mcp6002 question

[ello]

hey there. i am facing an issue where i need some help in understanding. i use an input-circuit like for example here (https://mutable-instruments.net/modules ... ds_v02.pdf) but i get an output from the mcp6002 between 0 and 3.7 volts. i tried different configurations. with or without offset voltages at the input. but the max output remains around 3.7V. i don't understand that. is it normal? the mcp6002 is connected to 5v and ground, so shouldn't it be able to output 5v?? even if i just measure the output and use the inputs as a comparator i get 3.7V

any ideas??

thanks,
ello

edit: here is a picture btw:

[guest]

thats correct, it should go to 5V. try disconnecting the line to the arduino and measure at the opamp output. if your arduino is running at 3.3V, then it will clamp the output at 3.7V or so.

[ello]

thats correct, it should go to 5V. try disconnecting the line to the arduino and measure at the opamp output. if your arduino is running at 3.3V, then it will clamp the output at 3.7V or so.

thanks.. i didn't connect it to anything until now, cause i wanted to be sure to get the 5V before taking the next step... which would basically be doing something like in this test (https://tinyurl.com/yye6w3df)

maybe i am being picky here, but the arduino can map the 0 to 5 v input to 0 to 1023 and thus i wanted to be able to use the full range. clearly i could script it the way using a range 0...3.7v , but i am mainly interested in finding the cause for the mcp6002 to not provide the 5 volt output.

[neil.johnson]

The MCP6002 is a dual. By the looks of it you have not connected anything to the other opamp, so it might be oscillating or doing other weird stuff. Connect it up as a unity gain buffer with the input connected to ground, to park it in a safe stable state.

Neil

[ello]

tried that. but still the same thing. it wont go above ~3.72v

checked the voltages again. but the chip is definately powered by +5 and gnd and the -5v for the offset are there. i'd so love to understand. i already tried different chips. could it be this specific type (MCP6002 I/P 3AE 1639)?

[neil.johnson]

I would start with a basic unity gain non-inverting buffer and feed it a 0 to +5V triangle and check the output on a scope.

Neil

[emmaker]

In the MI circuit either the jack switch or the pot is providing another path to ground.

What happens when you add another 100K input resistor to ground?

Jay S.

[neil.johnson]

Reading through the data sheet sections 4.2 and 4.5 might be useful.

Neil

[ello]

thanks for your input. putting another 100k to ground doesn't affect the output (as i'd thought).
connecting the unused opamp (like mentioned under 4.5 in the datasheet, or to ground ) doesn't change things, either. and using it as a unity gain buffer and feeding in 5V outputs the same ~3.74... i am absolutely clueless

what exactly does 4.2 mean in the regard of missing ~1.26 v?

could it be that a whole batch of chips is not working?

[neil.johnson]

Hook a scope up to the pins and see what's going on? Also, try the 10k load to Vref/2 and monitor the current through it.

Where did you get the MCP6002 from?

Neil

[ello]

i'll try to find someone with a scope to check it.

got them here a year ago: https://de.aliexpress.com/item/32958179543.html
maybe complete garbage. should have tried them right away but got distracted...

oh, btw. to what degree am i bound to the mcp6002 in this use case? i have a few INA105kp laying around. it seems that the mcp6002 looks a bit more forgiving considering maximum ratings on the inputs.

[guest]

what are the voltages on all pins of the opamp? and what are the voltages on the 100k resistors inputs (points marked -5V and CV on your schematic)? its important to measure right on the parts, as maybe there is breadboard issue. its possible the opamp is fried, but i wouldnt waste the INA on this application.

[ello]

when the cv input is for example at 5V

pin 1: 0
pin 2: 0.01
pin 3,4,5,6,7: 0
pin 8: 5,04

if cv in is 0V:
pin 1: 3.7
pin 2: 0.43..0.47

if cv in is -12V:
pin 1: 0.37
pin 2: 0.59

if cv in is +12V:
pin 1: 0.05
pin 2: 2.74

i already changed the 6002 three times. one exploded because i shorted it somewhere. however, they are all from the same order. maybe i should get some different one from another location, just to be sure there is not a whole batch of garbage

btw, i moved the 6002 to different locations on the breadboard and this didn't change anything. other chips work on it, for example 40106 or tl074

[joem]

I'd be suspect of IC's bought on aliexpress -- that could indeed be the problem. Sometimes you get good ones, but sometimes you don't. It's long been theorized that some of what's sold on there are things rejected by the manufacturers. MCP6002's are pretty cheap from the big distributors (like Mouser, etc), so there's not much reason not to get them from a big distrubutor, IMO.

[guest]

yeah, sounds like the chip is no good.

[neil.johnson]

Yes could be that. Never ever buy ICs from untraceable vendors.

Also would be worth adding a 100n supply decoupling cap across the power pins.

[Synthiq]

Measure the Vdd current into pin 8 and verify that it matches the specification (~200uA).

[ello]

i tried decoupling cap, but that didn't change a thing, so i left it for now. when soldered i add them.

cannot check the current cause my multimeter seems broken. it always shows 0µA .. but the powersupply is supplying modules, too, so i think that the supply is not the issue. guess it is a bad batch than. damn it. luckily it wasn't that pricey. too bad that when i order some parts from mouser or reichelt, the postal service will cost more than the parts themself

[rich_de]

If it goes from a clean zero to about V+ - 1.5V, my bets would be on a relabeled LM358. You could use that datasheet to see if other characteristics match (e.g current as suggested by Synthiq).

[migueldetrool]

I have had the EXACT same problem. I was trying to recreate the same MI circuit, with the same MCP6002 op amp (even has the same markings as yours), and got 3,7v on output with a -5v input.

I believe that it is a shitty chip from aliexpress, that is not rail to rail. Traditional op amps go to within 1.5 volts of the power rails, which, when supplied with 5 volts, is just about 3.7 volts... The vendor on ali was Lushigher IC Store, was it the same for you ? I'm waiting to try my circuit with parts from a more reputable seller, because I can't see any other problem, and MI's design seems perfect.

[emmaker]

cannot check the current cause my multimeter seems broken. it always shows 0µA

Just want to make sure, you measured current in series, not in parallel like voltage, right?

Also the - input pin on op-amps is at virtual ground and you measure it as a current and not a voltage.

Jay S.

[ello]

i believe this issue is settled than (bought garbage). gonna buy a bunch of them right away. thank you all.

Jay, i put it between the positiv supply voltage and pin8. like if it was a cable bridge. i already had this issue in another regard, so i really think my meter is broken

[paperCUT]

I'm wary of even getting LEDs from Ali these days since last batch were utter crap, life's too short to deal with dodgy people.

The bug fixing time is worth much more than the original part from real stores.

(Rant over )

[migueldetrool]

For future reference, can you confirm if it was the same reseller on aliexpress ? Lushigher IC Store ?

[ello]

For future reference, can you confirm if it was the same reseller on aliexpress ? Lushigher IC Store ?

seems so. the link i shared was taken from my order-history. guess i'll let them know what i think about such. sadly long time has passed since i got aware of this