Translation from 5V to 3V

[hltc]

This is just going out for opinions. The cheap and dirty way is the simple 2/3 voltage divider, and it works...sorta. For slow rise time signals, I don't see much of a problem but for faster signals, I'm see the results of the capacitance on the cable, etc that gives me the slow fall and rise times so my nice square wave now looks like...crap.

So, I've played some with resistor values but I haven't seen anything that I'm really happy with.

This is a bi-directional bus..driving an LCD module at 3V, so there are some reads going on.

Suggestions on what might be the better way...ie...what translation chips have others used to interface to 3V logic.

 

[geekEE]

What about using 3.3V logic that is 5V-tolerant? 74LCX or 74LVC would work.

Glenn

 

[geekEE]

I should have mentioned that the 5V-tolerant methods works as long as the 5V signals have TTL levels rather than CMOS levels.

Glenn

 

[itsmoked]

If it's the standard LCD modules just delay the strobe long enough to get the job done. You rarely need to talk to displays very quickly as they are poky slow devices anyway.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[MacGyverS2000]

TI has a fairly wide selection of translation chips... they tend to be a bit more expensive on average, about 10-20% more, but I often don't need more than a couple here and there for one offs, so the extra cost is a good trade-off for the selection.

Dan - Owner

http://www.Hi-TecDesigns.com

 

[hltc]

I was thinking that something other than voltage translation chips might work. Keith..I don't have access to the firmware on this, its an upgrade to an existing product and the code is another messy story that I am working on as well, or at least trying to get the original source (I came in late on the project). I'll see where this takes me, ought to be able to find something. I did notice the read line is being toggled, and not sure if thats just part of the opcodes to control or if the data bus is truly bidirectional - need to locate the controller data sheet to verify that as well.

Again, thanks for the suggestions.

 

[itsmoked]

I see...

Typically the reads are just checking to see if the last command has finished executing in the LCD.

What values are you using for your dividers? What processor are you using?

Keith Cress
kcress -

http://www.flaminsystems.com

 

[Noway2]

I have used the 74VLC4245A for several years now with great success. The devices are bidirectional buffers that take two supply voltages (your 3V and your 5V supplies). You can either use one device and switch the direction control or use two devices and have one in each direction. Last I checked, the cost about $0.50 from digikey.

These devices convert 8 signals at a time. If you are interested in just one signal, there are devices such as the SN74LVC1G14 or similar ( that one is an inverter too) that will do the trick.

If you are really cheap, you could use a simple transistor, such as a logic N-Chan mosfet. Use the gate to pick up the 3V signal and switch 5V with the source-drain.

 

[hltc]

I've ordered samples from my TI rep, they will be here tomorrow. Thanks for the suggestions.

 

[xray]

hltc,

I just finished a little FPGA driving a Noritake vacuum display. I ordered an analog part lpc 32 pin package?, quite small and hard to work with utlizing a soldering iron. I was looking around in my supply as found a TI74ach273 and worked great. Put a quick design on eagle produced some gcode and sent it to the shop bench pcb printer, taig cnc. a couple hours and shazam an FPGA voltage translator for a noritake vacuum display. a couple of TI74ACH273 (16 I/Os) and that TAIG CNC sure is a delight.

Regards,

Cy

 

[itsmoked]

Sounds like fun! Too bad those Noritakes burn out in just a few years. Every appliance I have that has a display has those displays and they are all screwed up now. I have a VFD with a backlit LCD that has been on for 14 years straight. It has not changed in any visible way.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[Noway2]

One of the things that I recall the Noritake rep recommending is to use a form of screen saver as they admit that they will loose about 50% of the brightness when left running continuously for a few years.

Depending on the application, this may or may not be feasible. In the last application where I planned one of these displays, the machines usually sat in a basement and got ignored except for when they reported trouble or were getting their monthly test. During the periods of inactivity, we reduced the display brightness to about 25% and bounced a "press any key" message around the screen, greatly reducing the drain on the phosphor elements.

 

[xray]

hltc,

Your handle almost sounds like the right logic choice, but have a look at this logic and part: SN74LVC4245A. My project went the other way cmos to ttl this one should go from ttl to cmos 5 to 3. There is so much logic on TI, but if this does not work look on TI site it has what you need.

regards,

Cy