Voltage Divider Help

Status
Not open for further replies.

AtomSoft

Well-Known Member
Ok i know i ask a ton about this kind of stuff but this is important.

I now use ARMs (pics too) but this is a more ARM power question

The core uses 1.8v and the vcc is 3.3 v

So i usually have 2 regulators on board ... A 3.3v reg and a ADJ one or a 1.8v reg...

If i use 1/8 watt resistors how much current can i draw without frying the resistor? or making it too hot?

I need to know what wattage to use.

Thanks!

Or should i simply stick with the 1.8v Reg and 3.3v reg pair?
 
Last edited:
1/8W = V*I, so if for example, you're dropping 3.3V - 1.8V, you can do a little algebra:

1/8W = (3.3 - 1.8)*I

1/8W = 1.5 * I

=> I = 1/8 * 1/1.5 Amps

I = .083A or 83mA.
 
awesome thanks so doing that and using 1/4w wouls be

1/4 * 1/1.5 = 167mA (rounded up)

Awesome. That way more than enough.

So i uploaded a picture of the info it stated 70mA max



you think a 10mA cushion is fine?
 

Attachments

  • max.jpg
    52.4 KB · Views: 466
Last edited:
Last edited:
At 70mA, a voltage divider would make a poor regulator. You can use the zener, but you should buffer it with an opamp, or else just use a 1.8V regulator. Otherwise, you can use a zener and resistor voltage divider, but make sure you have some spare current delivered from the resistor ( ie, (5 - 1.8)/R > 70mA ), and that no max power spec is violated.
 
3.3 - 1.8 = 1.5

1.5 / 20 = 75mA
1.5 / 15 = 100mA

So a value between 15-20ohm is best correct?
 
Regulators are cheap. Your time spent debugging flaky circuits is not. I would stick with off the shelf regulators.
 
was just a thought heh.. I just got some 1.8v regulators in ... but wanted to check other ways.
 
3.3 - 1.8 = 1.5

1.5 / 20 = 75mA
1.5 / 15 = 100mA

So a value between 15-20ohm is best correct?

Careful, 1.5V*100mA = 150mW, so use 1/4 watt resistors, at least. Also, the power in your zener will be 1.8V * (100mA - 70mA) = 54mW average. If the uC were to become disconnected, then you have, for the zener 100mA*1.8V = 180mW. Always make sure you're spec'd for enough power.

It's always good to analyze options.
 
Last edited:
What ARM7 are you using? The ones I used all had a 1.8v reg built in for the core, though you still had to externally route it.

Really, it's time to move to Cortex chips. Hardware is easier, firmware is simpler. ARM7 is old news.
 
Last edited:
What if you dropped 1.4volts from your 3.3 volt supply with two diodes in seriese to get 1.9volts? Andy
 
Im playing with the LPC210x (2101 and 2103 to be exact) From what i know they require a 1.8v input and dont have it internally generated.

I heard about the Cortex Stuff. But it seems to have so many variates... Can you recommend something? Can i still program via Wiggler Clone?
 
I've gone through a few manufacturers, but I'm basically sticking with the LPC1xxx chips. Mostly the LPC1313 (Cortex-M3) and LPC1114 (Cortex-M0). Unfortunately you need to go up to the LPC17xx devices to be able to use regular JTAG. I don't know what there is for SWD options out there right now, but I'm using an SWD adapter from Rowley on my Olimex AMR-USB-TINY JTAG. I think finding a decent SWD programmer will be the hardest.

The M3 and M0 parts are nice. Fast, 100Mhz, 72Mhz, and 50Mhz. Cheap for the LPC13xx and LPC11xx parts, down to $2 for 8k LPC11xx parts from NA suppliers. The LPC17xx are the 100Mhz do it all parts with all the crazy peripherals like Ethernet and they get close to ARM7 prices $6 to $12 for NA distributors.

Simpler to program, the peripherals, at least on the SAM7 chips I used were crazy. My SPI library was nutty. The cortex chips are meant to be simpler and the LPC13xx CM3 and LPC11xx CM0 have compatible peripherals mostly, so I don't have different libraries for both.

Simpler hardware. Just power and a crystal oscillator. No external PLL or need to power the VCORE yourself. I can actually put one on a single sided board with a minimum of jumpers. The SWD connector is a much smaller connector, though I think I'll be moving from the standard header to a 4 pin header. The little SWD connector are expensive and I really don't need them.
 

Attachments

  • 20100308-lpc1114-breadboard-002.jpg
    197.2 KB · Views: 264
Last edited:
Cool! I paid $2.50 for LPC2101 which is cheap to me! If i had more info on this SWD programmer i wouldnt mind taking a look into those devices. I made my Wiggler Clone with free items so it didnt cost me anything really. Only $2 for a PCB i used 2x2in of so i could make about 6 with 1 pcb heh

If you can find me info on the SWD please or someone with more info heh..

Thanks guys
 
Looks like the Segger EDU might be the best choice.
**broken link removed**

Something might have come up in the last few months, though. I think the OpenOCD guys have been doing some things and might have broken some proprietary SWD debuggers like the LPCXpresso. My Crossworks adapter wouldn't be an option. I don't think it works with the wiggler (not really sure).
 
Hey Mark, me and Paul Curtis from RA (Rowley Associates) came to a agreement on a argument we had on the LPC2000 Yahoo group forum. I now am the owner of :

Crossworks for ARM (v2)
CrossConnect Classic For ARM
ARM SWD Adapter

Thats $515 USD for FREE. He sent me it and it just arrived! I feel so excited heh... I now can play with many ARMs since i have the tools.

Ill be looking into the LPC1313 and LPC1114 ... Thanks for the tips!

https://atomsoft.wordpress.com/2010/11/20/ra-tools-just-in/
 
Last edited:
I was trying to source connectors for that small 2x5 connector and suck!!! i will just use:
80-pin 0.05? PCB Male Header - dipmicro electronics

I might just make a breakout like convert the 2x5 (1.27mm(0.05in)) to 2x5 (2.54mm(0.1in)) This way i can simply use the common sized headers.
I think though hole is better though since its stronger. Since Dipmicro sells those so cheap i will get a few and test them out.
 
Status
Not open for further replies.
Cookies are required to use this site. You must accept them to continue using the site. Learn more…