What causes the 74hc595 to draw the spec'd 70ma? I have a project that is 12v powered and with two of those and the arduino it comes out to 120ma draw at times and 50ma other times, the arduino draws like 35-40ma so they draw between 10-80ma combined, but it varies randomly it seems, moving between the two values kinda quick,
the outputs can all only draw 30ma max since its going to 2.7k resistors so what's the rest being used for?
the reason I ask is I wanna use a cheap and small 7805 powered off 12v but at 120ma its a bit much heat, the 50ma isn't so bad
Double check your wiring somethings shorted to ground one or more outputs are shorted to ground.
Doc
None are shorted for sure, its all multimetered out good
and I tried just one output and it does the same thing, just goes back and forth slow enough atleast that the multimeter has time to settle on each usually
And I mean it does say supply current 70 ma max and it seems to be between 5 and 40 ma each, so I just wonder when it actually needs all that so I can minimize it, and I don't think its when loading the data because im doing that constantly
- Are any of the chips getting warm.
- Do you have any inputs not pulled up or down, any at all.
- Does the power change if you pick it up or touch it.
- Is it (the board well bypassed a 100 uf on the Vcc line and .1uF caps on Every chip.
- Can you step this thing trough to find an address that is the high current one or ones
- Can you repeat it in a regular pattern or is it random in appearance or regularly cycling
- do you know that the IC's are all good.
- does the quiescent current change at all when not being driven with serial data
- is the Vcc stable at all times and if not can you determine the range and it any change in voltage during periods of high current draw.
- Do you own an O'scope and can you use it well.
... There are a lot of possibilities most of which I have covered here most as you can see are power supply related have you checked it a third time?
When I worked for a living I was a design engineer for a small company in Costa Mesa Ca. I worked there for 20 years and I had a rule that always worked for me
After I finished a design, all modules were checked I had a proof board made a sample run W/O silk or mask just solderplate and I assembled one and verified and many times repaired mistakes. I would then modify my final documentation and board work prepare the paneled art work as I usually made 6 or 8 boards on a panel and then I would go home regardless of the time of day eat and go to bed and the next morning recheck everything again before I told my employer that the project was done. He always asked me If I had "Checked It" and my answer was always yes to the limit of my ability.
The questions I asked above were for PSU, oscillation in any of the chips and pattern dependent issues that might reveal where in the work your board deviates from your design goals and specifications. For Engineering this is a critical evaluation. What you are doing in the above questions is looking for that sore tooth in your design.
Many years ago I was a techie and a good one by the standards of the day... I Never "Applied" for a job after the first one I had. I would every 5 or 6 years get a call and decide that it was a good time to change my neighbors for new ones. One of the things I learned is that there is a pattern that follows all trouble shooting and that is
"Kick it and see which way it jumps". Things that oscillate will usually change when you change feedback with your digital voltmeters... ("fingers" remember I worked with tubes for all my life, from TV's and Radio's to KW class transmitters from 50 mhz to 800 mhz) often times just 'feeling' the PCB or breadboard is enough to alter conditions and ca8use a current change, if the board is quiet (not clocked) and you can make any change in current draw you have a by-passing or layout issue and so on and so forth.
I write too much... it is hard though sometimes to get an idea across. put it through as many tests as you can you may well discover a great deal.
If you own or have access to at least a 20 Mhz BW scope so much the better probe the power and ground rails clock and data lines ir or are any at all of the output voltages from the S.R.'s out of spec, especially low? check it out and I am on california time and by the computer a good part of the day and evening...
So Dude, Let me know.
Doc
I've got a .47 uf and 470uf cap for both chips, still waiting for 50 .1uf caps off ebay..... but my ancient oscope says the power is clean not even a slight ripple or noise
no warmth, when I took the current measurement I was just toggling one output, but it didn't seem to match the current change
the outputs are driving the inputs of uln2803 so those have built in 2.7k resistors and so every output should have an even draw,
ill have to do some more tests after work today, see if touching it does anything and ill make a sketch where it cycles all the pins for a few seconds and a few seconds no data transmission and see how the current changes
I thought if anything it would have a current draw when data is being fed in, but I was doing so at a pretty slow rate, like 2 bytes every second or two so the time its clocked in would be so short the multimeter would never even know or so I would think
Okk, so after some tests I found out something interesting
I put it to one output on and off really slow and saw a steady 120ma
I put it to one output on and off really fast and saw a steadyish 120ma
I checked by touching pins and stuff, re checked oscope and it was clean as possible
I did notice during my setup phase the ma was around 50ma most of the time, which baffled me since it was busy but setting them on and off fast would give me 120ma
I've found that if I write all the outputs high it is drawing about 50ma total,including arduino but all low it draws 120ma
I find this a little odd, why would it draw more to have the outputs off? I know its got tri state outputs, but driving a transistor array it shoulding be sinking any current
it does this with or without the led vcc connected
And besides im measuring the +5v supply not the negative return so if I was sinking current it wouldn't affect my measurement (not saying it is or should be)
I didn't see a circuit attached, could you post one... I do have an Idea. If you are using someone elses drawing, I want to see what ever changes you made and likely you will have your answer it does sound like a wiring issue but not yours directly... I wish I could write code as well as I can diagnose and fix other's issues...
I personally think Smiley Faces are silly...
Doc
Ill make one up after my afternoon shift, its my own design although really not complex, just decoupling caps and two bigger ones for the 5v and led supply
Other than that just two faisy chained74hc595 with the outputs going directly to the inputs of two uln2803s
the only current draw I expected is from the internal 2.7k base resistors of the uln2803+ quiescent current of the 74hc595
Everything is working as expected. Just not the draw I expected at tge times i expected
Yeah but sometimes it's what is down stream that is frequently the issue, I await your pleasure.
Doc
srry for the shoddy shematic, didnt have much time before bed so this is it for now hope its clear its pretty sinmple
ill have a more proper cad schematic up tomorow after work , just wanted to get it up preliminarily