PDA

View Full Version : Testing DAC(digital to analog converter)


shantanu
05-27-2008, 12:05 AM
hi,
I have made current mode DAC chip and got it fabricated. I want to test;

I am facing certain problems;

Input to my circuits are both voltage and current (DC), is there any DC current supply lab equipment I didn't find any in my lab which says current supply source ( with DC voltage option).

2. I am planning to use counter to test my 12 bit and 6 bit DAC can anybody please let me know whether i should use synchronous or asynchronous counter. I am aware of the fact that output will be a ramp voltage ( analog form) .
Counter's information with specific part no. from any manufacturer will be really helpful.

" Testing is more complicated then design"

beenthere
05-27-2008, 12:14 AM
Resistors make excellent current-to-voltage converters. Do you have an accurate enough bench meter to detect the LSB change out of the DAC?

Why should the counter matter? If the count changes in a controlled manner, it hardly matters if it is synchronous or not. A couple of thumbwheel switches are really better sources for the conversions, as you have much better contrrol over the digital value.

mik3
05-27-2008, 12:14 AM
Your inputs will draw as much current they need but you have to supply them with the correct voltage as not to destroy them. Whats components do you use for the inputs?

You should use a synchronous counter otherwise your output will fluctuate rapidly before the final value of the counter is set on its outputs.

shantanu
05-27-2008, 12:39 AM
Thanks for the reply. First of all my circuit needs 5 Volts of VDD and I want to have desirable current value to supply. So yes, i can try using resistor but it will be cumbersome since then i have to characterize the channel resistance of MOSFET then calculate the current by using different resistor.

Is there any other way of supplying desirable current through a DC voltage supply in lab settings.

I am not using any component as if now for my chip. I thought same thing that a synchronous counter will be more accurate to use, but can anyone tell me specific part no. or how to find it on web.

I want to supply 1mA current through 5 volts VDD supply to my current input terminal. Is there any way of doing it without using any component and directly using power supply equipments in lab.

mik3
05-27-2008, 12:50 AM
Some power supply units have current limiting options (for example they can output 5 volts with a maximum current of 1mA), check yours.

Can you post a schematic of one of your input's circuit to see it?

shantanu
05-27-2008, 12:58 AM
ok ..I will show you the schematic..... I checked my power supply it does has current limit settings but ...i can't change current and voltage together...there...

the picture just shows the circuit, consider it having in a DIP chip with Iref/2 as pin for input current and other terminals Iref/4, Iref/8 etc ...being other pins of the chip where I should measure the current in the scope by connecting ammeter between VDD and MOSFET drain.

thanks
shantanu

mik3
05-27-2008, 01:02 AM
ok ..I will show you the schematic..... I checked my power supply it does has current limit settings but ...i can't change current and voltage together...there...

You set your voltage and then set the current limit to 1mA. If you short circuit it it should limit the current to 1mA.

Where is the picture?

thingmaker3
05-27-2008, 02:38 AM
Shantanu, the image files are on your C drive, not on the web. The "IMG" tags only work for image files on the web.

You also have the option of uploading your files to the AAC site as attachments: http://forum.allaboutcircuits.com/faq.php?faq=vb3_reading_posting#faq_vb3_attachment s

shantanu
05-27-2008, 04:34 PM
sorry for the delay, I am posting schematic as a .pdf file and coments from my last post....


the picture just shows the circuit, consider it having in a DIP chip with Iref/2 as pin for input current and other terminals Iref/4, Iref/8 etc ...being other pins of the chip where I should measure the current in the scope by connecting ammeter between VDD and MOSFET drain.

thanks
shantanu

beenthere
05-27-2008, 04:46 PM
An ammeter may introduce errors. You would do better by making a current to voltage converter. See figure 13 in the Wikipedia article - http://en.wikipedia.org/wiki/Transimpedance_amplifier.

shantanu
05-27-2008, 04:59 PM
when you say " current to voltage converter" I presume you are saying that i should use inverse of circuit to supply constant current to my chip , since the input is current coming from a dc source.

thanks
shantanu

" I hate testing "

beenthere
05-27-2008, 05:33 PM
Read the article. The converter circuit will take the current output of the DAC and convert it to a proportional voltage.

mik3
05-27-2008, 06:20 PM
You have to use op-amps to make voltage-to converters, look at the website beenthere said (you find more if you search in google), to control the amount of current at each of your inputs. Otherwise you have to have many power supplies with current limiting option which is expensive.

mik3
05-27-2008, 06:33 PM
There is something i dont like here. This is a DAC which converts the digital input to an amount of current according yo the value of the digital input, right?

As far as i know for a DAC you just apply at its inputs a high logic and low logic without "caring" about the amount of current it will draw. So, why do you want to control the current at its inputs?

shantanu
05-28-2008, 04:59 AM
Hi,
Thanks for all the input , I had lab supply in school and I tested my current mirror circuit. But I have another problem due to limited component supply.

I have 4 bit synchronous counter chips from which I have to make a 6 bit counter and one 12 bit counter.... I have part number:- 74LS161A ( its a 4bit synchronous binary counter ).

I tried going through data sheet of the part there is nothing mentioned about cascading and I am looking on net right now to see if I could find the implementation.

thanks
shantanu

SgtWookie
05-28-2008, 05:50 AM
Pin 15 is the ripple carry output.

All counters get the same clock pulse.

The 1st '161: Set pin 10 high using a pull-up resistor; it's the "T" Enable. That enables the ripple carry output; if not high the counter won't count. Pin 7 is the Enable "P"; it should be high when you want your 1st counter to count.

For the rest of the counters in the chain, set all pin 7's high using pull-up resistors.

Connect the ripple carry output (pin 15) of the 1st counter to the Enable T (pin 10) input of the 2nd counter. Connect the ripple carry output of the 2nd counter to the next counter's Enable T input, etc.

They will all then count synchronously.

Here's a link to Texas Instrument's datasheet:
http://www.ti.com/lit/gpn/sn74ls161a
Read the first couple of pages, then skip to the timing diagram on page 11.

shantanu
05-28-2008, 04:46 PM
Thanks for the reply. On the pull up resistor issue ..is it necessary to use pull up resistor...?

what should i connect parallel inputs too....?

Actually I want to use counters as ADC for my DAC's input ....

thanks
Shantanu

SgtWookie
05-28-2008, 10:43 PM
If you're trying to build an ADC, you're going about it the "brute force" way. It would be faster to use a successive approximation algorythm.

shantanu
05-28-2008, 11:16 PM
I am just trying to simulate my chip ..I got from foundry.....I don't want precise ADC ...all i want to do is test my DAC chip which needs digital inputs.

I though easiest way will be counter .....I am aware of the fact that ..it will generate ramp ...output...

thanks
Shantanu

SgtWookie
05-28-2008, 11:21 PM
OK, sure. Then use the 74161's.

Instead of pulling the individual Enable P's high, connect them all together. If you want them to count, put a logic "1" on the enable p line.

Unless you want to preset the counters to a specific value, leave them open. Just use the reset pin (tie those all together, too) to clear the counters.

shantanu
05-28-2008, 11:39 PM
Thanks for the quickie....but i didn't get it completely......when you said connect the enable pin all together ....it means i can connect all of them high or VDD ....( it is same as giving all of them logic one ) .

And if i understood you correctly .....you asked me not to connect anything to parallel inputs of counter ....it is not necessary , is that correct ?


I will attach my data sheet pdf , thanks a lot for your time .....

"learning is fun "... :)

SgtWookie
05-29-2008, 05:22 AM
Thanks for the quickie....but i didn't get it completely......when you said connect the enable pin all together
No, connect all Enable "P"'s together (CEP, Pin 7).
Additionally, you connect the Enable "T" (CET, Pin 10) to pin 7 of the first counter.
If they are held below 0.8v, the counters will not count.
If they are held above 2.8V, they will count.

The TC (pin 15) output of the first counter is connected to the CET (pin 10) of the 2nd counter. The TC of the 2nd counter is connected to the CET of the 3rd counter, etc.

ALL clock pins (CP, Pin 2) are wired together. This is where you apply your signal generator's square wave to cause the counters to count.

If you wish to use the reset function, ALL reset pins (pin 1) are wired together. You can connect these to the output of one of the counter's Qx pins if you wish for the count to reset early instead of "rolling over". If you do not wish to use the reset function, you don't have to connect them to anything.

....it means i can connect all of them high or VDD ....( it is same as giving all of them logic one ) .
Yes.
If they are 74LS161 IC's, the correct terminology is Vcc. Vdd is for MOS/CMOS/DMOS.

And if i understood you correctly .....you asked me not to connect anything to parallel inputs of counter ....it is not necessary , is that correct ?Yes. Unless you want to use the preset function and set the counters to a specific value, there is no need to connect anything to the Px inputs.

shantanu
05-29-2008, 04:03 PM
Thanks a lot, I will connect everything as you said. I hope it works for testing the DAC.

regards
shantanu