Five different DC supplies for VDD and other bias voltages

[STOIKOV]

DC supply

I have some chips to test working at 900MHz but need five different DC
voltages for VDD and other bias voltages. What option would be better to generate these voltages :

- Make 5 voltages through a power supply with voltage divisors or
- Apply some voltages with batteries (maybe more stable voltages ?)

Also have heard that if your chip includes oscillators, it's good to bias it with a different power supply than the rest of the circuit for better stability purposes.

 

[Miguel Gaspar]

DC supply

One voltage power supply with a transformer with one primary and 5 secondaries.

 

[STOIKOV]

Re: DC supply

but if it is DC how to use transformer ??
what about independient batteries ? Could them supply stable voltage ?

 

[Miguel Gaspar]

DC supply

Well the switching power supplys first converts DC to a pulse train that can be transformed.

Batteries can supply stable voltage only when they are well carged and you are extracting very low power. Remember they tend to discharge and variations of voltage occours when they are discharged and bad funcioning happens.

 

[memarian]

Re: DC supply

Dear,

please list that 5 voltages.if i know i can help better.usually you can use dc-dc

converters,but where noise is important you can't use them.

Good Luck.

 

[STOIKOV]

Re: DC supply

1.56V fixed
2.56V fixed
3.3V fixed
5V fixed
0-5V variable

know voltage regulators that have precision and low noise ?

 

[szekit]

DC supply

make sure you put many supply decoupling caps on your supplies to get a clean supply voltage

 

[ehsanelahimirza]

Re: DC supply

hi

it is recommende that u use batteries.isolated and fully charged or batteries like automotive batteries are the best for these things

 

[memarian]

Re: DC supply

Dear,

From one supply voltage,for 3.3v & 5v fixed you can use fixed linear regulators

for best noise.for 1.56V & 2.56V you can use Variable Linear Reulators.

For 0-5V variable,no linear regulator can provide you 0v,they usually give you

minimum 1.2~1.25V.if it is enough you can use one of them.if your currents are below 100~200 ma i think you should use Low Dropout Regulators.With them you can use lower input voltage & you give Lower Power Dissipation.usually they have
SMD packages and this reduce your PCB Size.

For Better noise i think you should use Tantalum & Ceramic Capacitors with at least 2 different capacitance.for example a 10uf Tant+100nf Ceramic+100pf Ceramic.


Examples:

LM2931 Series : Low Droput Regulator up to 100ma,3.3V & 5V Fixed &3V-24V Variable,TO220,TO263,TO92 & MicroSMD Packages are available.

LM317 : Variable Regulator from 1.2V to more than 20V.in TO223 & TO252 Sourface mount packages.

LD1086 : from STmicroelectronics (SGS-Thomson).Three terminal adjustable or fixed output voltages 2.5V, 2.85V, 3.3V, 3.6V, 5V, 8V, 12V.Low Drop out at high current (1.3V @ 1.5A),SMD Packages Available.

LD2981: (SGS-Thomson) ULTRA LOW DROP VOLTAGE REGULATORS WITH INHIBIT.Output voltage of 2.85V, 3.0V, 3.2V, 3.3V, 3.8V, 4.85V, 5.0V.Ultra low dropout voltage (200mV typical at 100mA load and 7mV at 1mA load).Output low noise voltage 160 µV RMS.Smallest packages: SOT23-5L

if you use variable regulators,for better regulation i think you should use
NPO Resistors & NPO Multiturn Potentiometers.

Check the
SGS-Thomson : www.st.com
& Linear Technology www.linear.com
& national Semiconductors www.national.com
Sites for better selection.

Good Luck.

 

[STOIKOV]

Re: DC supply

thanks for the part numbers you've suggested. One problem is that the best performance devices (noise, etc.) are very small to handle them with conventional solder tools but anyway, I will select the best option for my application.
One more question what do you mean with NPO resistor, I found NPO is a kind of dielectric in capacitors but did not find this term related with resistors.

Regards,

 

[memarian]

Re: DC supply

Dear,

About NPO Resistor:

I only want to say that you should use resistors with low temprature coefficient.
(such as npo capacitors)

I think if you can use resistors with low temprature coefficient (about 10ppm) you

can have voltages with low temprature drift.

Good Luck.