Cooked Balmar AT-200

[Jd1]

Originally Posted by transmitterdan

Let's look at some numbers.

A typical 6 diode rectifier producing 200A DC puts 67A through each of 6 diodes. Peak currents can be higher but we'll ignore that for now. So we have about 1.2-1.5 V and 67A on each diode which is somewhere between 80-100W per diode. Typical junction to case temp is 1.5C/W so it means the silicon is 90-150C hotter than the case.

If you want those diodes to last you need to limit case temp a lot lower than 104C or run less current.

External mounting helps a lot. Forced air (or better yet liquid cooled) heat sink can maintain a much lower heat sink temp.

The AT-200 uses 12 (!!) diodes and I wouldn't be surprised if your voltage drop figure is substantially too high - these are not regular silicon diodes. Nevertheless, I agree with your sentiment.
Add to all this the fact that the AT-200 temperature probe location is on the alternator frame and not the diode pack, plus there is a substantial delay/obstruction to get any heat from the diode pack to the frame, and you have a BIG problem.

When I get my repaired alternator back I will do some test runs with sensors on both the frame and the diode pack to look at the difference.

[transmitterdan]

The case of the alternator is a pretty good thermal conductor (but poor radiator of heat). So the frame and the place where the diodes are mounted are not much different temperature wise. Maybe 10C. The big delta in temp (50-100C) is from the junction to the case. Even if they use hot carrier diodes there is still a significant power loss in the diode bank. Using 12 diodes just divides almost the same power among more diodes. It doesn't reduce the total power loss a lot. Maybe 10%.

I think the lesson here is that the 200A rating means 200A for a while. And it does not mean you can run the case at 104C for a long time because the diodes are going to be about double that temp. Running diodes at junction temp of 150C will result in lifetimes measured in tens of hours not thousands. If you want to run at 200A continuously you have to keep the case at about 45C. It takes a lot of cooling air to do that. Maybe even impossible to do.

The solution is to use external diodes and consider cooling them with water. Replace the heatsink every couple of years because aluminum and salt water hate each other. Then you can have a 200A system that should be very reliable.

http://www.wakefield-vette.com/compa...heat-sink.aspx

[hellosailor]

jd-
There are all sorts of issues and solutions to diode cooling that only an engineer in the business or a hard enthusiast would be aware of. For instance, forced air cooling. If there is a heat sink on the diode frame, it is intended to have forced air cooling ALL THE TIME and your test without that, should have thrown a red flag. Typical small frame alternators have one "pusher" fan in the front of the alternator, blowing air into it. The heavy duty alternators often have two fans, one pushing, and a second on the rear "pulling" as well. That's done by necessity.
Then there is diode rating. Diodes rated more conservatively will be larger and more expensive, but they help ensure proper operation. And like all electronic components, a diode's life is cut in half with every 10C rise in temperature. Start with 100 years to 20C and raise the temperature six or eight steps...see how that gets chopped.
AC Delco (I'm pretty sure) at one point marketed a line of alternators with internal spike protection, since normal startup and shutdown throws spikes that can kill the electronics in a regulator and harm the diodes. They found out that the spike protection components failed more often than the rest of the alternator and killed it even sooner. ooops.
And then there are the made Germans over at VW-Porsche-Audi. They've been using water-cooled alternators (shades of video gamers!) in their top-end cars for some years now. Apparently cools great, although the US companies say that's got to create leak problems and shouldn't be necessary.
Remote diodes? More connections, more voltage drop, more complications. Can make cooling easier, certainly makes it easier to use robust diodes in a big heat sinked frame.

There's all kinds of things you can do, probably all have been done before and they are generally not done because of the possible complications. Forced air? With an air source outside the engine bay? That's a fairly easy one, compared to everything else.

[deluxe68]

Originally Posted by transmitterdan

The case of the alternator is a pretty good thermal conductor (but poor radiator of heat). So the frame and the place where the diodes are mounted are not much different temperature wise. Maybe 10C. The big delta in temp (50-100C) is from the junction to the case. Even if they use hot carrier diodes there is still a significant power loss in the diode bank. Using 12 diodes just divides almost the same power among more diodes. It doesn't reduce the total power loss a lot. Maybe 10%.

I think the lesson here is that the 200A rating means 200A for a while. And it does not mean you can run the case at 104C for a long time because the diodes are going to be about double that temp. Running diodes at junction temp of 150C will result in lifetimes measured in tens of hours not thousands. If you want to run at 200A continuously you have to keep the case at about 45C. It takes a lot of cooling air to do that. Maybe even impossible to do.

The solution is to use external diodes and consider cooling them with water. Replace the heatsink every couple of years because aluminum and salt water hate each other. Then you can have a 200A system that should be very reliable.

Build Your Own Liquid Cooled Heat Sink

But in my case (165 Balmar software derated by 28% and 614 regulator) why does my temp get up to 100c so quickly when the bulk charge is 130 Amps? Should I derate it more than 28%? I knew these things were not rated for continuous running at max output but I did not think they would overheat at 80% output.

[Jd1]

Originally Posted by deluxe68

But in my case (165 Balmar software derated by 28% and 614 regulator) why does my temp get up to 100c so quickly when the bulk charge is 130 Amps? Should I derate it more than 28%? I knew these things were not rated for continuous running at max output but I did not think they would overheat at 80% output.

I estimate that I will get around 110 to 120 amps continuously out of a 200A alternator if I do not depend on external cooling or the temperature probe. That is how I will set the derating and after that is set I will run external cooling to improve life.
I would think about 90 amps sounds about right for the 165 but of course that depends on a lot of little details of your install.

[hellosailor]

"why does my temp get up to 100c so quickly "
You mentioned running with zero airflow. Is that still the case?
Did you read the engineering notes for that alternator, including the recommended airflow, and are you meeting or exceeding the recommended airflow in all ways?
Yes, it can be that simple. That's what I would expect.

[deluxe68]

Originally Posted by hellosailor

"why does my temp get up to 100c so quickly "
You mentioned running with zero airflow. Is that still the case?
Did you read the engineering notes for that alternator, including the recommended airflow, and are you meeting or exceeding the recommended airflow in all ways?
Yes, it can be that simple. That's what I would expect.

The original alternator was OEM for the Yanmar and was rated at 80 amps. At the same rpms as the New Balmar is overheating the old alternator put out 50-60 amps. It was operating in the same conditions. I will check the Balmar documents. Where did I say there was zero airflow? Please enlighten with your expertise.

[transmitterdan]

A 28% reduction in current reduces the diode temperature rise (all other things being equal) by 28%. It reduces the winding rise about 50%. So the 150-200C diodes drop about 20-30C. Not enough to solve your problem I think. To make a significant dent in overheating you need 10X more airflow or a 50% reduction in current.

[DeepFrz]

I was always under the impression that an alternator drew in air from the rear and it exited at the front. Is this correct or am I mistaken?

[deluxe68]

Originally Posted by DeepFrz

I was always under the impression that an alternator drew in air from the rear and it exited at the front. Is this correct or am I mistaken?

I thought it travelled from front to back. I will verify, because I think I am going to try forcing air into the alternator. The engine compartment already gets ventilated while running, I just need to add another fan dedicated to the crappy designed Balmar.

[hellosailor]

Message #1 said "The tests I was running were with no forced air cooling" so the question arises, were those tests "in place" same as operating conditions, or was there forced air cooling during normal operations? First rule of troubleshooting, assume NOTHING. Except, perhaps, that there may be earth-normal gravity and atmospheric conditions. Perhaps.(G)

" I think I am going to try forcing air into the alternator." You don't say "more air" you make it sound like there has been no forced air flow, no fan.

Airflow on alternators is normally from front to rear, because "normally" the alternator is somewhat in front of the engine block in a car/truck, so that fresh cool air coming in through the radiator hits the alternator before it has been heated by the engine block or bounced off the hot firewall. That's not to say all designs will be that way, just that it is "normal". And those cooling fans probably can be installed "backwards" on the shaft by clever accident, worth checking.

Note that the engine compartment on a boat normally does not have the same massive air flow coming in the front grill [sic] so an alternator on a boat is already being over-worked due to lack of cooling air. This has been mentioned in older threads, where some folks have either rigged a simple duct to a set of air scoops in the side of a larger power boat, or to the main salon in a sailboat, in order to ensure a fresher air supply. Or, added an in-line fan in that duct as well.

Combined with a bilge blower and left running after shut-down, this also cools an engine down faster, which will minimize coking of the oil and help the whole boat cool down in hot weather, so it isn't really such a bad idea.

I know Balmar sometimes goes entirely AWOL (shuts down while there's a boat show to attend) but have you tried contacting them? I could see the fan being a "necessary but optional" component that might not ship as part of the "base" alternator, even if it was necessary. They've got the experience to know about all sorts of quirky oddball issues with their products.

Or perhaps, you got one that was made for use in the southern hemisphere. You know, everything has to turn backwards to be used down there.(VBG)

[Jd1]

Just received my repaired alternator back and had a look-see. Two diodes toasted but the interesting thing is that there is zero/none/nada/no thermal path between the alternator diode pack and the alternator frame !!!!

So moral of the story, the temperature sensor on the Balmar AT series alternators only protects the frame and windings, it does absolutely nothing to prevent diode pack meltdown.
It will be interesting to run temperature comparisons between the frame and the diode pack!

[wsmurdoch]

Originally Posted by hellosailor

Airflow on alternators is normally from front to rear, because "normally" the alternator is somewhat in front of the engine block in a car/truck, so that fresh cool air coming in through the radiator hits the alternator before it has been heated by the engine block or bounced off the hot firewall. That's not to say all designs will be that way, just that it is "normal". And those cooling fans probably can be installed "backwards" on the shaft by clever accident, worth checking.
(VBG)

All the alternators I am familiar with have either a single fan on the front of the alternator (like a 10SI alternator https://alternatorparts.com/x10si.html), or a fan on the front and an internal fan (like a CS130 alternator https://alternatorparts.com/xcs130.html), or two internal fans (like a CS130-D alternator https://alternatorparts.com/xcs130d.html). All suck in from the back over the diode pack and blow out at the front or side. The fans are centrifugal. They suck at their center and blow out at their periphery. The angle of their blades makes them a little more efficient when rotated in one direction than in the other, but the difference is not really huge.

[George DuBose]

Returning from a six week voyage that entailed quite a bit of motoring, I saw that my batteries were being charged at 18v!

Removing the Balmar Series 6 alternator and the ARS-5 regulator, a Dutch tech told me that the Balmar was burned out beyond any repair possibility, as was the ARS-5.

Reading through this forum, apparently additional cooling is needed for the alternator. I will install box fans in a duct and get more cool air into the engine compartment.

When a tech at Balmar wrote me about my alternator failure, well here is what he wrote:
Here is what the Balmar tech told me about "planned obsolescence".

George,
It is not planned obsolescence, it is an end-of-life failure. There are
thousands of metal solder joints on each circuit board. Every time the unit is
turned on the solder joints heat up and expand. When the unit is turned off the joints cool down and contract. This continual flexing eventually cause the
joint to fail. We have designed it as robust as we can (and keep the price
reasonable) but eventually it is going to fail. Any piece of electronics is
going to fail at some point. Most times the unit fails in an "open" condition
and just stops working with no output. On occasion they fail shorted and allow current to pass through them. Due to manufacturing variances and tolerances we can not accurately predict what and when something will fail. There is a 10 amp fuse in the power in wire and this should blow before any damage occurs to the alternator.

So I guess I need to cool the regulator as well?

Electrics are not my strong point, but for what these Balmar systems cost, I would expect better dependability. The Balmar system wasn't even 8 years old.

[NahanniV]

Originally Posted by George DuBose

...
George,
It is not planned obsolescence, it is an end-of-life failure. There are
thousands of metal solder joints on each circuit board. Every time the unit is
turned on the solder joints heat up and expand. When the unit is turned off the joints cool down and contract. This continual flexing eventually cause the
joint to fail. We have designed it as robust as we can (and keep the price
reasonable) but eventually it is going to fail. Any piece of electronics is
going to fail at some point. Most times the unit fails in an "open" condition
and just stops working with no output. On occasion they fail shorted and allow current to pass through them. Due to manufacturing variances and tolerances we can not accurately predict what and when something will fail. There is a 10 amp fuse in the power in wire and this should blow before any damage occurs to the alternator.

So I guess I need to cool the regulator as well?
...

Or, look at something other than Balmar !!!