ADS1244 Current

Other Parts Discussed in Thread: ADS1244, LM1117, ADS1254, OPA350, OPA335, CC2530

I am having an intermittent problem with the ADS1244. Most of the time, it works exactly as planned. However in some cases it draws a significant amount of current (sometimes up to 1 amp). I was breaking down my circuit and one thing I have noticed that may be different is the order of the supply voltages. Would it make a difference to receive the 5V analog Ref before the 3.3V digital Ref? What could also be happening to make the ADS1244 to heat up?

Thanks.

over 13 years ago

0 hj over 13 years ago

Intellectual 650 points

I think I found a possible problem. I found that at times my circuit applied slightly more than 3.6V to the DVDD pin for short periods of time. Could that possibly damage the ADS1244 and cause high current draw? Maybe causing a short inside the ADS1244?

0 Bob Benjamin over 13 years ago in reply to hj

TI__Guru** 113765 points

Hi HJ,

For the device to initialize properly including the initial calibration sequence it is best that both supplies come up together at the same time. What you are seeing is a latchup condition. If DVDD exceeds 3.6V the absolute maximum is being exceeded and could cause the latchup. You might want to prevent the condition using a zener or TVS diode on the DVDD supply.

Best regards,

Bob B

0 hj over 13 years ago in reply to Bob Benjamin

Intellectual 650 points

Thanks Bob. Going back to the other problem. Does it make a difference to the ADS1244 to receive the 5V supply for the analog pin before the 3.3V supply?

0 Bob Benjamin over 13 years ago in reply to hj

TI__Guru** 113765 points

Hi HJ,

I will have to do some more investigation with respect to supply sequencing as it relates specifically with this device. I do know that we have some parts that require specific power sequencing. The datasheet for the ADS1244 does not specifically say that you must apply power in a specific way, but Figure 11 shows AVDD and DVDD coming up together, and in the Power-Up discussion at the top of the page (10) states that both supplies should be up and settled withing 100ms for minimizing any offset and gain errors. Unfortunately it does not state anything regarding latchup.

Best regards,

Bob B

0 hj over 13 years ago in reply to Bob Benjamin

Intellectual 650 points

Hi,

Did you find any information about powering ACDD and DVDD separately causing latchup? I was able to shorten the timing between the powerups but I still have similar issues with the ADS1244 in fewer cases. Thanks Bob

0 Bob Benjamin over 13 years ago in reply to hj

TI__Guru** 113765 points

Hi HJ,

I have a request in to some designers, but have not heard back from them yet. I will respond as soon as I have more info.

Best regards,

Bob B

0 Bob Benjamin over 13 years ago in reply to Bob Benjamin

TI__Guru** 113765 points

Hi HJ,

The designer told me that there is the potential for latchup if one supply ramps before the other. He told me I would have to verify it on the bench as he doesn't remember for sure which supply should come up first. He thought it was the digital as the digital resets and controls the analog portion.

I will need to setup a test to verify this and hopefully I can verify this next week. When you have the issue, is it destructive to the ADS1244 or will it work again if you power down and reapply the power?

Best regards,

Bob B

0 Bob Benjamin over 13 years ago in reply to Bob Benjamin

TI__Guru** 113765 points

Hi HJ,

I set this up in our lab, and repeated power up cycling numerous time over both AVDD first then DVDD, and DVDD first then AVDD. I did not see any adverse effects or high current demands with either situation.

I would look for possibilities on your board where an inductance or capacitance may apply an over voltage (or under voltage) condition which may be exceeding an absolute maximum condition for the part.

Best regards,

Bob B

0 Wilson Chiu39 over 10 years ago in reply to Bob Benjamin

Prodigy 220 points

Hello Bob,

I am facing similar issues as HJ. The DVDD and AVDD of ADS1244 are supplied by two different power sources. Initially it was ok. When it start measuring the voltage for a variable resistor. It draws a lof of current which heats up the regulator supplying AVDD. Then in a short while, the ADS1244 break down.

A very unstable outcome. Burn quite often.

0 Joseph Wu over 10 years ago in reply to Wilson Chiu39

TI__Guru** 100925 points

Wilson,

As mentioned in the previous posts, you want to make sure that the maximum power supply is not exceeded. What supplies do you use? Are they higher than the 6V for AVDD or 3.6V for DVDD mentioned in the datasheet.

There isn't normal case where the ADS1244 starts to draw a lot of current. In the previous post Bob mentioned that he tried cycling power with different start-up cases to see if he could find a latch-up condition, which did he did not.

It might help to give a couple of examples where you see this happen. Does this happen in certain conditions? Have you done this and monitored the input? As Bob mentioned previously, do you have any inductances and capacitances on your board that may cause any resonances that cause an over-voltage condition?

Joseph Wu

0 Wilson Chiu39 over 10 years ago in reply to Joseph Wu

Prodigy 220 points

I connect DVdd to a 3.3 v from a regulator LM1117 3.3. Avdd is connected to a 5v supply from LM1117 5v. The voltage being observed works normal, and no oversupply.

I did connect ainp to a filter with capacitor. But it's only like 0.1u.

There is no inductor on the circuit.

I used ads1244 to measure potentiometer. It works like a variable resister with 5v supplied.

From previous reading by hj and Bob, it looks I m not the only one experiencing this issue.

0 Joseph Wu over 10 years ago in reply to Wilson Chiu39

TI__Guru** 100925 points

Wilson,

Do you have a schematic that you are able to share? If you can, show how you have the potentiometer set up as well. Which starts first - DVDD or AVDD?

I know that Bob was able to test this device power-up and didn't find a condition where the part went into a latch up to draw excess current.

I would note that hj stopped posting, meaning he's figured out his problem or he decided to move onto another solution. If he's still receiving email notifications, perhaps he can respond to this as well.

Joseph Wu

0 Wilson Chiu39 over 10 years ago in reply to Joseph Wu

Prodigy 220 points

ADS1244.pdf

0 Wilson Chiu39 over 10 years ago in reply to Wilson Chiu39

Prodigy 220 points

What I did is just very standard voltage regulation. I have two power supply 15V and 5V from the same switching power and I made them common ground. step down from 15-8-5V and 5-3.3V.

0 Joseph Wu over 10 years ago in reply to Wilson Chiu39

TI__Guru** 100925 points

Wilson,

Have you attached an oscilloscope to look at the voltages on the device? Specifically, I'd track the 3.3V supply and the 5V supply. Could you please plot this on startup and post the result? You may want to also get a longer extended shot.

I've looked over your schematic, and I don't really see much that could be a problem. However, I did notice one thing. The LM1117 does have a minimum load current typ 1.7mA, max of 5mA. If the load current is not high enough, then the device may lose regulation. This is listed for the adjustable version, so I'm not sure if it applies to the non-adjustable versions.

I would also check to make sure that all of the input power to your board also has a ground attachments. I counted two inputs - the 15V supply and the D5V. You may also want to track how they power up with the oscilloscope as well.

I think it's more likely that you are having problems with the power supplies and regulation.

Joseph Wu

0 Wilson Chiu39 over 10 years ago in reply to Joseph Wu

Prodigy 220 points

Normally, ADS1244 works fine for a week or two. I just found out that one or two days before ADS1244 die, the reference part was meant to be 2.5V as I used 10K Ohm and 10K Ohm in series. Then It dropped to 1.6V for a day or two. Then It draw large current then burnt.

So far 12 ADS1244 has burnt. No survival!

0 Wilson Chiu39 over 10 years ago in reply to Wilson Chiu39

Prodigy 220 points

I use ADS1244 for measuring potentiometer, which has 3 pin output. 1) VCC 2) Sense 3) GND. I connected Sense to the amplifier as buffer. When I discount the output end of the amplifier and measure the voltage. It shows a perfect voltage follower. Once I connect to the ADS1244, the voltage drop significantly. Ex. If it was 3 volts, it now shows 0.9 volts.

0 Joseph Wu over 10 years ago in reply to Wilson Chiu39

TI__Guru** 100925 points

Wilson,

There are plenty of posts in E2E that describe a chip burning up, but most really are problems with the schematic. The most common issue is problems in the power supply, but there can be issues with the input measurement and digital I/O. I've seen your schematic and it's really basic so there's not much to comment on. Do you have a layout that I can look at? I'd take a printout of the layers on the board.

What are the supplies that you are using (before the regulators)? Have you tried looking at the supplies with an oscilloscope? I want to make sure that the AVDD and DVDD aren't oscillating in any way. In general, if the device is pulling excess current it is possible that some voltage is going over the absolute maximum rating. If it takes time to reach a failure, there could be some intermittent over-voltage condition. Have you tried removing the regulators and replacing the supplies with some sort of laboratory supply?

Are you sure that the part is damaged and that the problem isn't a short near the device? If there is a short near the device, or a very low impedance path, then there could be a nearby short that causes the device to get warm. I've heard of cases with different types of solder and flux that can cause this issue.

Also, you describe cases where the device immediately pulls out current so that the supply cannot be regulated to the proper voltage, and other cases where the device will operate for a week before failure. I'm not sure how you see your primary failure, but I'm still having a hard time understanding what you normally see.

At this point, other than the hj comments, I haven't heard of any problems with the ADS1244 having a destructive failure.

Joseph Wu

0 Wilson Chiu39 over 10 years ago in reply to Joseph Wu

Prodigy 220 points

Hi Joeseph,

I did check the AVDD and DVDD starting up.

It should be fine. I conducted the same experienment as BOB.

I have also tried the ADS 1254 as well. The same thing happen.

I discover that

for ADS1244 I set REF (-) to GND. REF (+) to 2.5V ( According to datasheet, If AINN =GND and AINP range should be from 0-5V)

Nonlinearity happens AINP larger than 2.5V. You see voltage drop when measures with voltmeter. The ADS1244 start draws large current.

for ADS1254 I set REF to 4.2V ( according to datasheet, If AINN =GND and AINP range should be from 0-4.2V)

Nonlinearity happens AINP larger than 2.1V. You see voltage drop when measures with voltmeter. The ADS1254 start draws large current.

Both ADS, I use a buffer amplifier OPA 335 or OPA350 to feed to the AINP.

Could you please advise if any setting in the circuit leads to this behaviour?

How do i solve it? So i can have an ADS measures from 0-5V.

0 Wilson Chiu39 over 10 years ago in reply to Wilson Chiu39

Prodigy 220 points

Hi Joseph, if you check ADS1244 datasheet, page 7. Figure two. It looks quite possible that switch S1 and S2 both on at the same time. Therefore, if AINP is << 2.5V. It will draw a lot of current and makes the capacitor not charged up properly. Hence, it creates non-linearity and possible burnt down after short period of usage. my CLK is at 2.4576MHZ 3.3V. 1) Would it improve if I change the CLK voltage to 5V? 2) if it is the case that both S1 and S2 are on at the same time. How do i avoid it ? I think HJ also faced this problem.

0 Joseph Wu over 10 years ago in reply to Wilson Chiu39

TI__Guru** 100925 points

Wilson,

For the non-linearity in measurement, there is no reason the input measurement should draw large amounts of current. It is strange that both the ADS1244 and the ADS1254 would both do the same thing. What are you using as your source voltage? Is there any input protection on the board that may be pulling the current?

Just to make sure, it might help to record both the input voltage (measured by a precision multimeter) and the output data (in code - not converted to voltage). I just want to make sure that the ADC is giving the correct output code when the input voltage is on the lower end.

You say that you are using the OPA335 or the OPA350 as the input buffer amplifier. How do you have that connected together? Is this a simple unity gain buffer? Are you using the same 0V to 5V supplies? Do you have a schematic for this circuit? I just want to make sure there isn't something on the buffer that is causing this problem.

Have you tried adding any series resistance to the input? The input impedance to the ADS1244 is about 6.5M and you should be able to insert a small 500 Ohm resistance in front of the ADC inputs. This would serve two purposes. First, it could protect the inputs in case there's any over voltages. Second, it could be used to measure any input current if that occurs.

As for Figure 2 on page 7 of the datasheet. This is just the way that the device samples the input. The input is sampled onto capacitors for the ADC. This sampling is non-overlapping and correct for the ADC or else the ADC wouldn't work at all. There should be no way for S1 and S2 to be closed at the same time.

Is it possible that the device is damaged? Is there a possibility that the device input pins or supply pins see a voltage outside the absolute maximums? You do uses a chain of LDOs to get your 5V and 3.3V supplies and I was wondering if it's possible that any part of the device sees a high voltage.

It shouldn't make a difference what the CLK voltage is. The datasheet says that the CLK can be driven by 5V regardless of the AVDD or DVDD voltage. However, the datasheet does recommend a 10 Ohm to 100 Ohm series resistance though.

Joseph Wu

0 Wilson Chiu39 over 10 years ago in reply to Joseph Wu

Prodigy 220 points

Hi Joseph,

Thank you for your reply. I am using TI CC2530 as my MCU.
I recently solve all the problem by changing the pins.
I used P.2.0 pin on CC2530 as the Drdy/Dout pin to communicate with ADS1244 or ADS1254.
I could have made anything to work properly for 3 months.

Recently, I just change to use P.1.6 pin on CC2530. CC2530 is a 8051 core.
Some port has pull up resistance like P.1.6 then it works without any hassles.
I have no idea why it works perfect now just using different port.

So, now problem has been sorted.
Either ADS1244 or ADS1254 works fine now.

According to the datasheet,
This ads 1244 if i set

REF+ = 2.5V
REF- = 0V
AINN= 0V
AVDD= 5V
DVDD= 5V

The range of AINP should be +/- 5V ????
According to datasheet is +/-2Vref..

0 Joseph Wu over 10 years ago in reply to Wilson Chiu39

TI__Guru** 100925 points

Wilson,

I'm glad that you are able to finally get some good results. When there are problems with communications, it helps to use a scope photo or the output from a logic analyzer to look at the digital communications.

When VREFP=2.5V and VREFN=0V, the full scale range is 5V to -5V. However, as I mention in your other post, the inputs should not go beyond GND-0.1V and AVDD+0.1V. Since the input is fully differential, you get negative measurements when AINN is higher than AINP. For example if AINN=5V and AINP=0V, the ADC will measure -5V.

Joseph Wu

Data converters

Data converters forum

ADS1244 Current