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.

Digital Potentiometer and Power Supply

Status
Not open for further replies.
Are woried about human safety?
Yes.
I had thought about putting the battery on the inside of the vest. Battery loss will heat the human. :) The battery will be worm and work better than if the battery was at -40C, out side temperature. But batteries burn up some times. Do you want a battery pushed against your skin. On the news they showed some pictures of E-cigarettes catching on fire in a picket. :eek:
----------edited-------
How is the computer being powered? 5V?
If this is just a prototype then it does not matter. Use a wall wort and don't worry about it. But you need to power the computer and 3V is low.
 
Last edited:
40 yrs ago , I did the same with HP power supplies and used a DAC R ladder to control the level. In my case for immunity, I used isolated mercury wetted relays. Now there are better options. The digital pot has more resolution but limited range.

Normally a constant low current drives the pot to generate the control voltage.. But dont guess, test it

A suitable smoothing cap prevents glitches.. Twisted pair or short leads is important to reduce ingress.
 
Yes.
I had thought about putting the battery on the inside of the vest. Battery loss will heat the human. :) The battery will be worm and work better than if the battery was at -40C, out side temperature. But batteries burn up some times. Do you want a battery pushed against your skin. On the news they showed some pictures of E-cigarettes catching on fire in a picket. :eek:
It would not be advisable to put the battery pack in contact with the human body for this exercise for three reasons:
(1) Uncomfortable mechanically
(2) Uncomfortable thermally
(3) Safety hazard. But don't forget that there are millions of people walking around with LiIon batteries in their pockets (puffers), not to mention mobile phones, and cameras, and for a production item, which this is not, there are fairly standard mechanical and electrical measures to mitigate any risk.

But the idea of putting the battery in thermal contact with the body has been used, especially by the military. Self heaters have also been used. As you say, the battery will internally self-heat anyway, but internal self-heating will be inversely proportional to the number of batteries used.

How is the computer being powered?
First thought was to power the MCU from the 3V to 4.1V (3.6V nominal) battery pack, the same as the heaters. I can't see a problem with this as far as the MCU is concerned but, as I said, the problem is driving the three MOSFETs with just 3V, but there are simple ways around that.

In this application we may even have the luxury of using low side pulse width modulation (PWM), so NMOSFETs, which have a better performance than PMOSFETs, could be used.

As you imply, it may be better to power the MCU from 5V, in which case a simple low power converter could do the job, possibly charge pump, or even a couple of CR3032 lithium coin cells in series to run the MCU at 6V.

As the gate charge of the switching MOSFETs is liable to be high, to get a low RDss with a low voltage gate drive, gate drivers might be needed to turn the MOSFETs on and off fast enough. But on the other hand, heaters would probably operate OK with very low speed PWM, say 10Hz.

Some more material here which should impress the OP's professor. :happy:

spec
 
Last edited:
40 yrs ago , I did the same with HP power supplies and used a DAC R ladder to control the level. In my case for immunity, I used isolated mercury wetted relays. Now there are better options. The digital pot has more resolution but limited range.

Normally a constant low current drives the pot to generate the control voltage.. But dont guess, test it

A suitable smoothing cap prevents glitches.. Twisted pair or short leads is important to reduce ingress.
Yeah, you could make the e-resitor control a constant current to give a linear relationship of R_total versus output voltage but, in this case, the MCU could sort all that out without any added hardware.

spec
 
Hi, I have read all your comments and I find them very helpful. Thanks a lot, I appreciate it!

I was brainstorming regarding the setup for the elements: 3 groups in parallel, and each group has elements in series (pictured attached). I thought this was a good idea so if a group fails, others can continue functioning. What do you guys think?

Also, how would the set up of the batteries affect the circuit? From what it is understood, batteries in series increase voltage and in parallel increases current. Would a having batteries in parallel and in series work too?

Regarding MOSFETs, I completely agree with you. MOSFETs will be implemented in design. How to know which MOSFET would work best? I am worried they might get too hot for safety.

Thanks a lot. You guys have been great. Your input is valued!
 

Attachments

  • cirt1.PNG
    cirt1.PNG
    22.3 KB · Views: 137
Hi, I have read all your comments and I find them very helpful. Thanks a lot, I appreciate it!

I was brainstorming regarding the setup for the elements: 3 groups in parallel, and each group has elements in series (pictured attached). I thought this was a good idea so if a group fails, others can continue functioning. What do you guys think?

Also, how would the set up of the batteries affect the circuit? From what it is understood, batteries in series increase voltage and in parallel increases current. Would a having batteries in parallel and in series work too?

Regarding MOSFETs, I completely agree with you. MOSFETs will be implemented in design. How to know which MOSFET would work best? I am worried they might get too hot for safety.

Thanks a lot. You guys have been great. Your input is valued!

Yes, you can put batteries in parallel to increase current capability and you can put batteries in series to increase voltage. Both arrangements are quite normal. In both cases the energy, in Joules, would be doubled.

Take a 3.6V 3 Amp/Hour (38.88 Kilo Joules of energy) LiIon battery: two in parallel would give 3.6V at 6 Amp/Hours = 77.76 Kilo Joules while two in series would give 7.2V at 3 Amp/hours = 77.76 Kilo Joules, exactly the same. Note that a Joule is one volt at one amp for one second so you have to multiply by 60 seconds * 60 minutes = 3600, to cope with hours (Battery Voltage * Amp/Hours * 3600= Joules of energy stored in a battery).

It is the energy stored in a battery, in Joules, that defines the battery duration in a particular application, just like the gallons of gasoline in your automobile tank defines how far you can drive. If you use the energy 100% efficiently in both cases, the battery duration would be exactly the same for parallel and serial batteries.

You would specify a MOSFET that can be turned on with 3V and can handle at least 5A and have a low RDsss (on resistance). I can specify a suitable MOSFET if you would like.

Don't worry, a correctly specified MOSFETs would not get very hot. An effective RDss (on resistance) of around 0.100 Ohms would be easily achievable giving a MOSFET power dissipation of 0.100 Ohms * 4.5 Amps = 0.45 Watts, which is nothing. The actual dissipation in a MOSFET using pulse width modulation (PWM) heater temperature control is more complex than this but this dissipation figure will not be far off.

Heating elements tend to fail open circuit rather than short circuit which is another advantage of wiring the heaters in parallel rather than series. In a series arrangement if one heating element went open circuit it would take out a whole heating bank, but in a parallel arrangement only the open circuit heater would be lost*. Also, with a parallel heater arrangement it is simple to add or remove heating elements and batteries*. Finally, the resistance of heating elements is not normally well defined so the parallel arrangement would give optimum heat output from each heating element*.

The first * point is derived from Failure Mode Analysis (FMA) and the second two * points are connected with modularity, one of the fundamental tenants of good design. Theses two terms would also impress your professor. :)

spec

PS: I see you have started a new thread on this topic:
https://www.electro-tech-online.com/threads/heating-jacket-controlled-by-arduino-for-winter.148520/
 
Last edited:
I can specify a suitable MOSFET if you would like.

Hi Spec, thanks a lot for the useful info. I will make sure I put it into the information of the project. Unfortunately nobody advised me regarding a MOSFET so I bought the IRFZ44NPBF MOSFET N-channel. Do you think that was a good purchase? If you have a better suggestion please let me know, and yes another thread so I can have more expert advice :)

Thank you
 
Here is the schematic for the circuit that I propose:

spec

Issue 7 of 2106_07_18
2016_07_12_iss2_ETO_BODY_HEATER_CONTROLLER_VER1.png

NOTES
(1) All capacitors are through hole (not surface mount) ceramic X7R dielectric, +-10% or better.
(2) All resistors are quarter Watt (250mW) minimum, metal film through hole, +-5% or better.
(3) The circuit should be connected up as indicated on the schematic.
(4) Thick traces on the schematic indicate substantial wire/PCCT traces to handle high current.
(4) Batteries are LiIon size 18650: LG INR1865-HG2 or Samsung INR18650-30Q. No other batteries will do (unless the relevant parameters are checked on the data sheet). Do not be tempted to buy batteries from any but a dead reliable source, never mind how low the price or how good the paper specifications are. I am advised that this is a reliable vendor with good prices: https://www.fasttech.com/category/1420/batteries
(5) R6, R7, R8 are gate stoppers and must be connected physically to the PMOSFET gates.
(6) Only the PMOSFETs specified will work in this circuit (unless the data sheet parameters are checked for other PMOSFETs).
(7) You can put as many batteries in parallel as you like to extend the battery duration but keep all the batteries the same type.
(8) The lowest frequency pulse width modulation (PWM) should be used to control the heater power dissipation: 2Hz is probably about right.
(9) Pins 5, 6, 7, & 8 (drain) of each MOSFET case should be connected together. Likewise, pins 1, 2, and 3 (source) should be connected together.
(10) Heater temperature sensing, as I mentioned in an earlier post, is not included but can be added at a later date.

DATA SHEETS
(1) PMOSFET
https://www.digikey.co.uk/product-d...age/TPH1R712MD,L1Q/TPH1R712MDL1QCT-ND/5323104

 
Last edited:
(1) All capacitors are through hole (not surface mount) ceramic X7R dielectric

Sorry for the late response, I did not get notified in my email of your notification. Thanks a lot for your help. Everything looks great. I bought the batteries already and I am almost finishing the microcontroller code. I am pretty excited. One quick question, are we using the capacitors as protection?

Thanks you
 
It will work if only one pin is connected. Because inside the part all the Source pins are connected together.
The part is rated for 200A pk (short time) and a single pin con not do that. Normally you connect to all the pins to keep the current ability high and the resistance low.
Pins 5----8 connect to a large piece of copper on the PCB to be a heat sink.
upload_2016-7-17_18-3-56.png

Look at the bottom of the part you will see that pins 5,6,7,8 are all connected to a piece of copper.
upload_2016-7-17_18-5-55.png
 
One quick question, are we using the capacitors as protection?

Hy cmore82,

(afraid that the alerts on ETO are inconsistent)

It will be interesting to know how you project develops.:cool:

The capacitors will give protection against transient voltages and currents but their primary function is to provide a low impedance (like low resistance) for the PMOSFETS and microcontroller. Because the PMOSFETs are conducting a high current they have relatively large decoupling capacitors while the MCU only consumes a small amount of current so the MCU only has a relatively small value decoupling capacitor. To be effective the decoupling capacitors need to be as physically close as possible to the pins on the device and all wires/traces need to be short and substantial.

Think of the capacitors as little local batteries- they store energy locally and supply energy at a low impedance. Ceramic capacitors, unlike most other capacitors, work from DC to high frequencies; that is why they are specified (but ceramic capacitors, especially small surface mount types, have some weird characteristics so you need to take account of this when designing ceramic capacitors in).

A practical circuit is not the same as a schematic because wires, PCB traces, and solder joints add resistance, inductance, and capacitance, which can form circuits that you are not aware of and do not want. This added resistance can also reduce efficiency, by wasting power, especially in a power circuit like this. Not only is the schematic not the same as a practical circuit, but the components themselves are not perfect either. For example, resistors have parasitic capacitance and inductance as well as resistance.

One of the fundamental problems with anything in the world is stability. This includes physical structures like automobiles, bridges, buildings, and trees, but it is exactly the same in electronics. By the way, everything in the world has a resonant frequency; even simple resistors resonate around 200MHz (as a consequence of their parasitic capacitance and inductance).

MOSFETs, in particular, are prone to oscillate, typically between 2Mz and 10Mz. This is because they have huge parasitic capacitors formed internally between their terminals and they also have a very high frequency capability. Ordinary bipolar junction transistors (BJTs) are not so twitchy.

The resistors connected directly to the gates of the MOSFETs are called gate stoppers, which reduce the MOSFETs gain at high frequencies and thus prevent the MOSFETs from oscillating. The gate stoppers, in this circuit, also improve the waveform at the gates of the MOSFETs when the MOSFETs are switched. Although you will only be switching the MOSFETs at a low frequency, the edges of the switching waveform from the microcontroller are very fast and thus contain high frequencies.

spec
 
Last edited:
MOSFETs

spec

Hi Spec, I have been reading about decoupling capacitors. Thanks a lot for the information. Just to make sure, in order for our chosen MOSFET to be active a voltage of 4.5 V should be supplied to its gate correct?

Thanks
 
Just to make sure, in order for our chosen MOSFET to be active a voltage of 4.5 V should be supplied to its gate correct?

Hy cmore82,

In engineering it greatly simplifies things and avoids confusion if you are specific (not aimed at you- just a general point).

(1) There are two sexes of MOSFETs: N chanel MOSFETs (NMOSFETs) and P channel MOSFETs (PMOSFETs).
(2) NMOSFETs and PMOSFETs are exactly the same except they operate with opposite voltages.
(3) With an NMOSFET the drain must be more positive than the source. Under that condition if you make the gate the same voltage as the drain no current will flow from the drain to the source. If you make the gate more positive than the source, current will flow from the drain to the source. Note that the gate is insulated from the drain and source, so the gate does not take any current.
(4) With a PMOSFET the drain must be more negative than the source. Under that condition if you make the gate the same voltage as the drain no current will flow from the drain to the source. If you make the gate more negative than the drain, current will flow from the source to the drain. Note that the gate is insulated from the drain and source, so the gate does not take any current.
(5) Your circuit has a PMOSFET which, when turned on passes current from the positive supply line into the heaters.
(6) The voltage needed to turn a MOSFET on (gate threshold voltage) varies between MOSFETS and ranges from 12V to as low as 200mV. I chose a PMOSFET for your circuit that has a gate threshold voltage between 0.5 Volts (500mV) and 1.6V. Notice that a PMOSFET that had a gate threshold voltage of 12V would be totally useless because the supply line will only be 3V to 4V from a LiIon battery thus the MOSFET would never turn on.
(7) Assume that the battery voltage is 3.6V, which is the normal voltage for a LiIon battery. The PMOSFET source will be at 3.6V and the PMOSFET drain is connected to zero volts via a heater bank.
(8) If the MCU output is high (3.6V) there will be no difference between the PMOSFET gate and drain, thus the PMOSFET will be turned off and no current will flow from the PMOSFET source to the PMOSFET drain so no current will flow into the heater bank.
(9) If the MCU output is low (0V) the PMOSFET gate will be 3.6V more negative than the PMOSFET source, so the PMOSFET will be turned hard on and current will flow from the PMOSFET drain to the PMOSFET source and down to 0V through the heater bank.
(10) The PMOSFET has been chosen to have a very low resistance (RDS) of 0.002 Ohms (2 mili Ohms) between its source and drain when it is turned on, so that:
(10.1) the power dissipated in the PMOSFET will be low, so the PMOSFET will not exceed its maximum internal temperature of 150 deg Centigrade. Power dissipated = I*I* RDS. In this case the average power dissipated in the PMOSFET will be 4.5A * 4.5A * 0.002 Ohms = 0.0405 Watts (40.5mili W).
(10.2) as the power lost in the PMOSFET is low little power will be wasted, thus battery life will be maximized.
(11) Note that although MOSFETs gate are insulated from the source and drain, there are huge parasitic capacitors inside MOSFETs. As a result you need substantial current to switch MOSFETs on and off fast.

spec
 
Last edited:
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top