Acceptable practice to power memory ics from voltage divider ?

[iso9001]

Like subject says really, I'm seeing if its Good... erm, Acceptable Design to power two EEPROM ics from a voltage divider (knowcked down to 3V) off of my 5V regulator ?

I have two 8pin EEPROMs, one is 1.8-5V tolerant, the other has a maximum of 3V. If I split off a divider from my 5V regulator and these two chips are both writing at the same time I think the current should be around 20mA (writing both at the same time will never happen, but lets just say)


There shouldn't be a problem with this correct ? Maybe I should set it up for a zener instead ? I think my only concern is making sure the resistors that make up the divider are relatively of low. I could add a 3V regulator, but I'm trying to keep costs down, these roms will rarely be read from, probably never written to (accept for initial programming)

Just getting opinions, Thanks

 

[dknguyen]

Like subject says really, I'm seeing if its Good... erm, Acceptable Design to power two EEPROM ics from a voltage divider (knowcked down to 3V) off of my 5V regulator ?

I have two 8pin EEPROMs, one is 1.8-5V tolerant, the other has a maximum of 3V. If I split off a divider from my 5V regulator and these two chips are both writing at the same time I think the current should be around 20mA (writing both at the same time will never happen, but lets just say)


There shouldn't be a problem with this correct ? Maybe I should set it up for a zener instead ? I think my only concern is making sure the resistors that make up the divider are relatively of low. I could add a 3V regulator, but I'm trying to keep costs down, these roms will rarely be read from, probably never written to (accept for initial programming)

Just getting opinions, Thanks

It depends on how much current the ROM will draw and what voltage drop is can tolerate when it draws said current.

 

[blueroomelectronics]

Sounds like a lot of effort, why not just replace the EEPROM with a 5V one.
You need a regulator of some sort, the EEPROMs current varies with read an write modes and is just poor design practice.

 

[donniedj]

Sounds like a lot of effort, why not just replace the EEPROM with a 5V one.

But you have to remember, electronics is going forward with low voltage designs. Sticking to a 5v EEPROM will put a cap on not just future but everyday designs. The whole 1.8 to 5v tolerance is just a transition spec. One day 5v will be taboo. Today, I can not buy the PIC chips I want because they do not operated at 5v. The numbers are raising for 1.8 to 5v ICs that only cover the inputs/outputs and will not take 5v as the supply.

The only concrete issue with using a divider is power wasted. Today they are down to 3mA max making the dual use load to 6mA. Not much waste if fueled from mains powered, auto powered, or some other source of renewing energy. I say who cares. EEPROMS are not IGBTS.

 

[Papabravo]

Like subject says really, I'm seeing if its Good... erm, Acceptable Design to power two EEPROM ics from a voltage divider (knowcked down to 3V) off of my 5V regulator ?

I have two 8pin EEPROMs, one is 1.8-5V tolerant, the other has a maximum of 3V. If I split off a divider from my 5V regulator and these two chips are both writing at the same time I think the current should be around 20mA (writing both at the same time will never happen, but lets just say)


There shouldn't be a problem with this correct ? Maybe I should set it up for a zener instead ? I think my only concern is making sure the resistors that make up the divider are relatively of low. I could add a 3V regulator, but I'm trying to keep costs down, these roms will rarely be read from, probably never written to (accept for initial programming)

Just getting opinions, Thanks

If you worked for me and I saw this in one of your designs during a design review I'd have a long talk with you about why you think it is a good idea. I'd be surprised if you could justify it. A voltage divider is quite simply inappropriate for any application that will draw enough current to change the voltage. This is not just a room temperature proposition; I would want to know what happens over extremes of supply voltage and temperature. I would want to know if there are issues of memory integrity during power up and power down. Abandon this idea sooner rather than later. Put in a second regulator and you will find all sorts of benefits as designs migrate to lower voltages.

 

[donniedj]

I would first ask how much current is being drawn and if this is a mission critical design or some back room hobby board that most likely is riddled with for more errors other than the believed to be taboo divider.

Why would a person working for you need to even ask basic questions in the first place. I bet you would fire someone because they spend the money on the extra regulator.

"If we have time to do it over, will it be done right or will we repeat this act for all eternity since we never have time to do it right? Or we doing it over because the act of doing it over is not equivalent to doing it right? Are repetition and correction even related? If we do it right after doing it over, does this not prove there is time enough at last." - by ACG, by Me

 

[blueroomelectronics]

The OP asked if it was "Acceptable Design to power two EEPROM ics from a voltage divider" the responsible answer is no.
The IC is going to change its load unpredictably and its also really poor practice. Just like so many folks don't use decoupling caps for whatever reason. 3.3V regulators are cheap, use em.

 

[Pommie]

Why not just replace the bottom resistor in you divider with a 3.1V zener. Set the top resistor to drop 2v at whatever the max current is and you should be set. You should include decouplers as well.

Mike.

 

[Papabravo]

I would first ask how much current is being drawn and if this is a mission critical design or some back room hobby board that most likely is riddled with for more errors other than the believed to be taboo divider.

Why would a person working for you need to even ask basic questions in the first place. I bet you would fire someone because they spend the money on the extra regulator.

"If we have time to do it over, will it be done right or will we repeat this act for all eternity since we never have time to do it right? Or we doing it over because the act of doing it over is not equivalent to doing it right? Are repetition and correction even related? If we do it right after doing it over, does this not prove there is time enough at last." - by ACG, by Me

I would never fire someone for something like that, but I would try to help him find a better solution. The cheapest cost solution in parts may in fact lead to years worth of support headaches. It is simply not worth the aggrivation to cut corners like that. Three terminal regulator are cheap cheap cheap. The next question I might be asking is "would there be a benefit to switching other parts to a lower voltage". The 800 pound gorilla in the room is that many 5V parts are now or soon will be going "End of Life". Can you imagine the pressure when production has to shut down because there are literally "no more parts". Then purchasing comes up with a batch from a "broker". You put them on a board only to discover that they are empty cases with the appropriate markings. Boy does that go over well with management!

 

[Pommie]

So, would I get talked to for suggesting a zener for regulation, assuming no power restrictions.

Mike.

 

[picbits]

Why not just replace the bottom resistor in you divider with a 3.1V zener. Set the top resistor to drop 2v at whatever the max current is and you should be set. You should include decouplers as well.

Mike.

That was going to be my suggestion for a cheap and nasty fix.

For the sake of a few pence I'd go with a dedicated 3.3v regulator

 

[Papabravo]

So, would I get talked to for suggesting a zener for regulation, assuming no power restrictions.

Mike.

Well I've never been tempted to use one for that application, and I'd still try to convince you it was a bad idea from a system point of view especially if there was a motivation to migrate other parts to a lower voltage.

 

[phil7890]

My 2 cents for the OP...

Scenario 1... it's a high-volume kid's toy and the useful life of this product is a few minutes/hours to a few days, ie. a disposable product,... two resistors is just fine. Don't need to decouple it.

Scenario 2... it's a kid's favorite toy and the useful life of this product is a few days to a weeks... resistor and zener diode will do. Have to decouple it too.

Scenario 3... it's a commercial product or something people will use for months or years. A separate voltage regulator should be used here (or select a different chip that doesn't require any voltage conversion). Certainly have to decouple it too.

Scenario 4... it's an industrial product. Other than a separate voltage regulator, should have a polymer/tantalum cap too (or select a different chip that doesn't require any voltage conversion). Certainly have to decouple it too.

Scenario 5... it's medical grade. Remove some caps... low leakage