Author Topic: Voltage Memory circuit  (Read 4718 times)

0 Members and 1 Guest are viewing this topic.

Offline FTVlogTopic starter

  • Newbie
  • Posts: 9
Voltage Memory circuit
« on: May 27, 2015, 01:29:57 pm »
Hi,

I'm looking of some kind of memory like "1 bit memory circuit" but instead of bits i want store voltages.

For better understanding like the "1 bit memory circuit" if I send 1 bit it store and output 1 until reset, so, what i looking for is similar but with voltages, if i send 1 bit with one volt it store and output 1 volt until reset or new bit with voltage.


I don't have any clue...

Any ideas???
 

Offline rolycat

  • Super Contributor
  • ***
  • Posts: 1103
  • Country: gb
Re: Voltage Memory circuit
« Reply #1 on: May 27, 2015, 01:34:15 pm »
 

Offline FTVlogTopic starter

  • Newbie
  • Posts: 9
Re: Voltage Memory circuit
« Reply #2 on: May 27, 2015, 03:54:38 pm »
Hi rolycat,
thanks for the fast reply.
I'm trying to get an working circuit but i only find bad ones
 

Offline FTVlogTopic starter

  • Newbie
  • Posts: 9
Re: Voltage Memory circuit
« Reply #3 on: May 27, 2015, 04:46:31 pm »
hi,
I already look for sample and hold circuits but i don't find one working diagram

my idea is very similar with the tft pixel



but instead of Liquid crystals i need handle a LED

Any Idea?
 

Offline paulie

  • Frequent Contributor
  • **
  • !
  • Posts: 849
  • Country: us
Re: Voltage Memory circuit
« Reply #4 on: May 27, 2015, 04:56:24 pm »
   ...
 

Offline dom0

  • Super Contributor
  • ***
  • Posts: 1483
  • Country: 00
Re: Voltage Memory circuit
« Reply #5 on: May 27, 2015, 04:57:32 pm »
If this is about controlling large matrices of LEDs: don't even try to do it with thousands of S/H+ driver stages and long multiplexing lines. Use either digital multiplexing and PWM/PDM to control intensity or smart LEDs with a built-in digital bus (i.e. WS2812).

By the way, there where real "voltage memories" once: bucket-brigade devices. They're essentially a chain of S/H stages.
« Last Edit: May 27, 2015, 04:59:29 pm by dom0 »
,
 

Offline FTVlogTopic starter

  • Newbie
  • Posts: 9
Re: Voltage Memory circuit
« Reply #6 on: May 27, 2015, 06:29:09 pm »
Hi paulie,

I think it work for me thanks
 

Offline FTVlogTopic starter

  • Newbie
  • Posts: 9
Re: Voltage Memory circuit
« Reply #7 on: May 27, 2015, 06:38:59 pm »
If this is about controlling large matrices of LEDs: don't even try to do it with thousands of S/H+ driver stages and long multiplexing lines. Use either digital multiplexing and PWM/PDM to control intensity or smart LEDs with a built-in digital bus (i.e. WS2812).

By the way, there where real "voltage memories" once: bucket-brigade devices. They're essentially a chain of S/H stages.

Hi dom0,
I don't get your point,
i just want a simple way to control an led light without stay always controlling in real time, but release the IC to do other stuff.
the idea is very similar with the TFT matrix just that,
I'm no inside of large Led matrix however on first view use this circuit to handled a led matrix maybe work.

using digital components means using a lot of stream of data when the matrix be larger and larger and far away to be user friendly

Can you explain better you point?

Thanks for reply 
 

Offline dom0

  • Super Contributor
  • ***
  • Posts: 1483
  • Country: 00
Re: Voltage Memory circuit
« Reply #8 on: May 27, 2015, 07:13:26 pm »
For controlling many LEDs this approach requires one S/H per LED, which is a lot of components, compared to digital row+column multiplexing or smart LEDs, which don't need additional components for every single LED.

If, however, you only want to control a single LED you can do that with a S/H. On the other hand I can't quite see how controlling a single LED poses any significant workload to any MCU.
,
 

Online Alex Eisenhut

  • Super Contributor
  • ***
  • Posts: 3502
  • Country: ca
  • Place text here.
Re: Voltage Memory circuit
« Reply #9 on: May 27, 2015, 11:32:46 pm »
If this is about controlling large matrices of LEDs: don't even try to do it with thousands of S/H+ driver stages and long multiplexing lines. Use either digital multiplexing and PWM/PDM to control intensity or smart LEDs with a built-in digital bus (i.e. WS2812).

By the way, there where real "voltage memories" once: bucket-brigade devices. They're essentially a chain of S/H stages.

Hi dom0,
I don't get your point,
i just want a simple way to control an led light without stay always controlling in real time, but release the IC to do other stuff.
the idea is very similar with the TFT matrix just that,
I'm no inside of large Led matrix however on first view use this circuit to handled a led matrix maybe work.

using digital components means using a lot of stream of data when the matrix be larger and larger and far away to be user friendly

Can you explain better you point?

Thanks for reply

Like dom0 says, this is hardly a problem even for a Commodore 64. What "IC" are you talking about? It's hard to help when all we can do is guess.

You absolutely want to control a LED with an analog voltage? Then why not just use a 6 pin I2C DAC, write to it, and there you go?
Hoarder of 8-bit Commodore relics and 1960s Tektronix 500-series stuff. Unconventional interior decorator.
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Voltage Memory circuit
« Reply #10 on: May 28, 2015, 01:45:29 am »
If we're talking period hardware, it may be relevant:

I've taken apart some older computing hardware (printers, tape drives), which had a few large boards chock full of logic, and ultimately, a Z80 processor or two.  The motor driver, I believe, had a pair of analog outputs for driving the tape reel spindle motors; probably a bunch of inputs as well.  On the main logic board, there was clearly an S&H design: an 8-bit DAC (R-2R ladder and switch chip), followed by op-amps and stuff, and a few analog muxes (CD4xxx something).  The op-amps weren't JFET, as I recall, so presumably they would leak down and need periodic refresh, but even with a Z80, updates in the 10kS/s range would've been easy enough to do.  (May've also been some SAR ADC action with a nearby comparator, I don't remember.)

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline bson

  • Supporter
  • ****
  • Posts: 2466
  • Country: us
Re: Voltage Memory circuit
« Reply #11 on: May 28, 2015, 02:08:06 am »
This is all very strange since LEDs aren't voltage devices.  They have very low resistance, a nearly fixed forward voltage drop, and light up according to the current they're fed.

Are you trying to replace an indicator lamp with an LED?  Or something else?
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Voltage Memory circuit
« Reply #12 on: May 28, 2015, 02:50:14 am »
Presumably with a current-transforming resistor inbetween...
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline FTVlogTopic starter

  • Newbie
  • Posts: 9
Re: Voltage Memory circuit
« Reply #13 on: May 29, 2015, 01:06:57 am »
Hi every one,

Lets suppose (I'm curious about this) I go make an LED matrix 10X10, i can handle like a TFT right?
Mount the apparatus with arduino just for play a bit.

Lets make the Digital VS Analog with this scenario....


Im getting lost, because, maybe we get fast delay with a limited connection...
 

Offline rolycat

  • Super Contributor
  • ***
  • Posts: 1103
  • Country: gb
Re: Voltage Memory circuit
« Reply #14 on: May 29, 2015, 09:40:19 am »
If you are curious, for a 10x10 LED matrix you could reduce the number of S/H elements by using 10 in a column and multiplexing the rows. Analog Devices make a couple of (expensive) 8-channel S/H ICs, so you could make an 8x8 matrix with just one of these.

However, if you just want to reduce the processing and the number of connections from your main microcontroller it would probably be simpler and cheaper to use a separate display controller.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf