X-Virus-Scanned: clean according to Sophos on Logan.com Return-Path: Sender: To: lml@lancaironline.net Date: Thu, 25 Oct 2012 09:21:41 -0400 Message-ID: X-Original-Return-Path: Received: from qmta14.westchester.pa.mail.comcast.net ([76.96.59.212] verified) by logan.com (CommuniGate Pro SMTP 6.0c2) with ESMTP id 5841077 for lml@lancaironline.net; Wed, 24 Oct 2012 18:04:09 -0400 Received-SPF: pass receiver=logan.com; client-ip=76.96.59.212; envelope-from=jmorgan1023@comcast.net Received: from omta14.westchester.pa.mail.comcast.net ([76.96.62.60]) by qmta14.westchester.pa.mail.comcast.net with comcast id F9bz1k0031HzFnQ5EA3fLU; Wed, 24 Oct 2012 22:03:39 +0000 Received: from [192.168.1.105] ([24.11.157.196]) by omta14.westchester.pa.mail.comcast.net with comcast id FA3e1k00W4EXR5U3aA3ekd; Wed, 24 Oct 2012 22:03:38 +0000 From: Jack Morgan Mime-Version: 1.0 (Apple Message framework v1283) Content-Type: multipart/alternative; boundary="Apple-Mail=_DEEDD330-7E3E-46CB-90E1-C59499C5A60A" Subject: How alternator diodes do their thing X-Original-Date: Wed, 24 Oct 2012 18:03:33 -0400 In-Reply-To: X-Original-To: Lancair Mailing List References: X-Original-Message-Id: <7D24A565-0056-4796-810E-9927EC84FAF6@comcast.net> X-Mailer: Apple Mail (2.1283) --Apple-Mail=_DEEDD330-7E3E-46CB-90E1-C59499C5A60A Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset=iso-8859-1 Hi all, I was with Motorola during the original design cycle for the first = alternator diodes for Chrysler. Shouldn't admit to being that old but I = can't resist putting in two cents here. I subsequently spent more years = than I will admit (with many) bringing electronics into the automobile = from the first alternator through the electronics evolution resulting in = more than 30 micro controllers in the basic car. It was a fun ride... = obviously retired now. Brent's description is correct... as the three windings produce their = sinusoidal outputs (each 120 degrees out of phase with one another) only = one winding at a time is producing the voltage required to cause = conduction through it's respective diode pair. I will leave the wye = versus delta differences to those who care. Brent's description is = easiest but you engineering types may want to draw three sine waves on a = piece of paper..... the tops of the sine waves show the ripple that is = present on an open circuit alternator. The alternator is actually a = current generator so there is actually negligible noise created by this = voltage ripple since it is clamped by the battery and the phases = smoothly take turns supplying the output current as controlled by the = field current. Thus no need for a large filter capacitor. The reason = that a failed diode created radio noise is that the smooth transfer of = current is disrupt by the offending phase. Paul's comments are also correct with one subtle nit. The alternator = whine is due to diode switching as Paul describes. Paul is correct that = the diodes have a turn on delay, however, the source of the noise is due = to turn off time of the diodes (reverse recovery time for the techies). = It takes time(micro seconds) for the output current in the diode to = clear and the diode appears shorted during this time. The resulting = reverse current spike creates the whine. The small capicator in the = alternator goes a long way to filtering out this high frequency spike. Sorry to bore most of the readers. Please send me an off list email if = you would like to pick bones or want more detail. Jack Morgan IV P serial 603 with 45 hours. On Oct 23, 2012, at 6:00 AM, Lancair Mailing List wrote: > From: Gary Casey > Subject: Re: TCM 60A alternator--diodes > Date: October 22, 2012 10:13:51 AM EDT > To: lml@lancaironline.net >=20 >=20 > Paul, > I was trying to say what Brent said, but he said it much more = elegantly. It really all depends on the methodology used in the rating = of the diodes. Is the diode package rated at the current of the = individual diodes in the package? Or is it rated at the current = produced by the alternator into which it is intended to be mounted? And = the the rating is also "derated" as a function of temperature, and at = the core each diode is rated depending on the junction temperature = within the diode. What I'm saying is that it gets complicated. I = assume a front-mounted alternator runs very cool, so that's a good = thing. As I recall, most alternators put out a frequency that is 7 = times their rpm (why I remember that I don't know :-) and the alternator = runs a a speed that I assume is at least twice crankshaft speed. But, = the electrical noise you hear is not the ripple from the windings, but = the voltage transient caused by the turn-on delay of the diodes. Diodes = don't turn on instantly, so the voltage in a winding will rise above = battery voltage and then suddenly drop as the diode turns on, and that = happens three times per cycle. That will give you the frequency of the = "whine" you sometimes hear in the headset, which using those numbers = comes out to at least 1,680 Hz at 2400 rpm. If there is a capacitor in = the alternator, that's what it's for, not to reduce the ripple. Look = carefully at the alternator voltage (also at battery voltage) and you'll = see small, very short spikes in the voltage caused by the diodes. A = "picket fence" look. > And I wasn't trying to be judgmental or anything - I was trying to = point out that our conservatism in any design is a function of intended = usage. If it is only for day VFR in sparsely-used airspace the whole = electrical system is a convenience and one might not be too concerned = about the reliability of the alternator. So the choices one makes are a = function of intended usage. > Gary >=20 > Gary: I have to disagree and I don't think this alternator works the = way you describe. This is multi-phase high frequency alternator that = has multiple poles and three windings. The result is that all the = circuits are producing AC high frequency 3-phase rectified output and = the output is indeed additive from the 6 diode outputs in the sinusoidal = sense. Plus, we're not running at 60Hz but a far higher frequency with = six pulses per pole, multiple poles (12 or 14) and somewhere around = 7000rpm at cruise. >=20 > The result in a very high effective DC output voltage with limited = ripple. That's why it gets dumped right into the aircraft buss with only = the small capacitor located inside the alternator All the diodes are = working all the time and they don't "take turns" in sequence. Each = diode is rated at 50A continuous and the three phases are actually = additive, overlapping sine waves. Take away a phase by removing a = winding and you will generate less amperage. The output from this = configuration is about 1.73x the peak phase voltage and I extrapolate = that to be directly proportional for current also. See the diagram = attached for reference. Therefore, I calculate each phase only needs = 19.5A (100A over 3 phases divided by 1.71) to produce 100A output. = That 19.5A maximum per phase is far less than the 50A rating we're = discussing and suggests a large margin exists. Realistically, 100A units = are run somewhere less than 100% max continuous load calculations which = suggests even more margin is built in. >=20 > Having said that, the unit described is typically use for high output = amperage truck alternators in the 90A range and you can do your own = research to verify the applications. You might also want to check the = diode sizing in the TCM 60A alternator to see how that compares with the = 50s I replaced those with. >=20 > Your last comment seemed to suggest some form of risky behaviour was = being contemplated by this thread. Some nonsense about night IFR or = whatever. I think it was unnecessary and a pointless addition to what = was intended to be an informative discussion. >=20 > Paul > Legacy, Calgary --Apple-Mail=_DEEDD330-7E3E-46CB-90E1-C59499C5A60A Content-Transfer-Encoding: quoted-printable Content-Type: text/html; charset=iso-8859-1 Hi = all,

I was with Motorola during the original design = cycle for the first alternator diodes for Chrysler. Shouldn't admit to = being that old but I can't resist putting in two cents here. I = subsequently spent more years than I will admit (with many) bringing = electronics into the automobile from the first alternator through the = electronics evolution resulting in more than 30 micro controllers in the = basic car. It was a fun ride... obviously retired = now.

Brent's description is correct... as the = three windings produce their sinusoidal outputs (each 120 degrees out of = phase with one another) only one winding at a time is producing the = voltage required to cause conduction through it's respective diode pair. = I will leave the wye versus delta differences to those who care. Brent's = description is easiest but you engineering types may want to draw three = sine waves on a piece of paper..... the tops of the sine waves show the = ripple that is present on an open circuit alternator. The alternator is = actually a current generator so there is actually negligible noise = created by this voltage ripple since it is clamped by the battery and = the phases smoothly take turns supplying the output current as = controlled by the field current. Thus no need for a large filter = capacitor. The reason that a failed diode created radio noise is that = the smooth transfer of current is disrupt by the offending = phase.

Paul's comments are also correct with = one subtle nit. The alternator whine is due to diode switching as Paul = describes. Paul is correct that the diodes have a turn on delay, = however, the source of the noise is due to turn off time of the diodes = (reverse recovery time for the techies). It takes time(micro seconds) = for the output current in the diode to clear and the diode appears = shorted during this time. The resulting reverse current spike creates = the whine. The small capicator in the alternator goes a long way to = filtering out this high frequency spike.

Sorry = to bore most of the readers. Please send me an off list email if you = would like to pick bones or want more = detail.

Jack Morgan
IV P serial 603 = with 45 hours.


On Oct 23, 2012, = at 6:00 AM, Lancair Mailing List wrote:

To: I was trying to say = what Brent said, but he said it much more elegantly.  It really all = depends on the methodology used in the rating of the diodes.  Is = the diode package rated at the current of the individual diodes in the = package?  Or is it rated at the current produced by the alternator = into which it is intended to be mounted?  And the the rating is = also "derated" as a function of temperature, and at the core each diode = is rated depending on the junction temperature within the diode. =  What I'm saying is that it gets complicated.  I assume a = front-mounted alternator runs very cool, so that's a good thing. =  As I recall, most alternators put out a frequency that is 7 times = their rpm (why I remember that I don't know :-) and the alternator runs = a a speed that I assume is at least twice crankshaft speed.  But, = the electrical noise you hear is not the ripple from the windings, but = the voltage transient caused by the turn-on delay of the diodes. =  Diodes don't turn on instantly, so the voltage in a winding will = rise above battery voltage and then suddenly drop as the diode turns on, = and that happens three times per cycle.  That will = give you the frequency of the "whine" you sometimes hear in the headset, = which using those numbers comes out to at least 1,680 Hz at 2400 rpm. =  If there is a capacitor in the alternator, that's what it's for, = not to reduce the ripple.  Look carefully at the alternator voltage = (also at battery voltage) and you'll see small, very short spikes in the = voltage caused by the diodes.  A "picket fence" look.
Gary: I have to = disagree and I don't think this alternator works the way you describe. =  This is multi-phase high frequency alternator that has multiple = poles and three windings.  The result is that all the circuits are = producing AC high frequency 3-phase rectified output and the output is = indeed additive from the 6 diode outputs in the sinusoidal sense. Plus, = we're not running at 60Hz but a far higher frequency with six pulses per = pole, multiple poles (12 or 14) and somewhere around 7000rpm at = cruise.

The result in a very high effective DC output = voltage with limited ripple. That's why it gets dumped right into the = aircraft buss with only the small capacitor located inside the = alternator  All the diodes are working all the time and they don't = "take turns" in sequence.  Each diode is rated at 50A continuous = and the three phases are actually additive, overlapping sine waves. =   Take away a phase by removing a winding and you will generate = less amperage.  The output from this configuration is about 1.73x = the peak phase voltage and I extrapolate that to be directly = proportional for current also.  See the diagram attached for = reference. Therefore, I calculate each phase only needs 19.5A (100A over = 3 phases divided by 1.71) to produce 100A output.   That 19.5A = maximum per phase is far less than the 50A rating we're discussing and = suggests a large margin exists. Realistically, 100A units are run = somewhere less than 100%  max continuous load calculations which = suggests even more margin is built in.

Having said = that, the unit described is typically use for high output amperage truck = alternators in the 90A range and you can do your own research to verify = the applications. You might also want to check the diode sizing in the = TCM 60A alternator to see how that compares with the 50s I replaced = those with.

Your last comment seemed to suggest = some form of risky behaviour was being contemplated by this thread. Some = nonsense about night IFR or whatever.  I think it was unnecessary = and a pointless addition to what was intended to be an informative = discussion.

Paul
Legacy, = Calgary

= --Apple-Mail=_DEEDD330-7E3E-46CB-90E1-C59499C5A60A--