When I worked for a supplier of electronics to Lucas I noted that an alternator control chip's test machine was connected to an alternator and ran to ensure the output voltage was as close to 14.35 volts as necessary. This was how the factory production testing was done. But I also noted that the voltage continued to creep up after the chip was deemed to have passed the test. Is this creeping normal? What causes it? Is there an industry wide specification for how much time must be spent proving the stability of the output voltage on such tests?

 

Is this creeping normal? What causes it?

Without access to the datasheet of the chip it's hard to say what's 'normal'. Probably caused by the chip temperature rising.

 

+1 on temperature. Most alternator regulators have a thermistor in them to adjust for temperaturebut generally when the alternator is cold, it will charge at a higher rate and then lower as it warms up. The only other variable may be the source they are using. The alternator has the ability to sense battery voltage and may change slightly as it charges. It is hard to say without all the information available such as unit #, power source etc.

 

Hi guys. Is there an industry wide standard for how much the alternator voltage can be allowed to creep up before it is deemed to be a failed unit? For what length of time must the test be applied to ensure that the test's voltage creep is acceptable in the car alternator department?
Alec-t I don't only ask what causes this creep but how much creep is allowable in the car industry and creep over what period of time.
bwilliams60 the alternator field control circuits that I came into contact with didn't have a thermistor. They had a zener voltage reference diode. My question isn't about what drift one can expect but what drift is acceptable in the industry. If the chip passes the test at one point in time and is put in the "pass" bin, it isn't much good if the alternator output voltage slowly drifts up way beyond 14.35V when in use in a car. Where did you get the idea that a thermistor could somehow control the alternator output? How is it supposed to do so.

 

When I worked for a supplier of electronics to Lucas I noted that an alternator control chip's test machine was connected to an alternator and ran to ensure the output voltage was as close to 14.35 volts as necessary. This was how the factory production testing was done. But I also noted that the voltage continued to creep up after the chip was deemed to have passed the test. Is this creeping normal? What causes it? Is there an industry wide specification for how much time must be spent proving the stability of the output voltage on such tests?

AFAICR: those are temperature compensated - the alternator output voltage isn't supposed to be completely fixed, it has a temperature chart on the datasheet.

 

When I worked for a supplier of electronics to Lucas

Doesn't that say it all?

Modern alternator regulators went PWM, after the simple electronic duplication of the old mechanical variety caused radio noise/whine.
My car goes through -40°c and +40°c throughout the four seasons and the digital display I have, rarely varies off of 14.5v reading.
Max.

 

MaxHeadRoo, glad your alternator gives a constant voltage. However, Ian Field doesn't seem to think that is what alternators should do. The alternator control circuits I'm describing were also PWM. Their o/p controlled the field voltage. But your car's alternator is only one sample. I was involved with making control circuits in their tens of thousands and many of them gave drifting o/p voltages of the test alternator. I never could find out how much drift was allowable in the industry. The drift I refer to is drift from any cause, not only temperature necessarily.

 

Where did you get the idea that a thermistor could somehow control the alternator output? How is it supposed to do so.

Well, lets see. I used to own a business, not anywhere nearly as big as Lucas, but probably rebuilt a few thousand alternators in my lifetime. I will refer you to this article which is just one of a few on the subject but you will note at the end where it says a thermistor is used to help regulate voltage with temperature. When a battery is cold, a greater charging current is necessary to push electrons into the battery. Opposite when warm. A zener diode does not do that. It creates a set point and as far as the limit, it is set by the manufacturer but it would usually be set at around 14.4-14.6, lower if GEL cell or AGM batteries are used.
https://alternatorparts.com/10si-15si-type-116-136-repair-manual-page-2.html
Max is correct. They are all PWM now and the field is controlled on a duty cycle. They will cycle upwards of 600 times per second or more, depending on manufacturer and system demands.
14.35 volts is not a concern and as far as how much they will vary, there is not a publicly known variable, but from experience I would say 0.1 - 0.4 VDC would be about it.

 

The consumer will be the ultimate decider of what is acceptable, and suppliers will react to their desires.
There is no national or international standards organization standard for the parameter you ask about.
When a company develops a reputation for selling alternators that cause damage to batteries which create noticeably higher/earlier failure rates in them, they either adopt better standards or join the ranks of failed companies.
The customer is king in the retail world, and with Internet communication word spreads really really fast these days.

 

Doesn't that say it all?

Modern alternator regulators went PWM, after the simple electronic duplication of the old mechanical variety caused radio noise/whine.
My car goes through -40°c and +40°c throughout the four seasons and the digital display I have, rarely varies off of 14.5v reading.
Max.

The terminal voltage of a lead-acid battery has a temperature coefficient.

Ignore that at your peril.................

 

Well, lets see. I used to own a business, not anywhere nearly as big as Lucas, but probably rebuilt a few thousand alternators in my lifetime. I will refer you to this article which is just one of a few on the subject but you will note at the end where it says a thermistor is used to help regulate voltage with temperature. When a battery is cold, a greater charging current is necessary to push electrons into the battery. Opposite when warm. A zener diode does not do that. It creates a set point and as far as the limit, it is set by the manufacturer but it would usually be set at around 14.4-14.6, lower if GEL cell or AGM batteries are used.
https://alternatorparts.com/10si-15si-type-116-136-repair-manual-page-2.html
Max is correct. They are all PWM now and the field is controlled on a duty cycle. They will cycle upwards of 600 times per second or more, depending on manufacturer and system demands.
14.35 volts is not a concern and as far as how much they will vary, there is not a publicly known variable, but from experience I would say 0.1 - 0.4 VDC would be about it.

A regular on Usenet electronics groups; Jim Thomson, has all the relevant info on his website for car battery tempco and probably snippets on alternator design.

http://www.analog-innovations.com/ - you'll have to do your own digging around if you want specific info.

 

The terminal voltage of a lead-acid battery has a temperature coefficient.

Ignore that at your peril.................

So what do you suggest?
Max.

 

So what do you suggest?
Max.

Look on the website I referenced earlier...........

 

bwilliams60, you say in relation to temp coefficient of the voltage output that a zener diode does not do that. You assume that the temp coeff of reference voltage from the zener is zero but it is only nominally zero for a 6.8V zener and there is a tolerance on that. Some samples will creep up with temperature, some go lower. What the manufacturer does during production is to set the alternator output voltage to 14.35 (or 14.5 as some say) and makes this process happen as quickly as he can so as to keep the day's output high. He rarely checks the creep of alternator output voltage as the zener warns up. I failed to find reference to the thermistor in the article you mention, by the way. Where is it?
Kermit2 I hear what you say and it seems to me to be true. But what a state of affairs! There really should be an industry-wide standard for this.

 

So how far are you considering voltage creep to be an issue at? Whats your idea of an ideal set point VS a creeped one?

I consider any charging system that puts out between 14 - 15 volts at any normal load point to be working just fine. Even 15 - 16 in the winter until thing warm up would not bother me any.

 

I assume nothing but sometimes I do talk in general terms as I do not know the entire audience. The article refers to the thermistor in the last line beside Figure 4.
As for what Lucas does in their test room, I have no idea what the conditions are but I am sure they will have an upper and lower limit to their voltage values and as long as they are in there somewhere, no alarm bells will go off. I am sure most electronic circuits "creep" up or down due to temperature fluctuations and that is probably very carefully engineered into most good products. Lucas has a very good name in the rebuilding industry so I am sure they are doing it right.
Tcmtech, I almost agree with your statement. Up to around 14.8, I am comfortable but after that, batteries start to get hot and things start to happen. More true if AGM or GEL as I said earlier.

 

bwilliams60 yes I can see the thernistor now but it's inside the regulator not next to the battery. So if the battery is at a different temperature to the alternator the latter will produce the wrong output voltage. Another fault with this design is that the reliability of thermistors is not very good. MIL Hdbk 217C shows this. A far more reliable and more accurate component for temperature control is an ordinary diode. The forward (conducting) voltage changes at -2.5mV per deg C and the conduction voltage graph is very stable and exactly the same for all samples of the same type of diode. But none of that is what we were discussing originally. My main point is that leaving the end user of the alternator to be the final inspector is not good practice. The industry needs a standard to which it can work. If you put yourself in the position of a designer of regulators for use in alternators, what temperature coefficient is he supposed to design for? And even worse than that, for what period of time is the output voltage to be allowed to change? You can say that the factory testing regime is up to the manufacturer but without a thorough specification to cover it anything is possible.

 

I guess I am having a hard time understanding where you are going with this. Although the battery and alternator may be at different temperatures, there is still a sense wire which tells the alternator what battery voltage is, and the alternator uses that as well to determine its duty cycle. These alternators and setups have been used for years without problems and from what I get, you want to fix "something" but if something works, don't muck with it. As for our initial conversation, you asked why there might be creep. I simply gave you temperature as one variable. I'm sure there may be more but that is the first one that pops out at me. I don't know the lab conditions.

 

OK bwilliams60. I hear you. Thanks for your input. So what does the thermistor do? What you call "lab conditions" is the lack of a national or even an international standard regarding the output voltage of an alternator, apart from the vague fact that ot must be somewhere between 13V and 15V at all times and at all temperatures. What about the result of other conditions that affect it? Are they not to be specified? And what the lab does is accepted by the factory and then passed on to the end user without further question. Without a national spec how can the lab possibly get it right? OK it has worked for a while but the setup is wrong as it stands.