looking for new project ideas!

[minimer]

Hello!

I'm looking for a new project idea! I want to build something new and exciting! Preferably something more hardware intensive than software.
I would like a project that can leverage my existing knowledge Python, AVR and power system design.
I would also like this project to leave my comfort zone requiring investigation into things I have never given enough time like FPGAs  & HDL, ARM, ??
I have designed battery chargers, power supplies, dc to dc converters, and rc ESC's (3 phase motor drivers)
I have programmed everything from websites to image classifiers in python.
Finally I would prefer to build something you or some one you know could use!

cross platform GUI + power systems + new and exciting = new project!?

Nothing is too weird, most things are not too expensive! Any Ideas???

[Psi]

Do you want a project that will produce something sellable at the end, or just something fun/challenging to make?

[digsys]

Sure do. I have a project I've been trying to finish for the last 4 yrs, but never seem to have time. It could fit in well with what you did last,
PLUS it is a very interesting emerging technology.     Active cell balancing for EV Lithium packs.
Most packs have 36-72 series cells. I usually break this down into groups of 8 (as a replaceable module). The topography I've chosen,
(and there are 3 main ones), is - each group charges / discharges (up to 10A equalisation) from a COMMON 14.8V (4x Li cells) independent
storage block. If you have any interest, look up "active cell balancing" and I can provide more info. We can go from there.

[han]

i suggest :

Biosensors and Bioelectronics: (make sure it save, use battery powered device only)
1. EMG / EKG / ECG transducer;
- you can make it using instrumental amplifier and notch filter
2. Pulse oximetry
- using infrared led and photodetector

Optical:
1. Optical Specrum analyzer
-using grating (from CD or DVD if real grating is hard to get)
2. raman spectroscopy
- Using Laser pointe + OSA + notch Optical filter

[minimer]

Excellent!

Psi, yes I would love to create something I can sell who wouldn't! But that said it is a secondary concern of mine. But please don't hold out on any ideas!.


Digsys,
Well as luck would have it, long ago I wanted to build my own lipo pack for a vw bug electric conversion because lead acid is hopeless. And well at the time I realized I could build my own battery packs for a fraction of the cost of buying one. At that time I was pondering charging circuitry as well. I have actually already designed a balancing chargers and all sorts of things I never ended up building for high voltage lipo packs. Any way Active cell balancing to my knowledge is a circuit that uses the supply total to buck down a localized current source to add to a deficient cell instead of passively balancing the voltage of all the cells in a pack this system tends to be about 93 percent efficient at the buck stage and saves battery performance and longevity. I am interested in this design and we should start a separate topic about it or talk personally aaron.a.dennis@gmail.com.

han,
From my quick look at EMG and ECG I found simple cheap solutions to EMG I'm assuming ECG isn't vastly different. Pulse oximetry is also a fairly simple concept. So what do you propose simply building the sensors and signal conditioning circuits is not new... When i see this suggestion I see wireless EMG nodes and a master acquisition software for prosthesis research?? Tell me what your thoughts are this could turn into a multi project type of thing because distributed wireless EMG sounds excellent and so does active cell balancing.

[minimer]

P.S.  Digsys, han

Both of these idea interest me because they are easily scaled down I can test the active cell balancing on small cheap cells in large packs first. Then on a couple large batteries to verify current performance. Also developing a multi purpose EMG, ECG, pulse oximetry wireless node sounds like it is very useful to research that could make broken people whole again and can be tested with a few nodes to verify its design. 

[minimer]

EMG/ECG Looks like there already is a product for arduino the only improvement I could see immediately is adding something like the electric imp.
https://www.olimex.com/Products/Duino/Shields/SHIELD-EKG-EMG-PA/

[minimer]

Daisy chain capable active cell balancing IC! The typical application in this data sheet shows a 24 cell battery pack. Claims support up to 1000 volt battery pack.
http://cds.linear.com/docs/en/datasheet/33001fb.pdf

I'm thinking 6 or 12 cell nodes per PCB, just to keep PCB size down. Nodes can be added as needed in a given battery pack.

[rexxar]

Claims support up to 1000 volt battery pack.

Holy hell! What would you even do with a 1KV battery pack?

[minimer]

Daisy chain capable active cell balancing IC! The typical application in this data sheet shows a 24 cell battery pack. Claims support up to 1000 volt battery pack.
http://cds.linear.com/docs/en/datasheet/33001fb.pdf

I'm thinking 6 or 12 cell nodes per PCB, just to keep PCB size down. Nodes can be added as needed in a given battery pack.

Well I see, it looks like even with this IC the secondary windings must be interleaved every six cells to keep the voltage down. 
Sounds like we just need a high efficiency high current high voltage DC to DC converter! 

Example 36 cell system
6 LTC3300 IC's and 36 (1:1 10 A 10uH inductors) with required diodes, FETs and current shunts.
But rather than interleaved every 6 cells we will just buck from 36 cell potential to a 6 cell potential at max efficiency possible.
Draw back, 90% ish energy conversion in 36 to 6 cell potential conversion this only adds to the 90 % efficiency of the  active cell balancing.
So we will only accomplish about 80 percent total efficiency, but not bad for cutting a development and price curve.   

[minimer]

Claims support up to 1000 volt battery pack.

Holy hell! What would you even do with a 1KV battery pack?

Well, think about this! Watts = Volts x Amps however resistance to current flow creates heat. This is why power transmission lines are HV.
If you have a multi Kilowatt system would you rather have huge cables and waisted heat at every resistance above fractions of an ohm or 1 KV supply and low current??

[minimer]

Sure do. I have a project I've been trying to finish for the last 4 yrs, but never seem to have time. It could fit in well with what you did last,
PLUS it is a very interesting emerging technology.     Active cell balancing for EV Lithium packs.
Most packs have 36-72 series cells. I usually break this down into groups of 8 (as a replaceable module). The topography I've chosen,
(and there are 3 main ones), is - each group charges / discharges (up to 10A equalisation) from a COMMON 14.8V (4x Li cells) independent
storage block. If you have any interest, look up "active cell balancing" and I can provide more info. We can go from there.

Here is a picture to go along with my answer to this idea!

[Psi]

Make a LED matrix array on flex PCB that can slot into a pouch sewn to the inner side of any T-shirt.  Include a EMG/ECG sense circuit that displays the wearers heartbeat, or other stats, on the led grid (shining through the t-shirt fabric).

When its "off" the T-shirt looks normal. when its "on" it shows realtime stats in text or graph form.

Sadly its more of a MCU project which isnt really what you're after.
The mention of EMG/ECG made me think of it, so i thought i'd share.

[han]

hi minimer,

what type of area that you interest most?
maybe i can give some idea...