XS1-L4A-64-TQ48 Test board

XCore Project reviews, ideas, videos and proposals.
User avatar
Folknology
XCore Legend
Posts: 1274
Joined: Thu Dec 10, 2009 10:20 pm

Re: XS1-L4A-64-TQ48 Test board

Postby Folknology » Tue Jan 28, 2014 6:16 pm

I suspect with the circuit as you have drawn it, the 1V supply may have risen to 0.4v before the 3v3 is properly stabilised which would break the Xmos VDDCORE vs VDDIO sequence requirements.

Because the sc189c will trigger at 1.2V (as the 3v3 is still coming up) 100usecs later the 1v output will likely exceed 0.4v (see sc189c datasheet page 10).

regards
Al
User avatar
AtomSoft
XCore Addict
Posts: 135
Joined: Mon Dec 14, 2009 3:02 pm

Postby AtomSoft » Tue Jan 28, 2014 6:20 pm

The 1.2v is the threshold for a logic high to enable the device, the devis will not ouput poweruntil it passes the undervoltage lockout, or at least thats how iI see it.

That snippet on the EN pin states that it will switch off either when its below threshold (1.2v) or below lockout

hence
(ULOV = Lock Out Voltage)

EN 0 = < 1.2v
EN 1 = 1.2v+

LOV 0 = < 2.6v
LOV 1 = 2.6v+

POWER = Switched On
NO POWER = Switched OFF

ULOV or EN = OUTPUT
-------------------------
0 or 0 = NO POWER
0 or 1 = NO POWER
1 or 0 = NO POWER
1 or 1 = POWER
Last edited by AtomSoft on Tue Jan 28, 2014 6:30 pm, edited 2 times in total.
User avatar
mon2
XCore Legend
Posts: 1679
Joined: Thu Jun 10, 2010 11:43 am

Postby mon2 » Tue Jan 28, 2014 6:27 pm

The APX803 is not able to monitor 1.0 volts so a different component is required for the reset.

The SC189C requires a min of 2.9 volts on Vin to operate which will effectively allow your 3.3 volt input rail to stabilize before this switcher turns itself on to generate the 1.0 volt rail. This is the reason why I do not believe an external power supervisor is required for this device.

Please note that the Vdd on the TPS3808 device is a min of 1.8 volts. That is, you must connect the Vdd of this device to 3.3 volts (not the voltage you are attempting to monitor = 1.0 volts). Leave the SENS pin still connected to the 1.0 volt rail.

Add a pull-up resistor on the pin 1 of the TPS3808 since this device is also open drain so it needs a pull-up to +3.3 volts. 10k should be fine.

If you have the room, consider to apply a resistor on Ct of the TPS3808 so you can experiment with the RESET delay as per their documentation. See figure 12a on page 9 of their datasheet.

Where did you source the LPS3015 from ? Is this Coilcraft ? Good company but will be expensive and may not be easy to source (unless you get directly from their webstore). Perhaps consider some alternatives like Wurth or better yet, Panasonic ELL-VGG1R0N ? We are using this Panasonic shielded inductor with our SC183C design with success and is fairly low cost.

http://www.digikey.com/product-detail/e ... ND/1938506

Not sure on your part values or footprint for the filter bulk cap at the output of the SC189C but if you are following the guidelines by Semtech then you will be fine. We prefer non-polarized ceramics for a long life of our products (X5R or X7R dielectric).
User avatar
Folknology
XCore Legend
Posts: 1274
Joined: Thu Dec 10, 2009 10:20 pm

Postby Folknology » Tue Jan 28, 2014 6:29 pm

Ok Nice, I can see what your saying about the UVLO (2.6v worst case) assuming that the 3v3 ramp up isn't extraordinarily slow it should probably work and save you a chip!

regards
Al
User avatar
AtomSoft
XCore Addict
Posts: 135
Joined: Mon Dec 14, 2009 3:02 pm

Postby AtomSoft » Tue Jan 28, 2014 6:37 pm

VDD is now 3.3v
Sense is 1v
pullup on rst (10k)

I guess ill add a resistor to tinker with, good idea. (on CT)

I got the inductors straight from Coilcraft in a past order. I have about 8 left. So i didnt want them to go to waste. (or my money :D) But ill look into another inductor (especially the mentioned)

As for part values on the caps i usually follow the datasheet and forget to put it in the schematic :( bad habits i guess. So none polarized are better? Ill see what i can do since they are usually more expensive in the higher uF range.
User avatar
AtomSoft
XCore Addict
Posts: 135
Joined: Mon Dec 14, 2009 3:02 pm

Postby AtomSoft » Tue Jan 28, 2014 6:43 pm

I think you meant FIG 14(a) but then would i connect that pullup to my VDD or SENSE ?

LOL Scratch that... to VDD it says
User avatar
mon2
XCore Legend
Posts: 1679
Joined: Thu Jun 10, 2010 11:43 am

Postby mon2 » Tue Jan 28, 2014 6:52 pm

Actually I was reviewing the low q version of the TPS3808 which is then figure 12a but you are also correct if reviewing the standard TPS3808 datasheet = figure 14a. Both are drawn the same way.

On the caps, you should be fine to apply a 22 uf in a 1206 footprint but consider what works for you. Since the voltage rail is only 1.0 volts you should be able to source many caps to fit the footprint. We suggest you allow for the largest footprint you can afford due to the following article which is a real eye opener:

http://www.edn.com/design/analog/440204 ... -capacitor

Be sure to do a quick read. Regardless, we prefer the use of ceramics in our design vs. electrolytic or tantalums.
User avatar
AtomSoft
XCore Addict
Posts: 135
Joined: Mon Dec 14, 2009 3:02 pm

Postby AtomSoft » Tue Jan 28, 2014 7:06 pm

mon2 wrote:Actually I was reviewing the low q version of the TPS3808 which is then figure 12a but you are also correct if reviewing the standard TPS3808 datasheet = figure 14a. Both are drawn the same way.

On the caps, you should be fine to apply a 22 uf in a 1206 footprint but consider what works for you. Since the voltage rail is only 1.0 volts you should be able to source many caps to fit the footprint. We suggest you allow for the largest footprint you can afford due to the following article which is a real eye opener:

http://www.edn.com/design/analog/440204 ... -capacitor

Be sure to do a quick read. Regardless, we prefer the use of ceramics in our design vs. electrolytic or tantalums.
I am in no way an expert in cap selection. The datasheet recommends these:
http://www.digikey.com/product-detail/e ... -ND/587424

as a selection. As they are small i might just use them. They are smaller than the caps i intended to use and since they are recommended i assume best for the case.

Also these inductors:
http://www.digikey.com/product-detail/e ... ND/1893675

Ill use whats recommended to avoid any headaches...
User avatar
AtomSoft
XCore Addict
Posts: 135
Joined: Mon Dec 14, 2009 3:02 pm

Postby AtomSoft » Tue Jan 28, 2014 7:25 pm

Ok Updated...

No Tantalum caps anymore...
New Inductor
Added that Pullup on CT (i think i mentioned that before)
You do not have the required permissions to view the files attached to this post.
User avatar
mon2
XCore Legend
Posts: 1679
Joined: Thu Jun 10, 2010 11:43 am

Postby mon2 » Tue Jan 28, 2014 8:43 pm

On the TPS3808, remove the connection to #MR. #MR is an external source for a reset. You could technically wire in a pushbutton here I suppose but not sure if the device will handle the switch bounce. Likely the delay internal to the device would assist here if you do consider the switch.

Leave the #MR pin open since there is an internal 90k to Vdd (+3.3 volts for your circuit).

I have not reviewed the JTAG nor direction of your MISO / MOSI, etc. but can do this later on but looks pretty good otherwise.

Who is online

Users browsing this forum: No registered users and 1 guest