I'm attempting to convert 32 VAC to 24 VDC, sourced from a steam driven turbo generator.

It's on a steam locomotive with AC lighting, but needing 24v DC for radio and speedometer. 24 VDC is currently sourced from 2 x 12v deep cycle lead acid batteries wired in series; charging them up after each day's travels. The aim is to keep the batteries charged, or at least slow the depletion rate, whenever the locomotive is 'live'.

I've made a partially succesful prototype using a bridge rectifier (BR354), filtering the output using a 100v 10000uF electrolytic capacitor, then into a 150W DC-DC converter. I've set the output at 27.2V (battery float voltage).

It seems to work, however there are some problems and concerns.

1/ Current spiking due to the filtering capacitor. This concern arises from circuit simulation showing approx. 90amp peak (albeit abrubt) each charging cycle of the capacitor. The turbogenerator is rated at 16A (500W @ 32VAC). Is that magnitude of current spiking correct? Could current spiking cause excessive heating/damage to the loco or generator wiring ahead of the rectifier?

2/ The DC-DC converter drops offline about half the time. I believe this is because as the turbo RPM drops depending on electrical load (and varying steam pressure), rectified voltage drops below the DC-DC converter minimum input of 30V to function. I also think 150W might be too much to bleed off. Does 75W seem like a sensible 2nd attempt? Is a wider range DC-DC converters eg 18-60V worth a shot?

3/ Longevity of the capacitor - I'm not sure it's going to last with the ripple it's getting. The current cap was really only meant to be interim to prove the concept. What capacitor choice would be appropriate?

4/ Is the significant voltage/current ripple output from the rectifier likely to damage the DC-DC converter? Is there a better way to filter?

5/ Is 27.2 VDC output to the batteries the correct choice?

I've looked at PWM / switching AC-DC converters, but they don't seem to handle significantly varying AC frequency, which I assume to be the case, or AC voltage's much below about 90 VAC so discounted that solution. Are there converters that can handle varying frequency? Searches so far have been fruitless.

I would greatly appreciate any assistance with this project. Any advice welcome.

 

since ac to dc switching power supplies ddont really have anything to do with the ac frequency coming in, tghey should work at varying frrequencies. the only problem is when the input frequency drops too low, it might introduce a lot of ripple into the internal filter cap. ac pwm supplies are very forgiving of input voltage, even working on dc input. how about using a transformer to get tieh 32 vac up to 125 for the supply?

 

This may be hard to answer without some more info. But.
The peak current for charging is probably lower due to the internal resistance in the generator. But it will cause heating there.
It sounds like a single phase alternator/generator not 3 phase like a car alternator?
If it is indeed single phase we need to know the minimum frequency and voltage at idle to size the cap for the load. If you don't have a way to measure the frequency you might be able to get close knowing the engine rpm and the pulley ratio. You could get an idea of the voltage by just measuring across the cap with a small load - maybe one of the small ac lights.

 

Why not look into controlling the turbo generator output for 29vdc for constant charge, I imagine the lighting will still be OK at that level?
It would just be a matter of adjusting the field control and placing a 3phase rectifier on the output, same principle as done in a automotive type, an automotive type could also be simple to modify.
That would be the simplest route I would be inclined to look at
Max.

 

If the AC is three-phase then you should use a 3-phase rectifier to minimize peak currents and ripple voltage.

 

Your batteries are a much better "capacitor" than the filter cap, which can be eliminated.

What size are your batteries? It's not obvious to me that you need anything other than the rectifier and a charge controller, if the batteries can absorb the max current available from the generator. If they cannot, you need a current limiter.

 

since ac to dc switching power supplies ddont really have anything to do with the ac frequency coming in, tghey should work at varying frrequencies. the only problem is when the input frequency drops too low, it might introduce a lot of ripple into the internal filter cap. ac pwm supplies are very forgiving of input voltage, even working on dc input. how about using a transformer to get tieh 32 vac up to 125 for the supply?

That's a good thought, I didn't think of a transformer to up the voltage first. I might experiment with that.

 

This may be hard to answer without some more info. But.
The peak current for charging is probably lower due to the internal resistance in the generator. But it will cause heating there.
It sounds like a single phase alternator/generator not 3 phase like a car alternator?
If it is indeed single phase we need to know the minimum frequency and voltage at idle to size the cap for the load. If you don't have a way to measure the frequency you might be able to get close knowing the engine rpm and the pulley ratio. You could get an idea of the voltage by just measuring across the cap with a small load - maybe one of the small ac lights.

It is a single phase unit (Stone's L.B.B turbo-generator, circa 1948) and has permanent magnets to create the magnetic filed, ie there's no field control of any sort. It's mechanically governed using a steam valve. I don't know rpm, but they do drop quite a lot under load. At a guess it's probably roughly 3000-3550 rpm no load. I don't know the frequency, but I'm thinking of upgrading my multimeter so I can try to measure the frequency. I assume frequency varies considerably with rpm but not sure on this.

Being singe phase, I assume the rectifier output ripple would be higher than that of a regular 3 phase alternator .

It's directly driven from the steam turbine, the shaft runs from the turbine though the the stator with a rotor in the middle and into the governor at the other end. The nominal AC voltage is 32V, and it measures close to this no load, but drops to approx. 24v or lower under full load.

Wiring heating in the generator is one of my concerns, it's a pretty tough unit, mounted in a very hostile environment om top of a hot boiler, but I don't want to damage the wiring.

Why not look into controlling the turbo generator output for 29vdc for constant charge, I imagine the lighting will still be OK at that level?
It would just be a matter of adjusting the field control and placing a 3phase rectifier on the output, same principle as done in a automotive type, an automotive type could also be simple to modify.
That would be the simplest route I would be inclined to look at
Max.

Unfortunately there's no way to control the unit, as it has permamanet magnets vice field control, and internally governed.

Your batteries are a much better "capacitor" than the filter cap, which can be eliminated.

What size are your batteries? It's not obvious to me that you need anything other than the rectifier and a charge controller, if the batteries can absorb the max current available from the generator. If they cannot, you need a current limiter.

I was wondering about the ability of the batteries to act as a capacitor, and indeed to cope with ripple. They are 12V deep cycle lead acid 105Ah each. I'm using the cap ahead of the dc-dc converter, maybe it's not needed? The dc-dc converter is current limited (150W).

Appeciate the input folks!

 

What is the AC frequency range, and what DC/DC are you using now? There a bunch of them to choose from, some designed specifically as battery charger devices. Capacitor ripple current can be calculated from the equivalent series resistance (ESR) on the cap datasheet. Generally speaking, larger caps handle more ripple. For example, the old fat "computer grade" electrolytics are much better for this than little pcb types.

ak

 

The turbogenerator is rated at 16A (500W @ 32VAC).

They [the batteries] are 12V deep cycle lead acid 105Ah each.

 

If the batteries were discharged and the generator running at full, it might drive 15-20A into those batteries if there is nothing but a rectifier in between. That current level is too much for a battery of that size. So you do indeed need something more in between.

 

A good device would be a constant current (or current limiting) buck boost converter. Unfortunately I haven't been able to find one that will take 50 volts or so on the input. Then you wouldn't need to worry so much about it dropping out at low RPMs and you could use a smaller cap on it's input. I'll look some more.

 

I can't find one for under a few hundred dollars. Seems they don't like the higher input voltage.
Are you handy? We could build a simple battery charger type that we could limit the current to lets say 3 amps and set the voltage to the battery float voltage.