Time for a new alternator?

[theAnonymous1]

I was testing a SMPS in my car today and was a bit shocked at the results.

No load - in 14.4v / out 98v
10R load - in 11.3v / out 64v

64v^ / 10 = 409.6

So, with a ~410W load my charging system dropped to 11.3v. Does this sound normal for a stock ~70A alternator at idle?

I'm leaning toward the alt is OK, just underpowered. Either way I'm going to buy a new alt; I'm just wondering if it's worth breaking the bank for a 160A unit.

 

[richie00boy]

At what engine RPM? You may find at idle the output collapses a lot more easily than say at 2000 RPM and above.

 

[theAnonymous1]

~700RPM idle. I'm going back out to the car to see what the voltage drops to with the engine held at 2K.

I think generally an alternator will put out 40-50% of its rated current at idle.

EDIT: Done. The results were pretty much the same, even at 3K. Also, two 10W 20R resistors in parallel cannot survive for more than 4 seconds at 64V. Time to find a new load.

 

[luka]

Hi

That is bad...but is your alternator/batter what is wrong, or wires going to smps?

 

[theAnonymous1]

The battery is only a few months old, but the alt looks original and possibly on it's last leg. The wiring is 10ft of 8 gauge.

 

[luka]

Hi

Where did you measure that voltage? On BAT? Is any bad grounding possible?

 

[theAnonymous1]

I was checking the voltage at the supply input. I'm going to check the voltage at the battery as soon as I find something to wrap this 10ft if nichrome wire around.

 

[luka]

Hi

Even at supply input there shouldn't be that low voltage

 

[theAnonymous1]

Ahha, definitely a wire problem. The voltage at the battery is 12.6v while the voltage at the supply is 11.15v. This is with the new load I made that is just under 10R.

I would like the try the supply connected directly at the battery, but there is a bit of a snow storm out right now.

Does 12.6v sound more reasonable for a 410w load?

 

[luka]

Hi

Lool snow storm, I have luck, nothing over here
Engine on @ 12.6v on battery = not good. Even my small clio can do better and I am sure you have bigger car...

You can check wire voltage drop, to see if you can get only by using bigger wire more out to smps.

 

[theAnonymous1]

The voltage drop would be 12.6v - 11.15v = 1.45v wouldn't it? The 12.6v was with the supply powering the load, it is closer to 14.6v no load at the battery with the car on.

The car is a Toyota Celica with a 2.2L engine. I think I'm just going to spring for the 160A alternator.

 

[luka]

Hi

Well if voltage is that low, with load, even if at higher RPM, regulator of alternator could be dead or working wrong...Probably can't do 70A anymore, if it would then voltage would stay high + 160A alternator is always good to have

 

[theAnonymous1]

Another question Luka. The output of the supply has 200v 16A TO-220 rectifiers (1 common anode and 1 common cathode per rail); is the current limit of each rail theoretically 32A or 16A?

 

[luka]

Hi

Can you read which one is it? There are probably two 8A diodes in ||, so thats why 16A, but if each diode is 16A then it is 32A, but current is probably given for whole device, not single diode

 

[theAnonymous1]

They are FEP16DT and FEP16DTA. I just replaced the the cheap FMG22x 200v 10A rectifiers with these.

I was thinking about making an external bridge for each rail using RURG8060 80A 600V UltraFast diodes.

The IRFZ44N are soon to be replaced with HUF75337P3 75A 55v UltraFETs.

 

[djQUAN]

same thing happens to mine. car is toyota rav4.

although, when the alternator is cold, at 750rpm, voltage is 14.3V. when alt heats up, drops down to 13.8V. when loaded with about 1KW or so, voltage drops to 12.2-12.4V (I have a truck battery under the hood) athough alternator output does improve when we get to 1500rpm and up.

you will probably not have problems when the amp is driving a speaker load and music signals, your ears wouldn't like the torture anyway. as I understand, in that test, your SMPS is directly connected to the load...

 

[theAnonymous1]

Yeah the 10R load was connected directly to both rails.

Below is a pic of the supply. TL494, 16x IRFZ44N (soon to be HUF75337P3), 8x 1mm primary, 3 x 1mm secondary, 10,000uf 71v per rail.

This will powering a 2R BTL via Tripath TDA2500 with 80A fets. There is about a 90% chance I blow something up.

[IMGDEAD]http://i32.tinypic.com/2ik2i5w.jpg[/IMGDEAD]

 

[luka]

Hi

You could make external bridge, if with very short wires, but I don't think you need to. Jup 16A diodes, but that is a lot if you will have 45V or more on rail, and if possible go for higher, say 4 or even 8R load, since in bridge you won't have problems, only if your supply works at +/-45V

 

[theAnonymous1]

The amp section that used to be attached had 10 pairs of A1694/C4467. It would run 1/2 stereo or 1 ohm BTL with no problem. The thing was an inefficient current hog though.

What do you think about taking the outputs of the transformers and making a +/-90v supply instead of +/-45v? Then run 4 ohm stereo?

 

[luka]

Hi

I think it would take more time, but all will run better, since output currents won't be that high, much lower than wih 2R, PCB will be able to handle currents better, output inductors, fets,...