Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Need help with MAX4172 Precision High Side Current Sense Amplifier Pls

Status
Not open for further replies.
hi Al,

Look at these plots from a spread sheet.

Showing.

1. Ideal Iload versus Vout , for reference

2. Showing Expected versus your Measured result [ they track very well]

3. Showing error difference between expected and measured

To me I would say the results are within measurement tolerances.?

Eric

Is not a problem trying to help you Al, keep the questions;) coming till we get it right.

In the last resort you have my home address.:rolleyes:
 
Last edited:

EDIT:

Set Rout for 2.442V at 1Amp.
This will give [2.442/5]*1024 = 500 decimal [0x1F4] counts in the ADC regs, do a left shift to give 1000 decimal [0x3e8].

Convert to ASCII for the display as 1000mA


E.

Hi Eric,

So from the sheet you posted I can assume the results are "good enough" :)

Sadly I still dont understand the above part you posted. I just don't get the conversion and how it takes care of the non-proportional nature of the results?

Also how does a left shift do what is required? For the last week or so I have been a little worse than usual in the concentration department, though they have not found any particular cause for my memory problems. But they have changed the tablets I take for it, maybe just for the fun of seeing what happens to me! ;)

Would it be easier to explain the theory if I phoned you? I'm sure I have your number.

Thanks again, Al
 
hi Al,
Give me 5, I will post.
E.

EDIT:
Al,
look at these values, I have shown 0 to 1000mA in one block and 0 to 100mA in the other block.
[the best resolution will be about 2mA, if you want a higher resolution you could use a 2.5V ref or amplify the Vout]

The ADC counts are shown in decimal, you will have to convert the ADC '10 Bit binary values to ASCII for displaying.

There are many binary to ascii convert subroutines

Eric

Edit2:
do you have the +V supply of the Max4172 connected OK to the input side of the 7805.???

Look at the Error plot on page #4 of the d/s,, less than 10mV Vsense, its awful!
 
Last edited:
Hi Eric,

As to my V+ connection, no it is connected after the reg. DOH! Now I see what you mean. I will reposition that immediately. Thanks again mate.

As to the 10mV, I see what you mean about the error. Are you suggesting that I use 100mV and if so how?

EDIT. just realised that I was using a zero in my additions, again DOH!

Yours, Al
 
Last edited:
Hi Eric,

As to my V+ connection, no it is connected after the reg. DOH! Now I see what you mean. I will reposition that immediately. Thanks again mate.

As to the 10mV, I see what you mean about the error. Are you suggesting that I use 100mV and if so how?

EDIT. just realised that I was using a zero in my additions, again DOH!

Yours, Al

Hi Al,

Assuming that your V input to the 7805 is say 9V, that means V out on the MAX can now be 9v-1.2V= ~7.5V.

You could now increase Rout to 4.88K, this would give in the ADC 1000 counts for 1Amp load current.

The way I read the d/s is you need to choose a Rsense that gives the optimum linearity for the application, see Table #1 on Page #5.

You could have a range select switch for a 0.1 Rsense from 100mA thru 1000mA and a 1.0 Rsense resistor for 10mA thru 100mA.

E.
 
Last edited:
Morning Eric,

The 7805 Vin is about 8.5v so thats enough but I thought of another potential problem. Would the contacts of the switch possibly have a resistance that could mess up the readings? I know many switches and relays boast a contact resistance of a few mOhms but that may be enough to screw up the readings?

Also I have been researching the way I was planning on limiting the current, using a 317T in current mode and for the life of me I can't find out if it needs a headroom voltage like it does in Vreg mode?

If it does need some headroom then I would have to put it before the 7805 and then I would also read any current used by that as well as the load.

This project is getting way more complicated than I envisaged! I only started it as a good way to use one of my recently acquired large 16x2 LCD displays!;)

Any further thoughts mate?

Al
 
hi Al,

Most regulators require a headroom voltage in order to operate correctly, for general purpose regulators about 2.5V to 3V is required.

If it does need some headroom then I would have to put it before the 7805 and then I would also read any current used by that as well as the load.

But your current sense is on the output side of the 7805.??

Do you have a negative supply on your power supply project.??

E.
 
hi Al,


But your current sense is on the output side of the 7805.??

Do you have a negative supply on your power supply project.??

E.

Hi Eric,

Doh! thats the old brainbox malfunctioning again! Of course it measures after the 7805!

As to the negative supply. Yes it does, I was planning on just measuring the actual voltage of this to display along side the positive variable side. I sometimes use the pos and neg rails without the GND to get the 30v that is needed to test various things like LED backlights from laptop screens which tend to have 8 or 9 LEDs in series, its quite handy for that, but since I don't do a lot with amplifiers and such I don't use it for what I presume is its original purpose.

Did you have something in mind that involves the negative rail?

Al
 
hi Al,
Do you have a 'concept' circuit diagram you could post for the complete circuit.?

It will give me an idea what I can suggest for the current limit and current measurement.

I assume when you say 'current limiter' you want an adjustable current limit.?

E.
 
Hi Eric,

Here is a quick sketch of what I'm doing.

Yes I do intend to limit the current output in 12 clicks ranging from 5mA to the full 1000mA.

Any ideas or tips would as always be much appreciated.

Thanks, Al
 
hi Al,
I am looking at the 5V current sense using an OPA and a 2N3819 [ which I believe you have?] or do you prefer to stick with the max4172.

The sense circuit is working simulation, but I will try it in hardware.

Do you still plan to use a 317T for the current limiter.?

As you are using a CT transformer with a full wave bridge for the +V and -V outputs they cannot be isolated, you can still connect across +V and -V for a higher SINGLE voltage.

E.

EDIT:

I see that you have a 15v, 0, 15v. 1A CT transformer, I guess you know you will not get +15Vout or -15Vout at full current load due to diode drops and the overhead voltage required for the LM317.

My best estimate would be ~12V

Whats the supply into the 7805??
 
Last edited:
Hi Eric,

I don't mind using other bits at all. I am happy as long as the job works.

I do have some 2N3819s but they are A's does this matter? I also have quite a few different op amps.

As to the transformer, I wrote the values that I actually measured at the outputs. Before the regs there is 18.5v on the variables and 8.5 on the 5v line.

Anyhow, just got to take my grandson home now, its the first time we've seen him for over a month due to his mum being awkward.

Al
 
hi Al,
Look at this LTSpice simulation.

I have assumed a 15v,0,15v CT, transformer, rated at 500mA.

The Vout for testing was set at +12V and the current limiter for 500mA.
The current limit resistors for the limit selection will be: I lim = 1.25v/Rlim

The highest current with 12Vout is approx 350mA, after that ripple is high on the output.

I will run some tests later, with current circuit attached.

Eric

Added:
Shows the +5Vreg with limiting and simple current measurement voltage circuit.
 
Last edited:
Hi Eric,

I have further measured the outputs of the PSU and I get 15v at 500mA on both variable rails. This is after all the smoothing and regulation.

I also notice that you don't know that the transformer has a separate pair of output wires for the 5v at 1A 5 wires out in total.

Also you have a "u3" in your schematic. What is it?

Al
 
Hi Eric,

I have further measured the outputs of the PSU and I get 15v at 500mA on both variable rails. This is after all the smoothing and regulation.

I also notice that you don't know that the transformer has a separate pair of output wires for the 5v at 1A 5 wires out in total.

Also you have a "u3" in your schematic. What is it?

Al

hi Al,
The U3 is a OPA, say a LM324 or equivalent

A 5V winding cannot be used to give a 5V regulated supply.

Any chance of your full circuit diagram, its a little difficult to fix any values in my circuits.;)

EDIT:
The transformer secondary outputs cannot be 15Vac, if you can get 15Vdc at 500mA after regulation.
 
Last edited:
Hi Eric,

Regarding getting 5v. The winding seems to be giving 8.5v to the 7805 so it has 3.5v headroom.

The other 3 connections are 18v before the lm317t's that controls the output voltage for each rail. So that side too has headroom.

One thing I have noticed though is that ONLY the variable side goes through a rectifier! The 5v side just goes to a large 4700u cap and straight on to a 7805. Weird that it has no rectification!

Edit: Just found the rectifier hidden under the transformer, obvious now I have stripped the unit!

I will try and do a circuit diagram but I am hopeless at drawing as you will have noticed. I wonder if I can find the diagram on the net? Nope!

But I did find this that has the same number that my original PCB bears "PSU206 Iss.4"
https://www.lascarelectronics.com/temperaturedatalogger.php?datalogger=130

They give exactly the same specs as mine has so I assume it is still around perhaps in a newer incarnation, for instance my transformer is open frame.

Anyway I will try to do a schematic for you Eric, just don't laugh at it too much. ;)

Al
 
Last edited:
Hi Eric,

Just a bit of extra info on the current available.

I am now certain that the 5v has 1A available but the variable rails share 500mA between them 250mA pos and a further 250mA for the neg side. I assume this means that even when using both rails without the center ground it will only give 250Ma?

Is that right?

Al
 
Hi Eric,

Just a bit of extra info on the current available.

I am now certain that the 5v has 1A available but the variable rails share 500mA between them 250mA pos and a further 250mA for the neg side. I assume this means that even when using both rails without the center ground it will only give 250Ma?

Is that right?

Al

hi Al,
If its a centre tapped transformer, say at 18V per winding, rated at 500mA max, then thats 250mA max per winding.

Without the CT being used thats 36V at 250mA.

Is the VA rating and voltage marked anywhere on the transformer.?

If not, can you measure the cross sectional area of the transformers centre, the end of the transformer with the windings on.

If its two separate 18V windings each rated at 500mA

E.
 
Hi Eric,

Sorry to take so long to get back, had some more grandson problems.

I have gone cross eyed and allsorts trying to do a schematic of my PSU. Tried a couple of ways of doing one on the PC and failed dismally so decided just to draw it and scan it.

I hope it is good enough and that I have got all of it right. Even though it is single sided I had problems seeing all the routes as the component side is mainly heatsinks and transformer.

I have further measured the transformers outputs and at no load the +5 is actually 8v only dropping to 7.2 at a full 1A load.

The variable voltages are 23v at the input to the regulators and still maintain 18.5 at full load, thats roughly the same for both pos and neg sides.

Hope this is suitable to get some good ideas and a working ADC/ IADC.

BTW the other thread I started about writing to LCDs got way beyond me and two people I respect and who have both given me advice in the past got into a rather heated discussion about methods and I was frightened to reply in case either thought I was taking sides. The truth is that both were way above my knowledge level. I just didn't want two really helpful and respected members to argue because of something I started. Both Mike and Roman have been helpful to me in the past and I feel a bit guilty about starting the thread.

Anyway below is the scan, done by my good wife Dawn for me, as I don't have a scanner now and she has taken over the computer guru role!

EDIT: Would there be any problem in permanantly joining the grounds from both? I have done it temporarily in the past to try things but I wondered if there is any detriment to this. Can you foresee any situation where they would NEED to be separate? I am thinking that to measure the variable voltages I would need to join the grounds if I use the 5v side to power the ICs?
 
Last edited:
Something else I forgot to mention is that there are definately two separate windings to the transformer that are split by a nylon divider.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top