Voltage Could this be correct?

[Indred]

Ok. I've got my DMM finally and I tested my PSU. But from my reading I think I would have cooked some thing by now but i'll post and let yall tell me if i'm doing it wrong.
I've tested off my 4 pin molex connector the red and yello wires (red and ground then yellow and ground) and came up with this.
Yellow = 16.94 DC V
Red = 7.16 DC V
I have a GreenLEE DM-20 brand new DMM Please tell me i'm doing some thing wrong LOL

 

[CrispyFried]

is the dmm on an ac setting? sounds about right for ac scale.

 

[Indred]

No it was on the DC scale I thought Power on that was DC is it AC. Am I stupid LOL

 

[Geniere]

Originally posted by: Indred
Ok. I've got my DMM finally and I tested my PSU. But from my reading I think I would have cooked some thing by now but i'll post and let yall tell me if i'm doing it wrong...

Well you want to measure the voltage referenced to common, not to ground. So put one probe on the yellow wire and the other probe on the red wire. You should measure 12vdc with a polarity depending on how the probes are connected, can't remember but I think the yellow wire is common.

The few times I've done this, the common and ground were electrically connected so it did not matter if a measurement was referenced to common or to ground (chassis, power cord ground pin, etc). In your case, I don't know.

Not hi-tech!

 

[jmcoreymv]

Here's the pinout diagram of the molex connector: http://www.helpwithpcs.com/courses/power-supply/atx-psu-pinouts.gif

You should be measuring between either red and black or yellow and black to get the voltages. If you measure between red and yellow you will end up with +/- 7V.

 

[Indred]

thanks i'll check that out. I was trying to read my 12v 5v (i think or may be its 6v) and 3v rails. But i'm leaving in the morning for a 5 day trip so I'll have to check it when i get back Thanks again.

 

[Bassyhead]

The black wires are common and ground! Red is +5VDC, yellow is +12VDC. If you measure higher than normal values, it may be because the power supply is not connected to a load.

 

[Matthias99]

Originally posted by: Bassyhead
The black wires are common and ground! Red is +5VDC, yellow is +12VDC. If you measure higher than normal values, it may be because the power supply is not connected to a load.

That was going to be my suggestion. You will probably *not* get accurate results with nothing hooked up to it (and you should not leave it running like that for extended periods of time). Try hooking it up to the motherboard and powering it up, then measure the voltages on an unused 4-pin 'molex' connector.

Well you want to measure the voltage referenced to common, not to ground. So put one probe on the yellow wire and the other probe on the red wire. You should measure 12vdc with a polarity depending on how the probes are connected, can't remember but I think the yellow wire is common.

Red is +5V, yellow (sometimes orange) is +12V, black is 'common' (0V). In all PC systems that I am aware of, the 'common' voltage is tied directly to the electrical ground (the case/PSU are both tied to the ground pin of the AC outlet, and then everything else gets its ground/power from the PSU).

 

[Navid]

Originally posted by: Bassyhead
If you measure higher than normal values, it may be because the power supply is not connected to a load.

Really? I thought most power supplies had voltage regulators in them, which would keep the voltage constant regardless of load!

 

[Matthias99]

Originally posted by: Navid

Originally posted by: Bassyhead
If you measure higher than normal values, it may be because the power supply is not connected to a load.

Really? I thought most power supplies had voltage regulators in them, which would keep the voltage constant regardless of load!

Yes, but voltage regulation (unless specifically built to work like this) generally requires a nontrivial amount of resistance in the circuit.

 

[Fin]

Need to measure the voltages under some kind of resistance load. The voltages you are reading are not under load.
-Fin

 

[Navid]

Originally posted by: Matthias99

Originally posted by: Navid

Originally posted by: Bassyhead
If you measure higher than normal values, it may be because the power supply is not connected to a load.

Really? I thought most power supplies had voltage regulators in them, which would keep the voltage constant regardless of load!

Yes, but voltage regulation (unless specifically built to work like this) generally requires a nontrivial amount of resistance in the circuit.

A voltage regulator is a simple negative feedback system. It compares the output voltage, or a portion (division) of it, to a reference voltage. If the output voltage is low, the regulator increases it. If the output voltage is high, the regulator decreases it.
This is done on a continuous basis.

So, the output voltage will always be constant to a reasonable degree (negative feedback gain has an impact on how much error may appear at the input of the comparator, but such error is negligible and is no where near 4V).

I am not familiar with all existing voltage regulators. So, I could be wrong.
Do you have a link to a datasheet or description of a regulator that falls apart with no load?

 

[Bassyhead]

Originally posted by: Navid

Originally posted by: Matthias99

Originally posted by: Navid

Originally posted by: Bassyhead
If you measure higher than normal values, it may be because the power supply is not connected to a load.

Really? I thought most power supplies had voltage regulators in them, which would keep the voltage constant regardless of load!

Yes, but voltage regulation (unless specifically built to work like this) generally requires a nontrivial amount of resistance in the circuit.

A voltage regulator is a simple negative feedback system. It compares the output voltage, or a portion (division) of it, to a reference voltage. If the output voltage is low, the regulator increases it. If the output voltage is high, the regulator decreases it.
This is done on a continuous basis.

So, the output voltage will always be constant to a reasonable degree (negative feedback gain has an impact on how much error may appear at the input of the comparator, but such error is negligible and is no where near 4V).

I am not familiar with all existing voltage regulators. So, I could be wrong.
Do you have a link to a datasheet or description of a regulator that falls apart with no load?

Most regulators (especially switching ones like in computer power supplies) are designed to regulate within a certain range under a load. The range is designed so that the regulator is able to adjust itself anywhere from a nominal rating to the maximum rating of the design during normal operating conditions. Because of Ohm's law, when the current draw on a regulator increases the voltage decreases and it's the regulator's job to adjust the voltage accordingly. However, when the supply is open or connected to a very low load outside the design specifications, the current draw may be lower than what the design was for so again, according to Ohm's law, the voltage will be high but it will be the lowest the supply can regulate under such a load.

 

[Indred]

Well to elabroate on what I tested. They computer had windows fully loaded Mobo, 5 hdds 1 DVD and 1 water pump connected plus 3 120mm fans and that was my reading. The only other thing i would say any thing about is that i tested it off a plug that was not under load. IE nothing else was connected to it, just stright from the PSU to my DMM. Unless i need to test it off of a connector that is in use i'm unsure if i messed it up (don't see how that bad) or what. I'll take some pictures once i get the charger for my camera back. Hope this helps in the comments. Thanks

 

[Bassyhead]

Originally posted by: Indred
Well to elabroate on what I tested. They computer had windows fully loaded Mobo, 5 hdds 1 DVD and 1 water pump connected plus 3 120mm fans and that was my reading. The only other thing i would say any thing about is that i tested it off a plug that was not under load. IE nothing else was connected to it, just stright from the PSU to my DMM. Unless i need to test it off of a connector that is in use i'm unsure if i messed it up (don't see how that bad) or what. I'll take some pictures once i get the charger for my camera back. Hope this helps in the comments. Thanks

All of the voltages from the supply will be the same so you can measure from any of the connectors. If you measured the right wires, then your readings should be valid.

 

[RossGr]

Originally posted by: CrispyFried
is the dmm on an ac setting? sounds about right for ac scale.

Actually not. If he had it on the AC scale he would read a few millivolts. That would be the ripple voltage.

 

[Navid]

Originally posted by: RossGr

Originally posted by: CrispyFried
is the dmm on an ac setting? sounds about right for ac scale.

Actually not. If he had it on the AC scale he would read a few millivolts. That would be the ripple voltage.

A regular meter cannot really measure RMS (what you supposedly measure when you use AC). What it really does is to first rectify the voltage. Then, the meter measures the average value. Then, it multiplies the measured value by about 1.2 to get the corresponding RMS value assuming it is measuring a sine wave.

Now, if you connect a DC voltage to it, its rectified value is the same and the multiplication will make the value wrong. This is very easy to test. Use your meter and measure the voltage of a battery on DC and then on AC. You will see that the AC measurement shows a higher value.
http://www.allaboutcircuits.com/vol_2/chpt_12/1.html

 

[RossGr]

Originally posted by: Navid

Originally posted by: RossGr

Originally posted by: CrispyFried
is the dmm on an ac setting? sounds about right for ac scale.

Actually not. If he had it on the AC scale he would read a few millivolts. That would be the ripple voltage.

A regular meter cannot really measure RMS (what you supposedly measure when you use AC). What it really does is to first rectify the voltage. Then, the meter measures the average value. Then, it multiplies the measured value by about 1.2 to get the corresponding RMS value assuming it is measuring a sine wave.

Now, if you connect a DC voltage to it, its rectified value is the same and the multiplication will make the value wrong. This is very easy to test. Use your meter and measure the voltage of a battery on DC and then on AC. You will see that the AC measurement shows a higher value.
http://www.allaboutcircuits.com/vol_2/chpt_12/1.html

Could you please relate this to my statement.

Perhaps you could be a bit more specific about the type of meter you are using.
If you measure a higher AC voltage then DC off of a battery, you need to get your voltmeter fixed, I measured 1.327 VDC and .3mVAC

Are you saying that you cannot measure the ripple voltage of a powersupply?

I will admit that the frequency of the ripple will be quite high in comparrison to an old linear supply but should still be present and measurable.

 

[RossGr]

Originally posted by: Indred
Well to elabroate on what I tested. They computer had windows fully loaded Mobo, 5 hdds 1 DVD and 1 water pump connected plus 3 120mm fans and that was my reading. The only other thing i would say any thing about is that i tested it off a plug that was not under load. IE nothing else was connected to it, just stright from the PSU to my DMM. Unless i need to test it off of a connector that is in use i'm unsure if i messed it up (don't see how that bad) or what. I'll take some pictures once i get the charger for my camera back. Hope this helps in the comments. Thanks

If your mobo was plugged in and running then you were checking a loaded condition at the connector, you do not have to be using the connector. What you do not want to do is to energize a ATX power supply without a load for each voltage.

 

[Navid]

Originally posted by: RossGr

Originally posted by: Navid

Originally posted by: RossGr

Originally posted by: CrispyFried
is the dmm on an ac setting? sounds about right for ac scale.

Actually not. If he had it on the AC scale he would read a few millivolts. That would be the ripple voltage.

A regular meter cannot really measure RMS (what you supposedly measure when you use AC). What it really does is to first rectify the voltage. Then, the meter measures the average value. Then, it multiplies the measured value by about 1.2 to get the corresponding RMS value assuming it is measuring a sine wave.

Now, if you connect a DC voltage to it, its rectified value is the same and the multiplication will make the value wrong. This is very easy to test. Use your meter and measure the voltage of a battery on DC and then on AC. You will see that the AC measurement shows a higher value.
http://www.allaboutcircuits.com/vol_2/chpt_12/1.html

Could you please relate this to my statement.

Perhaps you could be a bit more specific about the type of meter you are using.
If you measure a higher AC voltage then DC off of a battery, you need to get your voltmeter fixed, I measured 1.327 VDC and .3mVAC

Are you saying that you cannot measure the ripple voltage of a powersupply?

I will admit that the frequency of the ripple will be quite high in comparrison to an old linear supply but should still be present and measurable.

You state that a meter measures ripple. I don't think so!
Yes, I am saying you cannot use a general purpose volt meter to measure the ripple of a voltage.

An AC volt meter is supposed to measure RMS (Root Mean Square).
The RMS value of a clean DC voltage is not 0. It is equal to its DC value. If the DC voltage is not clean, and has ripple, the RMS value will not be equal to the ripple. The RMS value of a voltage, by definition, is equal to the DC voltage of a clean DC voltage that has the exact same amount of energy.

When we say that the line voltage in the states is 120V, that does not mean that the voltage has a 120V ripple! In fact, the US line voltage goes from negative 170V to positive 170V each cycle. So, the ripple is 170V peak. But, the RMS value for it is 120V, which means that if you measure its energy, it is equal to the energy of a 120V DC voltage.

The only way you can measure ripple is to first remove the DC component using perhaps a capacitor in series. But, that would not be RMS, which is what an AC volt meter is supposed to measure.

What meter do you have? What model? Does it have a peak-to-peak setting, that you used, as well as another RMS setting for AC?

 

[GalvanizedYankee]

I'm not an engineer and take the blue collar view of "where the rubber meets the road" to problem solving.

OP, wet cells in good condition develope 2.2V. Your car battery is 6x2.2=13.2V. Check your DDM against a known good car battery. Check it against the CMOS battery. Check it against a dry cell. Set it to VAC, check your mains. This testing might indicate a RMA is in order.

NOTE: With the engine at high idle the reading at it's battery should be 13.75~13.9V. The 13.2V would be with the engine off. Clean the probs of any sulfuric acid so as not to introduce it to the rig

For the blue collar service tech http://repairfaq.ece.drexel.edu/sam/smpsfaq.htm

For the well versed EE or up-and-coming EE http://www.smpstech.com/

The older SMPS were very tolerant of no load operation but the newer designs are not. Transformer saturation can/will damage/kill transistors. If one looks at any current SMPS spec sheet minimum load is listed. Cheap POS PSUs have weak if any no load shut down circuits or over voltage for that matter.

I've only tested/adjusted two PSUs on the bench. Several high output fans are clipped to the 5V and two old car headlights are clipped to the 12V. The unit is left to stabilize for an hour, then readings taken and pots tweaked. The cover is reinstalled and voltage checked again, because EMF can be altered by the steel cover being inplace. A variance of voltage was not noted but a change in EMF can effect voltage output to a minor degree.


...Galvanized

 

[RossGr]

Originally posted by: Navid

Originally posted by: RossGr

Originally posted by: Navid

Originally posted by: RossGr

Originally posted by: CrispyFried
is the dmm on an ac setting? sounds about right for ac scale.

Actually not. If he had it on the AC scale he would read a few millivolts. That would be the ripple voltage.

A regular meter cannot really measure RMS (what you supposedly measure when you use AC). What it really does is to first rectify the voltage. Then, the meter measures the average value. Then, it multiplies the measured value by about 1.2 to get the corresponding RMS value assuming it is measuring a sine wave.
EDIT:
I just went to the Fluke site and opened the operators manual for a Fluke 187 (I bellieve that this is a middle of the road model). Clearly they have changed the input circiuury from the Dvarsonal movement meters that I learned electronics on. With this Fluke you get BOTH the DC and AC components, they can be displayed seperately or combined. So YES you can do very nice PS ripple measuremens with this meter.
Now, if you connect a DC voltage to it, its rectified value is the same and the multiplication will make the value wrong. This is very easy to test. Use your meter and measure the voltage of a battery on DC and then on AC. You will see that the AC measurement shows a higher value.
http://www.allaboutcircuits.com/vol_2/chpt_12/1.html

Could you please relate this to my statement.

Perhaps you could be a bit more specific about the type of meter you are using.
If you measure a higher AC voltage then DC off of a battery, you need to get your voltmeter fixed, I measured 1.327 VDC and .3mVAC

Are you saying that you cannot measure the ripple voltage of a powersupply?

I will admit that the frequency of the ripple will be quite high in comparrison to an old linear supply but should still be present and measurable.

You state that a meter measures ripple. I don't think so!
Yes, I am saying you cannot use a general purpose volt meter to measure the ripple of a voltage.

An AC volt meter is supposed to measure RMS (Root Mean Square).
The RMS value of a clean DC voltage is not 0. It is equal to its DC value. If the DC voltage is not clean, and has ripple, the RMS value will not be equal to the ripple. The RMS value of a voltage, by definition, is equal to the DC voltage of a clean DC voltage that has the exact same amount of energy.

When we say that the line voltage in the states is 120V, that does not mean that the voltage has a 120V ripple! In fact, the US line voltage goes from negative 170V to positive 170V each cycle. So, the ripple is 170V peak. But, the RMS value for it is 120V, which means that if you measure its energy, it is equal to the energy of a 120V DC voltage.

The only way you can measure ripple is to first remove the DC component using perhaps a capacitor in series. But, that would not be RMS, which is what an AC volt meter is supposed to measure.

What meter do you have? What model? Does it have a peak-to-peak setting, that you used, as well as another RMS setting for AC?

To the best of my knowledge, AC voltmeters have a series capacitor on the input to block DC voltages. So when you take an AC Voltage measurement you read the AC which is riding on the DC signal. I am pretty sure that modern DVMs have not changed this.
EDIT: I just went to the Fluke site and read the AC measurement section for a Fluke 187. Looks like you get both a true RMS reading as well as the seperate AC and DC components. So this is an even better tool for checking PS ripple then I thought. Looks like you do need to be aware of how your DVM treats AC signals.