Omega2-Dash 5V and 3.3V

David Kerr

Does the Dash run off 3.3V (if this is supplied to the header) or does the TFT require 5V?
Thanks,
Dave

David Kerr

Okay, I've answered my own question and provide further information which might be useful to some.

I've checked the schematics (kudos to Onion for publishing them). I can confirm the display runs off 3.3V.
5V from the USB in goes via the on/off switch to a DC-DC converter which powers the whole unit. The converter is rated at up to 1amp.
Current drain of the Dash is very modest, averaging 240mA at 3.3V or 792mW. Power drain is slightly higher at the 5V input but current draw is lower because the converter is quite efficient. Above power figures are with the Omega2S doing frequent network and screen activity. Of course, as with all the Omegas, there can be short current peaks which are higher.
I examined the schematic of the voltage converter and there should be no problem applying 3.3V (e.g. from another power source) to its output. The bottom FET of the totem pole output will be turned off. It is possible the top FET will turn on because of effective negative gate voltage. However, all this will do is partially connect the large (400+ uF) capacitor to the output, which is a good thing for handling short current peaks.
I have been successfully running one of my Dash units direct from 3.3V applied to the header. For over 24hrs without a glitch. I needed to use a 68uF capacitor across the header as otherwise long lead inductance to my power supply was a problem.
I take no responsibility for any similar experiments you might perform. Onion might even deem it a void of warranty but I hope not.

The big plus is that it makes it easier to build a hand held device that can be run and/or recharged by USB. I'm putting at least one Dash unit into a box with a 3.7V 18650 LiPo cell running it via a 1A 3.3V LDO regulator direct (via an/off switch on the box) to the Dash header. The 18650 will be supplied by a cheap Li battery charger with inbuilt low voltage cell protection.

I'll update the forum when I've built this "in the real world".

Regds,
Dave

György Farkas

@David-Kerr
Hi Dave,

Since the Omega2 DASH board already has a high efficiency step-down DC-DC voltage converter you can connect your lithium accu (nom 3.7V, max 4.2V) directly to the Expansion Header's 5V pin.
In this case the DASH board's Power Switch must be always OFF!

Regards,
gf.

David Kerr

No you can't.
Well, you can, but the specifications for the switching regulator show that it will or may drop out because it is not LDO. It's great for what it's doing (buck converter from USB 5v to 3.3).

Best (in this situation) is a LDO linear regulator. Indeed, if you look at the efficiency of the Dash switching regulator, overall efficiency is likely better with the 3940. All you lose is the very tail of the LiPo discharge curve.

I am using a LM3940 with a 120mV drop at the average drain of 240mA.
You can see photos and discussion on another thread "running Omega from 3.7V LiPo".
I've been running a Dash for several days now.
2600maH LiPo (actual 2400 because it is sitting at 15degC) is 9+hrs of running against a theoretical (no loss) scenario of 10hrs.

Cheers,
Dave

David Kerr

Upon rereading the spec, it DOES claim LDO but it does not give actual numbers. I only have two Dashs so I'll run some tests to see the actual drop out figures on these two units.

Stay tuned!

Regds,
Dave

David Kerr

Okay- I stand corrected. I used a precision power supply, adjustable to 0.01V steps with a precision voltmeter +- .0001V.

Feeding the 5V Dash regulator from the power supply, the drop out voltage for the switching regulator is 108.5mV for unit 1 and 101.5mV for the second.

The other thing is that the regulator does not "drop out" but reduces output maintaining the drop. So, the LiPo can go to 3.408.5V (unit 1) or 3.401.5mV for the other before the 3.3V output stops dropping. You can go down a further 100mV before the output drops to 3.2V where the Dash will continue to work just fine. So at 3.3V the LiPo is just about out of gas anyway.

In actuality, I was able to run each down to 2.9615V (at the 3.3V terminal) beyond which they stopped and indeed needed about 50mV more than that to boot successfully.

For reference, my LM3940 (only tested one) was 90mV, so a little better than the switching reg. However, efficiency of the switching regulator is higher with 15mA saved. This should give an extra half hour from my 2600 (actual 2400) LiPo compared with the linear regulator. Given that the switching regulator is already there, it is a no-brainer. So Gyorgy is totally correct. Use the 5V pin with the switch off.

However, if supplying an Omega2+ or Omega2S in a portable device, I'd use the linear regulator because of cost and component count. For the Dash, it's great to have the regulator there already.

Best Regards,
David

David Kerr

Here is a pic of one of my Omega2-Dash units connected to a Ti TP4056 based charger/protection board.

LiPo connected to +_ battery terminals.
Load out connected to Dash 5V/Gnd with 68uF tantalum. 5V switch on Dash set OFF

Charges via USB input on TP4056 board. Battery under voltage protection. TP4056 cuts charge when complete (PROVIDED the Dash does not stay powered up) else it "floats" which is bad for LiPo life.

I'm doing a test to see if I do indeed get the extra 30mins run time compared with the linear regulator (which gave over 9hrs).

Cheers,
Dave

David Kerr

The data is in!

The Dash, working with networking running (UDP packets averaging about 2/sec), ran for 10hrs 55mins!
Of my two Dash units, this was the one with the slightly greater (i.e. "worse") drop across the switching regulator.

That's using a 2600maHr LiPo which I measured at 2400maHrs (average temp 15deg hence the approx 10% degradation).
3.3V was maintained until 1hr before the end whereupon it slowly dropped until the Omega stopped working and the LiPo protection circuit had not yet kicked in.
Drop across the switching regulator was 120mV for the last hour. This was slightly higher than the 108.5mV I measured with the regulated power supply.

So a Dash with a better LiPo or with two LiPos gives (in my opinion) excellent battery life for mobile projects.

Best Regards,
Dave

crispyoz

If we had a sleep mode, the device would be so much more useful.

David Kerr

That is true.
However, for my (several) applications, I find it to be great. I'm sure that many others will find the same.

I have a number of low power applications using Omegas. However what I do is run an ultra low power PIC processor, which wakes up from time to time and fires up the Omega (and ancillary devices, in sequence) and communicates with them before then shutting them (and DC-DC inverters) down. They run with several high efficiency DC-DC converters and consume a tiny amount of power on a daily basis. I would like the Omegas to boot a bit faster, but overall this scheme works well. I cannot conceive the Omega having a sleep consumption of well under one microamp to match a PIC or similar microcontroller.

The Dash is the same philosophy as the 4D systems people have had for some years (smart displays). However, it is cheaper and more powerful and more general. The 4DGL units have more powerful graphics, some sleep options and you can programmatically turn off their display, but I like the fact that the Dash is the computer as well as the display.
Cheers,
Dave

crispyoz

@David-Kerr I have used a similar approach but while sleep mode would not provide the same battery life, it would extend it. My comments really relate to Omega2x generally not specifically the Dash.

I just want it ALL!