Maker Pro
Maker Pro

2M+ ohm resistors not resisting

KJ6EAD

Aug 13, 2011
1,114
Joined
Aug 13, 2011
Messages
1,114
This is the area of the board we're interested in and the two resistors we need the values from. Your picture is still out of focus. .PNG files are preferred. Maybe you can measure the resistors.
attachment.php
 

Attachments

  • 12-24.jpg
    12-24.jpg
    12.6 KB · Views: 238

schmidtbag

Nov 8, 2012
36
Joined
Nov 8, 2012
Messages
36
This is the area of the board we're interested in and the two resistors we need the values from. Your picture is still out of focus. .PNG files are preferred. Maybe you can measure the resistors.
attachment.php

With a 10 ohms in place of the trimpot (I can't leave it disconnected or else the power source shuts down from overload), I get about 30V from the resistor on the left (according to that image) and about 12v with the resistor on the right. I have an analog multimeter so I can't get 100% accuracy.
 

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
Here's the long explanation for all of this. Please Google any terms you're not familiar with.

The regulator circuit is controlled by the controller chip, the 8-pin SOIC (small-outline IC) that may be a UC3843 "current-mode controller". This chip generates an ON/OFF control signal to the big switching MOSFET, the STP75NF75FP. This MOSFET forms a switching regulator using the non-isolated flyback configuration, along with the inductor (the large ring on the top side with heavy wire wound around it) and the diode (the STPS20H100CT), which steps the input voltage up to a higher voltage.

The timing of this control signal, from the controller IC, controls the voltage-step-up operation of the flyback circuit, and therefore, determines the output voltage. The controller chip monitors the output voltage, so it can generate the correct control signal to give the correct output voltage, through the "voltage feedback" circuit.

The voltage feedback circuit is a voltage divider (Google it), connected to the output, and it provides a certain fraction of the output voltage back into the controller chip on its "VFB" (voltage feedback) pin, which is pin 2. The controller chip adjusts its output so that the voltage on this pin is equal to an internal reference voltage, which (for the UC384x) is 2.5V.

The controller chip controls the switching MOSFET so that there's always 2.5V on the VFB pin. In simple terms, if the controller chip sees that the voltage on the VFB pin is too low (less than 2.5V), it "increases" the drive signal to the MOSFET to compensate; if the VFB voltage is too high, the controller chip "decreases" the signal. This is how the voltage regulation of the circuit is achieved.

SO, the output voltage is controlled by the voltage feedback circuit, which sends a certain fraction of the output voltage back to the controller chip as "voltage feedback". Here is what I think the voltage feedback circuit looks like. R1 is the little resistor nearest to the trimpot, and R2 is the other resistor further over, past the three empty component positions.

Vout ----------\/\/\/\/--------\/\/\/\/---------------\/\/\/\/-------GND
. . . . . . . . Trimpot . . . . R1 . . . . | . . . . R2
. . . . . . . . . . . . . . . . . . . . . . . . .-------------------------------> VFB (pin 2)

This is a "voltage divider" circuit. Its "input" is the board's output voltage, and its "output" feeds the VFB pin of the controller chip.

The controller chip adjusts the output voltage so that the VFB voltage is 2.5V. So you can draw the voltage divider as:

Vout ----------\/\/\/\/--------\/\/\/\/---------------\/\/\/\/----------GND
. . . . . | . . Trimpot . . . . R1 . . . . | . . . . R2 . . . . |
. . . . . | . . . . 5k . . . . . .??? . . . .| . . . .???. . . . .|
. . . . 24V . . . . . . . . . . . . . . . . 2.5V . . . . . . . . 0V

I've marked in the new output voltage that you want.

The trimpot's resistance is 10 kilohms end-to-end. We want to set it around half way, so we'll say its resistance is 5k. Now, if you know the resistances that are currently in the R1 and/or R2 positions, you can use the voltage divider calculations to work out what the resistances need to be changed to.

When you tell me the resistance of R1, I will do those calculations for you and tell you what to do. Ideally you should replace R1 with a resistor of a lower value, but the resistors are SMT (surface-mount technology) components, so this requires some careful desoldering and resoldering, and you can (at least temporarily) just tack another resistor in parallel with R1.
 

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
You need to turn the power supply OFF.

Disconnect everything from the trimpot position, then measure RESISTANCE, not VOLTAGE, of those two resistors.

Actually, just measure the one closest to the trimpot. Your analogue multimeter could damage the IC if you measure the other one.
 

MrEE

Apr 13, 2012
84
Joined
Apr 13, 2012
Messages
84
It may be that the value of the sense resistor (the large one that says R010) needs to be adjusted (increased) in order to obtain the desired range for the output voltages.
 

schmidtbag

Nov 8, 2012
36
Joined
Nov 8, 2012
Messages
36
You need to turn the power supply OFF.

Disconnect everything from the trimpot position, then measure RESISTANCE, not VOLTAGE, of those two resistors.

Actually, just measure the one closest to the trimpot. Your analogue multimeter could damage the IC if you measure the other one.

I figured that's what you meant but once you mentioned disconnecting the trimpot, that's what thew me off and made me think you wanted to check voltage. Anyways, the resistor you're looking for is a 10k.
 

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
OK here are my calculations.

With the trimpot wound down to 0 ohms you get about 35V out, right?

Vout ----------\/\/\/\/--------\/\/\/\/---------------\/\/\/\/----------GND
. . . . . | . . Trimpot . . . . R1 . . . . | . . . . R2 . . . . |
. . . . . | . . . . 0 . . . . . . .10k . . . .| . . . .???. . . . .|
. . . . 35V . . . . . . . . . . . . . . . . 2.5V . . . . . . . . 0V

This means R2 is about 770 ohms. And the values you want are:

Vout ----------\/\/\/\/--------\/\/\/\/---------------\/\/\/\/----------GND
. . . . . | . . Trimpot . . . . R1 . . . . | . . . . R2 . . . . |
. . . . . | . . . . 5k . . . . . .??? . . . .| . . . .770. . . . .|
. . . . 24V . . . . . . . . . . . . . . . . 2.5V . . . . . . . . 0V

R1 needs to be about 1620 ohms. You can get this value by putting a resistor in parallel with the existing 10K resistor. That parallel resistor should be about 1930 ohms. The closest preferred value is 1.8 kilohms.

So you should get a 1.8 kilohm resistor and tack it on in parallel with (directly across) the 10K resistor.

If you can, use a SMT resistor of the same size, mounted directly on top of the existing resistor. Solder one end at a time so nothing moves around.

If you need to use a THT (through-hole technology) resistor (i.e. one with wire leads), get the smallest THT resistor you can find, anchor the resistor firmly to the PCB next to the existing resistor, with glue (hot melt glue is good for this), then bend the wires into position, and solder one end at a time.
 

schmidtbag

Nov 8, 2012
36
Joined
Nov 8, 2012
Messages
36
OK here are my calculations.

With the trimpot wound down to 0 ohms you get about 35V out, right?

Vout ----------\/\/\/\/--------\/\/\/\/---------------\/\/\/\/----------GND
. . . . . | . . Trimpot . . . . R1 . . . . | . . . . R2 . . . . |
. . . . . | . . . . 0 . . . . . . .10k . . . .| . . . .???. . . . .|
. . . . 35V . . . . . . . . . . . . . . . . 2.5V . . . . . . . . 0V

This means R2 is about 770 ohms. And the values you want are:

Vout ----------\/\/\/\/--------\/\/\/\/---------------\/\/\/\/----------GND
. . . . . | . . Trimpot . . . . R1 . . . . | . . . . R2 . . . . |
. . . . . | . . . . 5k . . . . . .??? . . . .| . . . .770. . . . .|
. . . . 24V . . . . . . . . . . . . . . . . 2.5V . . . . . . . . 0V

R1 needs to be about 1620 ohms. You can get this value by putting a resistor in parallel with the existing 10K resistor. That parallel resistor should be about 1930 ohms. The closest preferred value is 1.8 kilohms.

So you should get a 1.8 kilohm resistor and tack it on in parallel with (directly across) the 10K resistor.

If you can, use a SMT resistor of the same size, mounted directly on top of the existing resistor. Solder one end at a time so nothing moves around.

If you need to use a THT (through-hole technology) resistor (i.e. one with wire leads), get the smallest THT resistor you can find, anchor the resistor firmly to the PCB next to the existing resistor, with glue (hot melt glue is good for this), then bend the wires into position, and solder one end at a time.

Nice, the 1.8K worked. I'm reading about 25V so I'm sure a 1.9K might've been better but I might've burnt the original 10K a little bit - its hard to solder such tiny pieces. 25V should be fine for what I'm doing though.

I'll mark this as solved. Sorry about my incompetence, hopefully the next time a situation like this comes up, I won't be the one trying to figure it out (I'm working on a project with someone who knows less about electronics than me - I'm a programmer). In the event I'm responsible for electronics again, I'll probably be back here but I'll wait a day before my first response so I'm less flustered, more awake, and more alert.

Anyways, thanks for the effort and the help.
 

KJ6EAD

Aug 13, 2011
1,114
Joined
Aug 13, 2011
Messages
1,114
You should be able to adjust the trimpot to get 24V now.
 

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
Cool. You're welcome, apology completely accepted. Now, as I said before, the regulator may not meet its full specifications at 24V. The components (specifically, the big inductor) were chosen for an output voltage range of 35V upwards, and it COULD make a difference. Things you should check for are:

1. The regulator may not regulate accurately with no load, or a light load. Put a load of around 1A on it, and check that the output voltage is still 24V or very close, and adjust the trimpot if it isn't.

If the voltage is too high with no load, or a light load, and if this will be a problem, you may need to add a resistor to permanently load it down enough to keep it in regulation. This resistor will waste power and will get hot. Let us know if you need to do this.

2. The regulator may not be able to provide its specified maximum current. Check that the output voltage stays in regulation at the maximum load current that you want to draw from it. If the input voltage comes from a battery or other variable voltage source, do this check with the input voltage at the lowest limit, i.e. the lowest input voltage that you will be using in your application.

One more comment.

Usually, when you have a resistor and a trimpot in series like that, the values are chosen so that the resistor is the largest part of the total resistance, typically around 80% of the total resistance needed, and the trimpot provides the "fine adjustment".

In this case, the total resistance you need (trimpot plus "R1") is about 6.5 kilohms, and it would be more appropriate to use, say, a 5.6k resistor and a 2k trimpot, which would be adjusted to about half way. This ensures that (a) the voltage can be set fairly accurately, even with a single-turn trimpot, and (b) it's not possible to adjust the voltage a long way away from the desired voltage of 24V.

At the moment, it's the other way round - the trimpot provides about 80% of the total resistance.

In this case, the trimpot is a multi-turn one, so you can adjust the voltage pretty accurately, and reason (a) doesn't apply, but reason (b) still does, and it's up to you whether you want to change the components so the adjustment range is limited. If you do, replace the trimpot with a 2k one, and replace the "R1" resistor with a 5.6 kilohm one.

Alternatively, if you just want to tidy up the mod, you can remove the tacked-on parallel resistor and the 10k resistor underneath it, and replace them both with a 1.5k or 1.8k SMT resistor.
 

schmidtbag

Nov 8, 2012
36
Joined
Nov 8, 2012
Messages
36
As of right now this seems to be doing what I want it to do. Its powering a 24v motor so I don't think a few additional volts here and there are going to cause serious problems, especially if the motor won't chronically be pushed to full speed. If this consistently performs the way I'd like it to and works with another IC I intend to use later on, then I'll likely buy the 12v-35v version of this model (the one I originally intended to buy) for another motor.

If the booster's output current changes, that's fine. If it ends up dropping then that's actually preferred.


I have noticed that trimpots are meant for extra precision and I too thought about your comment about the accuracy, but what I'm working on here is for prototyping so I'm not terribly concerned about the end results, just as long as its working and not catching on fire (or about to).
 

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
Cool. Then the only issue is the reliability of your mod. You know what will happen if your extra resistor does become disconnected... the output voltage will shoot up to about 45V. If you're confident that it can't get disconnected or broken off if you bump the board, or with vibration, then that's great.
 

schmidtbag

Nov 8, 2012
36
Joined
Nov 8, 2012
Messages
36
Hahaha I'm not worried about that happening any time soon - after ensuring my solder job worked, I used a rubbery super glue and covered the solder joint of the resistor so when it dries, it'll be a solid mass and shouldn't break off, even if I drop it. I did leave some room to replace the resistor itself, if I need to.
 

CocaCola

Apr 7, 2012
3,635
Joined
Apr 7, 2012
Messages
3,635
I find hot glue encapsulation to be a great way to solidify parts on boards... It's overall permanent but with some orange (citrus) oil based degreaser and a little heat it will release cleanly if every needed... It's cheap and easy to apply as well...
 
Top