ramp time to voltage

What!!?? No PWM outputs?

You could keep it simple by using an LM334 constant current IC (3 pin device) by National and discharge it into a capacitor. You can tweak the current on the LM334 with a pot, and make a nice linear ramp. This is a 3 part solution, not including any power supply.

Rocky

Dual 12-bit voltage-output DACs. One of the few with such a peripheral, but rather slow and power-hungry.

Best regards, Spehro Pefhany

P.S. This is more like it:

;-)

Best regards, Spehro Pefhany

Thanks everybody, most of you are over my head. I would like at least one hour at one shot. I would like a diagram,( I could build heath kits) I'm not in the electronics trade.

Two approaches of many

1. Use a binary counter with appropriate weighted summing resistors connected to the outputs Adjusting the clock rate into the counter determines the rate of rise. The number of stages in the counter determines the resolution.

2 Use a microcontroller with an analog output such as Analog Devices ADUC812 see

--

Dan Hollands

Most of the flatbeds had high (often infinite) input impedance on the low voltage ranges. So a simple resistor-capacitor will make a nice ramp.

So,

The capacitor size depends on how fast you want to sweep and which range of the recorder you're on. Try 10 microfarads to start, 1 volt range maybe. Stay at 1 volt range or preferably less, otherwise the sweep will get nonlinear.

Open the switch to run the ramp.

Ignore all the complicated electronics suggestions.

John

recorder+

recorder -

And that guy behind the curtain ;-)

...Jim Thompson

Add an opamp to make an integrator and you get a buffered output.

John

For one hour, you'll probably do best with a CMOS 555 feeding a

Best regards, Spehro Pefhany

He said "an hour at one shot". 3600 uF cap or giga-ohm resistors?

I would say attach the minute hand of a clock to a rotary pot; use a multi-turn pot for a multi-hour delay.

Jonathan

How fast will that 0.010 F electrolytic self-discharge? The tolerance will not be so nice either.

I see a lot of +80%/-20% with the double-layer capacitors. A mechanical clock and a pot would be much more accurate. The error is all due to the nonlinearity of the pot, and that doesn't accumulate. Or use a PIC.

Jonathan

Hello Dan,

Or a DAC08, or two of them for more precision. These should be found in the bins of any decent lab and would save a lot of resistor matching and soldering.

Regards, Joerg

Hello,

If you are in a lab look around whether one of the function generators allows for really long timing. Sometimes they will indicate in "milli-Hertz", sometimes as period times. Only digitally controlled ones will be able to do that. That way you may not have to build anything.

Regards, Joerg

Will this be running off the line? Maybe two opamps and an 6-bit counter at the core...Do you know enough to build a small DC power supply, or any other construction techniques?

In article , Spehro Pefhany wrote: [...]

I suspect it would be easier to find a complete DAC than an R-2R ladder, but the idea is a very good one.

Thanks John, NOW, I understand.

The flatbed is line voltage. I can build but not design. Can do vdc. Is a discharging cap more linear? I can run my flat bed x axis backwards if this helps.