My project is derived from only reading information from datasheets only. The project is to create a motor x 8 controller. I am using a uno board with one ULN2003 and a 74HC595. For this project I am only using vero boards.
I have already started testing the 74HC595. It appears that I have an output voltage issue already. I have a 6.2 DC power supply 4 x 1.5volt AA batteries. My output voltage on ULN2003 appears to be about half the VSS voltage.
Suggestions please!
Post a schematic, please.
It appears that I have an output voltage issue already.
Where? The output of the Arduino, 595, 2003?
My output voltage on ULN2003 appears to be about half the VSS voltage.
That makes little sense. [Vss](http://My output voltage on ULN2003 appears to be about half the VSS voltage.) is usually ground, did you mean Vdd (6.2V)?
My output voltage on ULN2003 appears to be about half the VSS voltage.
You do no that chip does not output any voltage. It just pulls down, that is, connect the output to ground or not.
Yes. As Grumpy has implied, the outputs on this device, "pulls down". These outputs are like switches with one end connected to ground. The other end needs to be tied to a load that is connected to a positive supply [up to 20V].
In other words, one side of the load is connected to one of the outputs, and the other side is connected to a positive supply. if the input is driven to 5V, the output can handle up to 140mA. If the input only goes as high as 3.3V, then the output can handle up to 100mA. And, in the proper "techno-speak": For a 5V input, the output can sink upto 140mA, etc.
Gates:
The project is to create a motor x 8 controller.
So you want to control eight motors with this chip?
How small are they?
Leo..
I am going to start testing with these motors.
Those motors claim a stall current of 0.25A so should be OK with the ULN2003 - be sure to connect the
diode terminal of the ULN2003 (labelled 'COM' in the datasheet) to the positive motor supply.
You should be using the ULN2803, not the ULN2003, if you want 8 channels (rather than 7).
thank you,
I am not concerned with the amount of channels available. At this stage I would prefer reliable operation.
The datasheet I am looking at currently is the 74HC595. I am going crazy over pins 10 to 14. I will place my oscilloscope probes over them tomorrow.
10, MR# should be HIGH
13, OE# should be LOW
12 is the latch, normally low
to clock in a new byte:
pulse the clock pin (11) once for each bit set on pin 14 (data in), then pulse the latch high then low
to copy the internal state to the output pins simultaneously.
To chain 2 or more together pin 9 on the first chip connects to pin 14 on the next, and so on, parallel
clock and latch.
Do not attach a capacitor to the latch pin (as seen in some broken tutorials out there).
A 100nF decoupling cap should be placed close to the '595's power pin.
Keep all wiring short. You cannot run shift registers reliably at the end of a ten foot cable, for instance,
as fast logic signals are degraded and can generate false clocks and crosstalk.
yeah, I already had a 100uf decoupling capacitor attached. I removed it from a development board.
BTW. I wouldn't go near my girl friends cat with a ten foot pole.
I know to keep signal lines as short as possible. keep away from medium to high EMF.
I have experienced problems with bread boards in the past. hence I stick to vero strip boards, etc.
They cost a little more. but a lot less pain.
Gates:
My project is derived from only reading information from datasheets only. The project is to create a motor x 8 controller. I am using a UNO board with one ULN2003 and a 74HC595. For this project I am only using Vero boards.
Inappropriate choice.
The ULN2x03 is an essentially obsolete device, using Darlington transistors which lose at least 1.5 V and get unnecessarily hot in the process. That is the reason you are losing so much voltage.
And you are driving it with a 74HC595! A TPIC6A595 combines the functionality of both and uses proper FETs as drivers - if your motors are rated at 250 mA stall current, then unlike the ULN2003, the TPIC can drive all at once.
By the way - for a serious project, use the Nano instead of the UNO - much easier to work with.
Paul__B:
Inappropriate choice.The ULN2x03 is an essentially obsolete device, using Darlington transistors which lose at least 1.5 V and get unnecessarily hot in the process. That is the reason you are losing so much voltage.
Wait-a-hold-it... There's a MOSFET version of the ULN2003 [just add an 'F' or 'V' at the end]:
Paul__B:
And you are driving it with a 74HC595! A TPIC6A595 combines the functionality of both and uses proper FETs as drivers - if your motors are rated at 250 mA stall current, then unlike the ULN2003, the TPIC can drive all at once.
But, yeah -- cool looking device that combines the two! But, an expensive little bugger. Whereas, the combination of a 74HC595 and ULN2003V is around a buck [vs the 4 buck TPIC6A595!]
AND, if those motors do, indeed, have a stall current of 250mA, then the ULN2003V12 should handle them with ease (with a caveat):
What the diagram (above) is telling us, is even when the ULN2003V12 is sinking 250mA, the voltage at the Drain will be around 450mV (MUCH less than the ULN2003 Darlington version). AND will be MUCH less when the motor is doing its thing [unless, of course, its thing involves a heavy load]. BUT, the ULN2003V12 won't do well if all those motors start up at once! So, the programming will need to space startups accordingly [limiting them to 2 or 3 at a time]. Also, a problem will be if more than 3 or 4 motors are running with big loads, or are simultaneously stalled.
The overall current load the ULN2003V12 can handle is 700 to 1000 mA, which is dependant on how hot it gets inside the ULN2003V12!
The TPIC6A595 on the other hand, CAN manage all those motors at once, no matter how much they misbehave [again, assuming the 250mA stall current is correct]!
Yes, a useful device indeed, the updated version of the - essentially obsolete - ULN2003 with a few limitations such as its working voltage of 12 (16 if you propose to argue the point). Certainly useful to know if you do not want the shift register function.
The cost comparisons get more involved.
Thank you all,
Reponses have been amazing. Looking at the 74hc595 shift register. appears that a single 74hc595 might be able to deliver a clock signal etc. to more than one 74hc595. The oscillator stays the same, but data shift concerns me a little.
but data shift concerns me a little.
Why?
