LOGIN
Gerry DeSeve
16
Mastermind
Inventor (Project Leader)
 · PhiladelphiaU.S.
Share
Report
Get Link
TASK
 Completed
Reward › Contribution award, recommendation letter and true gratitude!
Estimated Duration › 3 hours
Assigned to
Ravi Pujar
Sai Prasad

Power Supply Improvements

The Aurora power supply needs to be improved so it can handle a thermoelectric module drawing 2 Amps of power at 5 Volts. With the current Aurora design, the power supply overheats and fails by shorting the power and ground.

This is due to the power supply being too small to handle 2 Amps and to the placement of the power supply components. The supplier for the components, Texas Instruments, told my engineers that certain components had to be very close to the boost converter. We asked for a physical reason why but they couldn’t really provide one - they just said it was something that helped prevent shorts.

The current workaround is to by-pass the current power supply with an Adafruit PowerBoost 1000C board, which is expensive and isn’t practical long-term. (See https://www.adafruit.com/product/2465 for more on the 1000C, including the Eagle files.) The current power supply components should be replaced on the PCB with those from the 1000C - or something better - and the components should be positioned on the board so they do not short.

Also, the current battery connector has three pins for power, ground and a thermistor lead. The thermistor would be on the battery and is used to make sure the battery has not failed. Two pin connectors are more common, like a JST PHR-2. I’d like to get the team’s input to see if the third pin is really needed, or if a two pin setup would be better.

Abhishek Chakraborty has started working on this task.

Consumer Electronics
Electrical Engineering
Engineering & Tech
Hardware
Contributions
Ravi Pujar

Great tip on the fact that a component under consideration was buggy - a very important thing for the team to know.

3 likes 
Like
Award Contribution
Load 8 previous comments
Gerry DeSeve
16
Mastermind
Inventor (Project Leader)
 · PhiladelphiaU.S.
Share
Report
Get Link

Ravi Pujar

Here are the answers:

How much voltage do the Thermoelectric modules operate on?

Vmax for the target module, the TEC1-04902 (also known as a TES1-4902), is 5.78 V.

Advice from engineers has been to limit this to something less than the rated Vmax to prevent burning out the modules. So I think 5V is good as an actual Vmax ceiling, and typical operation should be 75% of Vmax (4.335 V) or less.

how much current they draw at different input voltages
and How much temperature difference can we achieve
with a given power source (VA)?

Unfortunately, I have not been able to find a professionally done datasheet with controlled test curves for the TEC1-04902.

But I do have three alternatives for you:

  1. Values can be estimated using the standard curves found in the document here: https://www.collaborizm.com/thread/SkewvpT_W

  2. Curves for the 6.3V / 4A version of the module can be found here: http://www.thermonamic.com.cn/TEC1-04904-English.pdf

  3. Bench test data for the TEC1-04902 when connected to a high precision variable DC power supply at 27.2C ambient, relative humidity of 50% and heat load of 32.8C (typical skin temperature) was:

Volts … Amps
5 … 1.458
4 … 1.219
3 … 0.933
2.5 … 0.796
2 … 0.635
1.5 … 0.476

I was not able to get a good temperature measurement during these tests.

Other info:

  1. To be effective, the system needs to create a skin temperature change of about 17.2C (from 32.8C to 15.6C).

  2. Here is a video of what an unloaded TEC1-04902 does at 5V constant voltage at 22.2C ambient and 42% RH over 28 seconds:

Starting temp is 36.7C and low temp is 11.3C, so total delta T is 25.4C.

Starting current was about 2A. At 11.3C current was 1.7A.

  1. A 1.1 CFM Sunon B0503AFB2-8(MS) was used as the blower in the video - see: http://www.hardwarespecialty.com/products/pdf/sunon/sunon_dc_brushless_blowers.pdf I mention this because the TEM’s Max delta T is limited by the amount of heat being ejected into the ambient air.

I will work on a block diagram and get it to you as soon as possible.

Thanks!

Gerry

Like
1 like 
Award Contribution
Gerry DeSeve
16
Mastermind
Inventor (Project Leader)
 · PhiladelphiaU.S.
Share
Report
Get Link

Ravi Pujar Here are the block diagrams - let me know if these are OK, or if I can provide any additional info/details.

Like
2 likes 
Award Contribution
Ravi Pujar
24
Community Expert
Embedded systems engineer, Founder @Valetron Systems, Blogger @www.raviyp.com
 · DharwadIndia
Share
Report
Get Link

Gerry DeSeve Thank you for the detailed block diagrams.
I can understand that we could use a 4100mAH as primary battery and provide charging option for it through Micro USB. But why the 10,000mAH extra battery is used? and also in a different path (external)?
I can suggest using a battery charger with power path like MP2617 which is capable of delivering current upto 2A. I use it in my GSM applications, but i have not tried continuous current as GSM current is pulsed and normal current is 50mA.
But i find this part much better compared to TI part BQ24266 which had severe heating issues. The MP2617 is able to handle this current very easily running at 12VDC input, BQ24266 used to get hot even at 5V input with a power pad as heatsink.

If we use a power bank type 10,000mAH battery at USB input then we should have to convert it to 5VDC to get the internal battery to charge. Is this where you were going to use 1000C board, to boost from 3.7V/4V to 5VDC ?

Like
0 like 
Award Contribution
Gerry DeSeve
16
Mastermind
Inventor (Project Leader)
 · PhiladelphiaU.S.
Share
Report
Get Link

Ravi Pujar Thank you for following up on this, and for your thoughtful questions ad advice.

But why the 10,000mAH extra battery is used? and also in a different path (external)?

The primary purpose of the extra 10,000 mAh battery is to provide a flexible way to extend the operating time of the system. The user can choose the size of the extra battery to suit their budget and how long they’d like the system to be operating.

In terms of why it is external, having a large primary/internal battery increases the weight, cost and size of the system. The path is really the same as the wall charging path.

Is this where you were going to use 1000C board, to boost from 3.7V/4V to 5VDC ?

Actually, it boosts the internal battery current from 3.7V to 5V. The 10000 mAh powerbank already delivers 5V through the USB.

One question for you - you mentioned that GSM current from the MP2617 is pulsed. By this, do you mean it uses PWM, or that the Voltage fluctuates frequently? Aurora does use PWM.

Like
0 like 
Award Contribution
Ravi Pujar
24
Community Expert
Embedded systems engineer, Founder @Valetron Systems, Blogger @www.raviyp.com
 · DharwadIndia
Share
Report
Get Link

The MP2617 uses PWM to generate the needed voltage of 4V for GSM as it is a switch mode battery charger. But what i mean by pulsed GSM current is, by nature GSM needs 2A bursts/pulses of current every few seconds during transmissions so the GSM module current draw is not constant or we can say it is pulsed.

Like
1 like 
Award Contribution
Gerry DeSeve
16
Mastermind
Inventor (Project Leader)
 · PhiladelphiaU.S.
Share
Report
Get Link

Got it - thanks for clarifying. I’m not familiar with GSM electronics, so this was very helpful.

Like
2 likes 
Award Contribution
Leave a reply...
DISCOVER
CHAT
HIRE
ACTIVITY
FEED
New Post
Help
Start Project
Online Users
Share Link
Write something before you submit it!
Photo updated
Request Sent!
Updated
Copied to Clipboard