AVR ISP MK2 + LM317 Regulator Tutorial

[thaking]

Hello all

I tried to make this modification on my ISP mkII, but have problem's

I use
-lm317
-2 x 0.1uF
-360 ohm
-220 ohm

I connected everything like dave in his video tutorial.

When I connected to USB, I get voltages:

5.1V and 2.83V  (switch on/off and 5/3.3 work OK )

I also try to read this voltages in AVR Studio 5(target voltages), which are the same.

But then I want to connect on chip, so I tried on xmega16a4. Before I connect I first try to connect via power supply to ensure that all work OK.    And it was, I can read xmega, set fuses....


Then I turn off power supply and try to connect ISP, with this modification. After I connect to chip and try to read, voltage drop to 1.5V. This also show in avr studio as target voltage(which we all know that it's too low)->1.5V

Any ideas how to solve my problem?

[thaking]

pictures...

[thaking]

Anyone? 

thanks

[ziq8tsi]

Slightly off topic, but related to Dave's earlier tips about using PCBs as front panels: I like how Atmel have used the high-contrast silkscreen on the underside of the board to show the logo through the transparent blue case.

[neoone]

First thing - you should take Dave's advice and use some fixed 3V3 LDO instead of LM317. But any way - did you completely disconnected any other power source from your uC ? Try to put some resistor like 100 ohms at the LM317 output (without uC) and check if the voltage drops. If it does then there's something wrong with your mod and if it doesn't change the resistor to smaller value like 33 ohms or even 10 (watch out for the power, if the resistor is only 1/4W do it quickly) and if it still does not drop significantly then the problem is on the uC side. for example current flows to the power supply circuit which is still connected but not powered from mains. Of course that's only one possibility. Also the voltage in "idle" is a bit strange - even in worst case with 10% tolerance resistors it should be greater than 2.9V.

[oPossum]

Looks like it is wired wrong.

[thaking]

Looks like it is wired wrong.

Yes first problem was that, I see that dave has turn around lm317, but I have in my mind lm7805 

Now worked fine on protoboard(I must change for LDO...as was said before)


Ok I will not include lm317 in my project, which 3V3 LDO do you recommend most for this

[jahonen]

Something like LM1117 might do the job, but it does not have very low dropout. It has the same pinout than LM317 that Dave used.

Regards,
Janne

[neoone]

whichever you can get cheaply just check first the data sheet for the capacitors since it's quite important thing for LDOs - when the caps are chosen wrong LDO can for example start to oscillate. As jahonen suggests LM1117 would do the job, it has 1.2V drop which should be OK.

[Zero999]

How much current is being drawn?

The LM317 is only guaranteed to regulate properly when the output is 3V less than the input but that's worst case at 1.5A and over the entire temperature range so you will be able to get away with less than that at lower output currents, still 1.7V is clearly not enough.

[thaking]

I will buy:


REG104-33
http://focus.ti.com/docs/prod/folders/print/reg104-33.html

Single Output LDO, 1.0A, Fixed(3.3V), Low Noise, Fast Transient Respons
Output Capacitor Type:   Capacitor Free     


I hope it will be ok, what you think?

[neoone]

well it is very good chip but did you check the price? I think it's over 10$ !! For one time thing you can go for it but it would hurt me if I where you

[oPossum]

LM1117T-3.3 or LD1117V33