(several hundreds of volts) of the waveform would have been saturating the
> >scopes amplifier.
>
> The voltage does not have time to build to several hundreds of volts before
> the diode is able to turn on.
>

Wrong again. The several hundred volts I'm talking about are right after the
output transistor turns off and the flyback and resonating capacitors resonate.
It is this very flyback pulse that produces almost all the voltages from the
flyback transformer. It is because of this very pulse that there used to be
warnings on schematics to NOT measure flyback voltages. It is safer now that we
have solid state output stages. In the tube days the output tube plate could hit
6,000 volts right after turnoff. The best horizontal output transistors are good
for only 1,500 volts and the BU406D isn't even good for half of that. To get the
same power out of a flyback transformer they had to trade peak voltage for peak
current.

The diode turns on, obviously, after the high voltage pulse when the current
through the transformer (and the polarity) reverses as part of the resonance
waveform. My point was that unless the scope is on a high enough range to keep
most ALL of this on screen the vertical amplifier in the scope can be driven into
saturation which will appear to move the entire DC reference (zero voltage point)
of the display. So you have two choices: one is to not saturate the amplifier,
but the part of the waveform you are trying to see is reduced to nothing due to
the high range you are on or 2) try to expand your view by using a lower range,
but getting errors in your reading due to overdriving the scope with the part of
the signal you aren't looking at.

Nothing to be ashamed of here, there are techs that have been in electronics for
years that make similar mistakes. Used to do it myself many years ago. You just
have to know where the limitations of your test equipment are and not pass them.

>
> Once that happens it's hard telling how much of your reading
> >is real and how much is from an overdriven vertical scope amplifier.
> PLUS, the
> >peak current across the 1N4006 is in excess of its 1 amp rating.
>
> What are your calculations regarding the theoretical voltages that are
> obtained using the Wintron transformer in the Amplifone HV board? I'd like
> to spot check what you have obtained with my quick calculations to see if I
> have made a mistake. In particular, I'm curious as to what the calculated
> values of the current through the diode are, and the calculated peak
> voltage obtained before the diode starts to conduct.
>
> I'm not saying that the 1N4006 was the best choice, it isn't. I'm still not
> recommending that diode to anyone else. I am merely supporting an
> alternative to using a BU406D ONLY IF the installer is comfortable with the
> alternative. The next time I take the machine apart for other repairs, I'm
> going to replace that 1N4006 with something more suitable. At the time, it
> was available to me and my initial calculations showed that that diode
> would protect the transistor from damage. Temperature measurements and the
> fact that the HV power supply has been operating correctly to this point
> seem to reinforce my point, although those facts do not completely prove it.
>

Remember back when you said you kept the leads short? Another benefit of keeping
leads short is that is allows the part to use the board as a heatsink. A prime
example of this is the MR750 series diode (a great part to replace overheating
rectifiers on game boards with). The part is rated for 6 amps maximum, UNLESS you
keep the leads short so the part can dump heat out the leads back into the pads on
the board. With short leads and heavy pads on the board the rating can go as high
as 22 amps!

And no, don't use those for dampers either.

>
> Using this diode was my choice on my game. I am comfortable with that
> decision, but recommend it to others with the exclusions I have mentioned.
> If you have contrary calculations, I'd like to see them - it expands the
> knowledge to the entire group. Some day we may not have access to the
> BU406D, and discussions like this will help find an alternate arrangement
> to keep the games going.
>

There are other parts, just many of them in the wrong case style.

Maybe we're trying too hard. At http://www.nteinc.com/press.html is an old press
release about NTE adding original generic parts (but NOT as NTE replacement
parts). If you follow the links it will tell you the BU406D is available through
them, it just has to be ordered under that number! I'll try to ask our local
dealer about this and get an answer back to you.

>
> - Matt
>
> I never met a Windows box I couldn't crash......
>

Neither have I. Not just Windows, though. I also could crash OS/2 with no
effort. DOS, also. Always hoped, for the worlds' sake, it was just me.

>
> ---------------------------------------------------------------------------
> ** To UNSUBSCRIBE from vectorlist, send a message with "UNSUBSCRIBE" in the
> ** message body to vectorlist-request@synthcom.com. Please direct other
> ** questions, comments, or problems to neil@synthcom.com.

--
Windows:
32 bit graphical interface for a
16 bit patch for an
8 bit operating system written for a
4 bit processor by a
2 bit company that can't stand
1 bit of competition.
---------------------------------------------------------------------------
** To UNSUBSCRIBE from vectorlist, send a message with "UNSUBSCRIBE" in the
** message body to vectorlist-request@synthcom.com. Please direct other
** questions, comments, or problems to neil@synthcom.com.