amperage requirements?

[the_doc735]

On 10/15/2018 at 10:08 PM, Tomslin said:

I haven’t read this thread from the beginning so I thought your intention was something about DIY.

1)   But now, your reason for doing these measurements, as you explained here, seems quite strange as I see it. I mean, why should you as a not professional builder/designer, serve them with these rather advanced measurements? I don’t see it as a sensible reason to make such advanced measurements just to get a price.

 

2)   You will probably pay expensive for the product in any case. It's not any rocket research behind to build a power supply to a computer, although some may want to make you believe it.

 

3)   Put there a pico-psu, you can power it from a regular 12V car battery, then measure the actual current draw during to the worst power conditions to the pico, and it's done.

 

4)   Also don't hurt to tell what things you have in your computer.

 

5) If they not can build you a good power supply after this, I would be really surprised.

 

....Then I’m also grateful for an explanation

Please remember I was only respectfully answering your questions!

 

1)   If you read the entire thread you'll see that I haven't mentioned DIY?

 

2)   One explanation from Paul hynes is that: "the power requirement will affect the parts like the case and heat sinks for dissipation etc. The lower the power requirement, then the smaller the case, the smaller transformers etc. Conversely, the more powerful and bigger the PSU the more it will cost, so if I have those figures I can tailor it precisely for its purpose; if I don't have those figures I would have to massively over specify the PSU and as a consequence the final cost would be considerably more."

Similar comments from others, but more modifying what they already have available rather than a complete build.

 

3)   I don't see how measuring the load to a pico from a battery would help me get the 3.3, 5, 12v load measurements though?

 

4)   My third post in this thread lists the components in my PC.

 

5)   I'm sure they can build good PSU's that are suitable for my purpose once they have the correct info.

 

   ....I haven't ruled out DIY yet though.

 

Nothing offensive or rude. 

 

a)    no I don't think you are stupid. I know from your posts that you are very knowledgeable in this area.

 

b)    But you asked what a quoted price had to do with amperage and other specifications, I am simply answering your question with comments I have received from the builders themselves.

 

c)    It is your prerogative to think that I am not in contact with, or have never made contact with any of these builders, that is your choice, mind you I don't know what evidence you have to suggest that assumption? How could you possibly know that?

 

d)    It is up to those designers if they wish to participate in this thread, not me.

 

e)    I don't see any shame in not having used a clamp meter before and researching how to use it, I don't feel inferior in any way just because this is the first time I have used a clamp meter, what's wrong with learning how to use a device? 

 

Nothing offensive or rude 'here', take care.

 

Many apologies if I offended you somehow?

 

[the_doc735]

On 10/16/2018 at 12:17 AM, Superdad said:

Who is "they?"  I don't see any reference here in this thread to who might be willing to build you a full custom multi-rail ATX LPS.  And are you prepared to pay the price?

Please remember I was only respectfully answering your questions!

 

As I already said above: "people like paul hynes, sean jacobs, Larry (HDPlex), T Pardo, etc."   

 

...some are mods. of gear they already have, others are 'ground up' builds. Others are: Kenneth Lau, John Wood, Paul Baldwin, Paul Pang. No they are not all ATX LPS, but serve the: 3.3, 5, 12v requirements.

 

I don't know if I am prepared to pay the price until they come back with estimates/quotes.

 

I haven't ruled out DIY at this stage though.  

 

This is more probing stuff that doesn't answer my original question, which nobody seems to know? That's why I am trying to find out my self with research and spending time and money on measurement equipment like clamp meters.

 

I wish I could afford your LPSU's but it's a little beyond my price range unfortunately. Never seems to be any decent secondhand ones hanging about out there! 

 

Nothing offensive or rude 'here', take care. 

 

Many apologies if I offended you somehow?

 

 

[marce]

14 hours ago, the_doc735 said:

I know that!

 

...."so whatever you put in front of it is not going to make much difference..."

plenty would debate this I think?

I would suggest you look up simultaneous switching noise as a start...

Then the is the issue of motherboard layout and house the interface chips have been designed to minimize the layers required for a basic motherboard and how this has an effect on noise levels... Generic motherboards are designed to a cost, they do the job, but I would not call them a low noise design.

[Tomslin]

No, I don't think you have offended me in any way. I was only greatly surprised by this strange demands of these designers, in order for you to get a price. Because I believe that it are unnecessary advanced measurements to do for an amateur, and it are information that a good designer can get in other ways. When you start explaining these things to me that are on the ground to basic level, so excuse me, I asked you if you think I'm stupid.

But anyway, if you and your designer have agreed to that methods and both parties are satisfied with this solution, so let it go. It’s not my problem, fortunately.

When you post up things in that way you do in this thread (and some others too), at least for me it looks like it's more of a school assignment. You want everything and everyones views at the same time.

I only point out that you are having trouble doing these measurements, and it’s also related to what I wrote here above.

How was the result of your last measurements, when you were to measure one cable at a time and sum them?

 

 

[the_doc735]

19 hours ago, marce said:

I would suggest you look up simultaneous switching noise as a start...

Then the is the issue of motherboard layout and house the interface chips have been designed to minimize the layers required for a basic motherboard and how this has an effect on noise levels... Generic motherboards are designed to a cost, they do the job, but I would not call them a low noise design.

also people 'rave' about NUC's but they are still PC's! Different manufacturers will design different PCB on which to put their 'mundorf's' etc (e.g. asus, supermicro etc.) but they are still PC mobo's of various sizes (except those that are server boards of course). There is an opinion that suggests that once you achieve 1 millivolt or less (e.g. uV, nV....) ripple, it becomes irrelevant, it is no longer the most important factor? Also I still remember reading about someone on here that preferred a NUC to a Zenith! I've read about switching technology and Ifi Ipower active noise cancellation. Which boards are a low power design? 

[the_doc735]

18 hours ago, Tomslin said:

No, I don't think you have offended me in any way. I was only greatly surprised by this strange demands of these designers, in order for you to get a price. Because I believe that it are unnecessary advanced measurements to do for an amateur, and it are information that a good designer can get in other ways. When you start explaining these things to me that are on the ground to basic level, so excuse me, I asked you if you think I'm stupid.

But anyway, if you and your designer have agreed to that methods and both parties are satisfied with this solution, so let it go. It’s not my problem, fortunately.

When you post up things in that way you do in this thread (and some others too), at least for me it looks like it's more of a school assignment. You want everything and everyones views at the same time.

I only point out that you are having trouble doing these measurements, and it’s also related to what I wrote here above.

How was the result of your last measurements, when you were to measure one cable at a time and sum them?

 

 

I'm sorry if I ask my questions like a school assignment, I apologise, this is no reflection on you and your knowledge. Can't remember what school assignments look like anymore as that was 42 years ago (LOL!).

Again, I am sorry and apologies, but it is true, yes I do like to hear both sides of the story i.e. for and against. Most things do seem to have two sides to it; like someone accused of murder is entitled to a fair case for a defence, no matter what anyone else might presume without any evidence. So, yes, I do usually ask for everyone's views on any particular issue before I make my mind up on what to do next. I just hope it doesn't annoy you too much? e.g. like the SMPS Vs. LPSU issue for example, some put forward a case for LPS and others a case for SMPS. I see now there is no best, it depends on the circumstances, but it's getting the right power supply in the right place in any given configuration. I suppose I will always be that way inclined.

 

Oh yes! ~ the summed measurements, I only just did in the last few minutes:

 

3.3v boot:

0.4??

0.54?

0.2??

0.4??

=1.54A

 

5.0v boot:

0.16?

0.2??

0.17?

0.34?

0.24?

=1.11A

 

12.0v boot: 

0.5??

0.5??

=1.00A

 

? = too fast for me to see!

No peak hold! NO max./min.

Set to: 2 DCA

 

Many Thanks!

[Tomslin]

Thus, power consumption below 30W without the CPU.

[the_doc735]

9 hours ago, Tomslin said:

Thus, power consumption below 30W without the CPU.

I've sold that clamp meter to buy another one with max./min. function in order to compare. Plus I will do the CPU readings as well.

 

many thanks!

[Tomslin]

12 hours ago, the_doc735 said:

I've sold that clamp meter to buy another one with max./min. function in order to compare. Plus I will do the CPU readings as well.

 

many thanks!

 

Have any manufacturer or anybody else said you should do it, exactly in that way?

[the_doc735]

23 minutes ago, Tomslin said:

 

Have any manufacturer or anybody else said you should do it, exactly in that way?

the only person who has said to sum up all 3.3, 5, & 12v cables in the 24 pin ATX connector + CPU, is you. So that is what I am now doing.

Kind Regards.

[Tomslin]

52 minutes ago, the_doc735 said:

the only person who has said to sum up all 3.3, 5, & 12v cables in the 24 pin ATX connector + CPU, is you. So that is what I am now doing.

Kind Regards.

 

No, I have not said that. I have always maintained that it are unnecessary measurements to do. Definitely when it's not about DIY, and especially for you who barely know what you are doing. I’ve also stated that you can as well measure the power consumption via a pico psu, which becomes much easier to do. Before I even came into this thread, I saw you did measuring at only one cable at a time on one rail to 24 pin ATX. And it’s clearly visible on your attached photo earlier in this thread. In this case you have to sum up all together. Do you understand the difference here? If you doing this only because you think I've said it, you can finish it. Don't buy any more expensive clamp meter, simply because you will not achieve any outstanding results.

[the_doc735]

20 hours ago, Tomslin said:

 

No, I have not said that. I have always maintained that it are unnecessary measurements to do. Definitely when it's not about DIY, and especially for you who barely know what you are doing. I’ve also stated that you can as well measure the power consumption via a pico psu, which becomes much easier to do. Before I even came into this thread, I saw you did measuring at only one cable at a time on one rail to 24 pin ATX. And it’s clearly visible on your attached photo earlier in this thread. In this case you have to sum up all together. Do you understand the difference here? If you doing this only because you think I've said it, you can finish it. Don't buy any more expensive clamp meter, simply because you will not achieve any outstanding results.

 

No, I have not said that. I have always maintained that it are unnecessary measurements to do.

But you said: “you were to measure one cable at a time and sum them”

 

As I stated earlier Paul Hynes & others have asked for my Volts/Amps/Wattage figures and I already explained to you why they asked for this in an earlier post.

 

Definitely when it's not about DIY, and especially for you who barely know what you are doing.

Please refrain from personal/judgemental remarks, please don’t keep on and on saying that I don’t know what I am doing, I ALREADY KNOW YOU HAVE FAR SUPERIOR KNOWLEDGE THAN ME!

 

 I’ve also stated that you can as well measure the power consumption via a pico psu, which becomes much easier to do.

Yes know I can measure with a pico on its DC input cable, WITH A CLAMP METER.

 

Before I even came into this thread, I saw you did measuring at only one cable at a time on one rail to 24 pin ATX. And it’s clearly visible on your attached photo earlier in this thread. In this case you have to sum up all together.

Yes I did do that, I admit! Thank you for suggesting to sum the results together, which I did in an earlier post, with the results in this thread. Remember?

 

Do you understand the difference here?

No, I don’t see what difference you are talking about? Sorry!

 

If you doing this only because you think I've said it, you can finish it.

I am doing ‘this’(?) i.e. taking measurements because people like paul hynes have asked for these measurements and I am fulfilling their requests.

 

Don't buy any more expensive clamp meter, simply because you will not achieve any outstanding results.

I am not looking for an outstanding result, I am looking for a peak result. Which my previous meter could not provide (because that function wasn’t available).

Many Thanks! 

[the_doc735]

NOTE: on the pico!

 

how would I measure: 3.3, 5, 12v separately via a pico? (given that it has a 12DCV single rail input)

[Tomslin]

I really hate when single sentences or pieces are taken from their context, and interpreted individual and/or differently. You are a master of doing that (not in positive sense, though). When it happens I completely lose interest and ambition.

I really need to consider if I would like to go another new round in this.

My gut feeling usually is correct and it tell me: NO!

[the_doc735]

32 minutes ago, Tomslin said:

I really hate when single sentences or pieces are taken from their context, and interpreted individual and/or differently. You are a master of doing that (not in positive sense, though). When it happens I completely lose interest and ambition.

I really need to consider if I would like to go another new round in this.

My gut feeling usually is correct and it tell me: NO!

Sorry, I'm only answering your questions, one by one. I apologise.

I can lump it all together if you prefer, but I thought 'this' 1 by 1 response was easier to understand?

Same reply in one go!

 

As I stated earlier Paul Hynes & others have asked for my Volts/Amps/Wattage figures and I already explained to you why they asked for this in an earlier post. Please refrain from personal/judgemental remarks, please don’t keep on and on saying that I don’t know what I am doing, I ALREADY KNOW YOU HAVE FAR SUPERIOR KNOWLEDGE THAN ME! Yes I know I can measure with a pico on its DC input cable, WITH A CLAMP METER. Yes I did do that, I admit! Thank you for suggesting to sum the results together, which I did in an earlier post, with the results in this thread. Remember? No, I don’t see what difference you are talking about? Sorry! I am doing ‘this’(?) i.e. taking measurements because people like paul hynes have asked for these measurements and I am fulfilling their requests. I am not looking for an outstanding result, I am looking for a peak result. Which my previous meter could not provide (because that function wasn’t available). Many Thanks 

 

I completely understand if you are not able to answer my latest post. I would like to thank you for the effort you have made up to this point AND YOU HAVE HELPED ME A LOT! (with your words and pictures and diagrams from various threads). I know quite a bit more now because of you, and your help. i.e. wiring and measuring a 24 pin ATX wiring loom, very grateful. I am still getting a clamp meter with a peak hold which will come in very handy for all sorts of things in the future, that involves electricity (LOL). Of course I will use this on the individual 24 pin ATX wires and sum them up in each group, and then send them to paul hynes and other LPSU builders. As you don't wish to participate any further in this thread, I will only post these results for you - if you want me to? If you ask?

I may of course post the results for OTHER MEMBERS to look at, in this thread.

Cheers! 

take care!

[the_doc735]

With: Protmex MS2108 6600.

Set to DCA/Max.HOLD. Zero function “off”.

23/10/2018.

 

 

3.3v boot:

00.34

00.33

00.05

00.15

=00.87A

 

5.0v boot:

00.12

00.14

00.08

00.04

00.23

=00.61A

 

12.0v boot: 

00.26

00.09

=00..35A

 

These figures are to help someone supply me with LPSU’s.

Has anyone else tested the MSI Z370i GAMING PRO CARBON AC mini-ITX mobo 24 pin ATX socket values? 

[the_doc735]

does anyone know if you can measure DCA like this pic. with a conventional DMM, instead of a clamp meter with hall effect?

 

[Superdad]

30 minutes ago, the_doc735 said:

does anyone know if you can measure DCA like this pic. with a conventional DMM, instead of a clamp meter with hall effect?

 

Yes, that is the better way to do so--just as I suggested to you previously. 

[the_doc735]

6 minutes ago, Superdad said:

 

Yes, that is the better way to do so--just as I suggested to you previously. 

Ar yes, thank you! I must of misinterpreted your suggestion - Sorry! It's part of my learning difficulty and how information is conveyed to my brain in an unorthodox fashion. Even at my age I still get frustrated by it sometimes.

I'll use a cheap extension and butcher that rather than the actual cables themselves.

Cheers!

[jabbr]

Ok this is killing me ...

 

https://web.archive.org/web/20160414133738/http://www.formfactors.org/developer/specs/Power_Supply_Design_Guide_Desktop_Platform_Rev_1_2.pdf

 

page 79

[marce]

Also note the requirement for a separate +5VSB supply. so if someone puts something dodgy in a USB port it doesn't bring the system down.....

[Solid-State]

revision 1.2

February 2008

 

 

and in 2008 we used some good old Seagate Cheetah SCSI HDD that drew power like a whole house,
The power supply would also provide 1-5 HDD with power at least

[jabbr]

6 hours ago, TubeMan said:

revision 1.2

February 2008

 

 

and in 2008 we used some good old Seagate Cheetah SCSI HDD that drew power like a whole house,
The power supply would also provide 1-5 HDD with power at least

Yep ... but it has guidelines for current at various voltages — and diagrams. More recent specs aren’t as detailed. But read those also!

 

Two possibilities:

1) you are buying something already designed (linear ATX) : get a minimum of 80W for very lowest power ATX. 120W-160W for low power and 200W is more reasonable. 

 

2) you are DIY — either find a project that has already been built on somewhere like DIYAudio OR if you are doing this yourself then you need to read everything you can find about ATX power supply design.

If you want to wing it go 4-6A for each rail... or 10x what you measure. 

[the_doc735]

On 10/23/2018 at 8:33 PM, Superdad said:

 

Yes, that is the better way to do so--just as I suggested to you previously. 

you and others say this is the preferred way to measure, but why is a DMM preferred over a clamp meter please?

Even paul hynes said: "inserting a low value resistor in series with the rail output, then measuring the voltage drop across the resistor is preferred". I don't pretend to understand this one though! Care to explain this voltage drop method as I thought I was measuring current/load/amps?

 

Many thanks!

[Superdad]

1 hour ago, the_doc735 said:

Care to explain this voltage drop method...

 

Not really, but I will partially. With an ammeter you break the wire and insert the ammeter into the circuit.  If you have only a voltmeter then measure the drop and calculate.  Here is a easy Youtube example:

 

I still do not understand why you don't just pick a 5-7 amp LPS and be happy.  What's your budget?