TI E2E Community
/etc... Amplifiers & Other Linear
/etc... Amplifiers & Other Linear Forum
Why is the LM2904 behaving this way?
I'm using an LM2904 as a current sense amplifier as shown in the attached schematic and I don't understand why when there's no current flowing through R5, the current sense resistor, there is still an output of 2 v?
I tried putting a trimpot in place of R6 to eliminate the offset across the inputs, but it didn't eliminate the 2v on the output.
The LM2904 can only sink ten's of micro-amps and maintain a near zero output voltage. At higher sink current a PNP emitter provides current, but the minimum output voltage is about 0.7V or more.
A rail to rail output op-amp is required to operate properly.
Regards,Ronald MichallickLinear Applications
TI assumes no liability for applications assistance or customer product design. Customer is fully responsible for all design decisions and engineering with regard to its products, including decisions relating to application of TI products. By providing technical information, TI does not intend to offer or provide engineering services or advice concerning Customer's design. If Customer desires engineering services, the Customer should rely on its retained employees and consultants and/or procure engineering services from a licensed professional engineer (LPE).
Thanks for the reply.
I had some INA121's laying around , so I threw in one of those and it works well, even with a single supply.
Would a TL082 do the trick? (They're much cheaper.)
I noticed that the V+ rating is max +/- 18 v, but the INA121 seems to work fine with a 24v single supply.
The TL082 will have the same problem.
It is OK to use the INA121 at 24V singal supply.
Maybe you could recommend a rail-rail device that is pin for pin compatible with the 2904?
Given that I need a device that can handle input voltages around 20 v, I need to run V+ above that, which I currently have at 24v.
What I need is a device that handles the above and can maintain a Vol somewhere much closer to ground, given that I'm using it in a single-supply configuration.
As an option, I will try raising the output impedance of the circuit. Do you happen to know how high I can go with the resistance in the feedback voltage divider for a LM22680? That's what this circuit is driving. According to the LM22680 datasheet, they recommend keeping the voltage divider total resistance less than 10k ohms, which is interesting because WebBench specified a resistor pair that totals 17k ohms, above the recommendation.
Will it cause any problems if I eliminate the feedback resistor voltage divider for the LM22680 and just drive it with the op-amp output, which will typically be below the 1.25v reference input?
One way to make the circuit work is to offset the zero current output voltage. The termination of the differential amp from non-inverting input to ground sets the output voltage when in balance (inputs equal). That way the opamp does not need to go to VCC- (GND). However you have to deal with the output offset.For an answer on the LM22680, please start a new thread with that part number.
1033.GROUP SENSOR DS Rev1.pdf
I am Paolo Montresor E/E engineer at UFI FIlters SpA company, located in Italy.
We developed a heater driver control circuit using LM2904 op-amp, one as inverting amplifier, and the other one as comparator with hysteresis.
You can find the schematic attached.
We noticed a "drift value" of the low output of the comparator (U2 pin7) at around 800mV, when the signal in the non-inverting input (U2 pin5) is roughly between 2 and 3.5V (3.5V is the fixed voltage level of the inverting input).
Such output "drift value" actually keeps the MOSFET driver circuit switched on, causing the mosfet to be weakly switched-on, sometimes going to work in linear region and thus dissipating a lot of power (then overheating).
Can you please confirm/correct our analysis?
If it is correct, which could be the best solutions?
I see these options:
a) changing op-amp to a rail-to-rail
b) adding a 3k resistor between Q3 base and gnd
c) increasing the value of hysteresis network (R6 and R8) keeping the ratio constant
What would you suggest?
Thank you in advance for your help, best regards,
I received a phonecall confirming that Ron is out of office till Thursday.
I write for asking if there is any possibility to have a very quick reply by today.
Unfortunately customer is pushing to have a report by this evening. We already analyzed the situation and hypotized several possible solutions, but we need your confirmation before sending it out.
I know I am requesting a lot, but we have no time and we are pushed minute by minute.
Sure of your collaboration, I thank you in advance.
I am looking in to this and will try to respond soon Today. Thanks.
The low output should be sitting near ground. Only the Voh is high. Do you get the same error if you try another omamp.
Thanks for your reply.
Actually we have not had the chance to change op-amp, but we have the same behavior on every piece (so far I tested myself about 10 pieces, every one with the same behavior).
When the second op-amp triggers low, then we have a drift voltage at the output of the op-amp, around 0.6 - 0.8 Volts.
I think it is related to sink current. On that branch we have a current up to 180uA, and according to the figure at page 7 of the datasheet (output carachteristics current sinking), it seems that for 180uA the op-amp actually gives an output of 800mV.
I would just like to have a confirmation of the analysis as explained in my previous message, and your feedback on which of the proposed solutions would fit better.
In case I misunderstood anything, plase correct me.
Thanks again for your support, best regards,
Interesting. So the current is flowing through R6 into the amp. If that's the case, then you need bigger resistors through that path. Did you try that?
We tried using bigger resistors (R6 and R8 100k and 200k) and the problem is no more present. But we are afraid that limiting so much the current, then the circuit becomes more sensitive to noise, and we need to run again EMC tests.
Instead putting a voltage divider to Q3 base (3.3k resistor between Q3 base and gnd) we still solve the issue, with no major modification of the circuit, currents and so on.
What do you think?
I think a 3K or lower resistor to the base of Q3 is the best option.
Thank you very much Andrew for your prompt feedback.
All content and materials on this site are provided "as is". TI and its respective suppliers and providers of content make no representations about the suitability of these materials for any purpose and disclaim all warranties and conditions with regard to these materials, including but not limited to all implied warranties and conditions of merchantability, fitness for a particular purpose, title and non-infringement of any third party intellectual property right. TI and its respective suppliers and providers of content make no representations about the suitability of these materials for any purpose and disclaim all warranties and conditions with respect to these materials. No license, either express or implied, by estoppel or otherwise, is granted by TI. Use of the information on this site may require a license from a third party, or a license from TI.
TI is a global semiconductor design and manufacturing company. Innovate with 100,000+ analog ICs andembedded processors, along with software, tools and the industry’s largest sales/support staff.