New USB port P/S and cable, voltage too low

[ke1g]

I've just today received an AdaFruit PID 501 5V 1A USB power supply, as suggested on the RaspberryPi products list, and a USB A/MicroB cable ordered at the same time on the PID 501's page. I'm seeing voltages on (both) my RaspberryPi model B's tp1 and tp2, presumably the source of the dreaded trouble initializing the SD card error I'm seeing.

Note: I didn't get the RPi(s) from AdaFruit, just the power supply and cable. (I ordered my RPis before I knew about AdaFruit.)

With nothing connected but the power supply and SD card, I'm seeing about 4.74 V on the test points. With HDMI connected, I see more like 4.69 V and even less with my keyboard and mouse plugged into USB as well. Additionally, if these voltages glitch lower with current spikes (such as when trying to write to the SD), my DVM is unlikely to catch it. (I see the same voltages on both of my model Bs.)

A further note: The A/MicroB cable listed on my invoice is PID 592, while what was shipped is labeled PID 4867.

I purchased these parts with the thought that they were suitable for running a RPi, including model B. If they should be, then I can think of several possibilities:

1. The PID 501 isn't keeping up its advertised 5.25 V under the load of a model B Pi, especially when running HDMI.
2. The PID 501 is making it, but the cable is too resistive (inadequate gauge of conductors).
3. The PID 501 and PID 592 are OK (presumably you've tested this combo and your customers are successful using it), but the shipped PID 4867 is too resistive. (If this is the problem, you may want to be careful what you are shipping.)
4. The AdaFruit parts are fine, but somehow my model Bs (ordered from Allied last July and finally received in December) require far more than the usual current (I consider this unlikely, since the ultimate source of all the Pis is presumably the same).

So I'm looking for suggestions on further checking before I as for an RMA. I'm an electrical engineer by training, so I'm willing attempt non-destructive disassembly of the PID 501, or destructive disassembly of the A end of the cable, to measure the voltage at the PID 501 output to further isolate the problem (I don't have a USB A to A M/F breakout connector), but not at the cost of not being able to return out of spec parts for refund or replacement.

Please advise.

Bill, KE1G

[adafruit_support_bill]

I agree. The Pi is the least likely suspect.

I'm willing attempt non-destructive disassembly of the PID 501

If you can measure the output at the P/S with and without load, that will narrow it down to the P/S or the cable.

(The 4867 is probably the manufacturer's #)

[ke1g]

Measuring on the PID 501 circuit board (DMM is a BK probe meter):
No Load: Varies between 5.12 and 5.24.
Pi Load: Varies between 5.17 and 5.2.

So, unless the connector is bad (and all the solder joints look clean and bright), the PID 501 is solid, and the cable is droping more than 0.4V under the lighter non-HDMI load.

PID 4867 could be a manufacturer's number, but since they are using literally "PID:" as a prefix, rather than the more common (in my experience) "P/N", "Part No:", "PN", "Part #", etc., I figured that there was a chance that they were labeling with your designation. The other stuff on the tag on the bag is "MONOPRICE(tm)" (with R in a circle for (tm)), "USB-MicroUSB5-03" (which looks more like a manufacturer's designation to me), a UPC type barcode with "44660-04867" under it (and "8" before the bars, and "1" after) - note the last four digits match the PID number - I don't know if that's their inventory control bar code, or one they put on for you, and "Made In China".

Bill

[adafruit_support_bill]

Sounds like the P/S is fine. Let's try replacing the cable. Contact support@adafruit.com with a link to this thread.

[AlWa]

I too am having power issues with my new #501 USB power supply.

Am experimenting at the moment with a Mega 2560R3 bread boarded to 2 10K thermistors and a I2C 16X2 LCD display with resistor reduced display LED brightness. I have written a sketch to display the temperatures and turn on a 4 relay module based on a temperature difference threshold.

The final project will be measurement of 7 thermistors and driving 10 relays. Each relay uses a opto-coupled coil circuit that is driven by a 15-20 ma draw Mega digital pin. Each relay coil circuit carries 75 ma @ 5V when activated. So the total 5V current for the project would be nearly the rated 1000 ma for this power supply if all the relays are on (not knowing what the LCD draws).

First let me explain my experimental power layout. The thermistors and LCD are powered off the Mega 5V pin and nearby ground pin. The 4 relay module 75 ma coil circuits share a common power pin that I connected to the Vin pin. It appeared to me based on the Mega circuit diagram, that Vin is connected to power-in before the Mega 5V regulator and I hoped the relays would then ride on the USB power supply regulator, not the Mega's. The relay module ground (common pin for both 20 ma and 75 ma circuits) is connected to the ground adjacent to the Aref pin in an effort to isolate the high relay ground current from the Mega's Vcc. Can't tell what the Mega ground layout is from the circuit diagram, just hoped closer to power-in was shorter path. I am concerned about sinking a 1000 ma through the Mega ground since I thought the Mega is a 200 ma max device.

Here are some voltage measurements at the bread board power busses with a 1 ft USB cable:
Laptop USB powered
Relays off / 4 Relays on
Vin 4.48 / 3.78
Vcc 4.96 / 4.69

#501 USB powered
Vin 4.61 / 3.92 (for only a 300 ma draw??????)
Vcc 5.11 / 4.85

I am seeing temperature jumps with my bread board setup of 3 oF at times (but not always) when the relays come on. I am using the Aref circuit and analopReference(EXTERNAL) call. I think I want to power the thermistors with the 5V supply because some are 50 feet away. I hope that clearing up the power issue will help this.


So, did I pick the wrong power supply for this project? Also, I am unclear as to whether I can use a seperate 7-12V power supply through the jack concurrently with a USB power supply. What is the best way to power and ground this project?

[adafruit_support_bill]

Here are some voltage measurements at the bread board power busses with a 1 ft USB cable:
Laptop USB powered
Relays off / 4 Relays on
Vin 4.48 / 3.78
Vcc 4.96 / 4.69

#501 USB powered
Vin 4.61 / 3.92 (for only a 300 ma draw??????)
Vcc 5.11 / 4.85

How do you get a VCC higher than VIN? Do you have snubber diodes on the relay coils?
Can you post photos or a wiring diagram so we can see how you have everything connected?

[AlWa]

Now that you mention it, good question!

Yes, there are coil diodes ... but so what? I've attached the circuit diagram - one for each of the 4 relay coils. There is a Vcc to JD-Vcc jumper that I have removed to connect a Vin jumper for coil power. I now realize that the ground pin is only for the coil circuit and has nothing to do with the Arduino side of the opto coupler.

I would be more than happy to draw up a schematic of the breadboard setup if you would suggest a free way to do this. A picture is worth .... yada yada.

Sainsmart relay circuit from Ebay

4channel-relay2.jpg (29.99 KiB) Viewed 131 times

[adafruit_support_bill]

We use Fritzing for breadboard diagrams. It is a free download. http://fritzing.org/
We also have a library of Adafruit components for Fritzing:
http://learn.adafruit.com/using-the-ada ... rom-github

[AlWa]

My breadboard schematic:

BdBrd_bb.png (414.53 KiB) Viewed 113 times

I realized that I had another 1000 ma USB power supply for charging an IPod. When I hooked that up for power, I get similar results as the #501. I'm seeing the handwriting on the wall now. I need a better power supply. Does Adafruit have one that will hold 5V when there is a 1000 ma being drawn by the 10 relays of the final setup?

I've got a 9VDC/350 ma Class 2 AC adapter to experiment with but am reluctant to hook it up because it measures 16.25V open circuit. Should I be concerned about this - will it lug down to the 12V max spec when it just powers the Mega? Would like to see how the higher voltage works with the relays.

I must say though, it does surprise me that both USB power supplies lugged down to 3.6 V when all 4 relays are on. I measured 195 ma flowing from Vin; lower than the expected 300 ma draw because the voltage was well below 5V. So why call these 1000 ma power supplies; they can't even hold voltage at a 1/3 of advertised current! Where's truth in advertising?

Another question is whether I can use Vin and the Mega ground pin to power the relays. Will the Mega PCB traces handle 1000 ma assuming I get a better power supply. After looking at the Arduino Mega schematic, I see Vin coming to what appears to be the Vcc regulator (IC1) but don't see where Vin breaks out to go to the Vin power pin. I also see Vin going through a comparator to T1 that "interfaces" with USB Vcc going to IC6. What's that all about?

Any help you can give will be appreciated.

[adafruit_support_bill]

You are missing some details in your diagram such as power supply connection(s?) and base connections to the relay drive transistors (with current limiting resistors?).

From you description, I am assuming that you are powering the entire circuit via the USB port on the UNO. If so, that explains why your VIN is lower than 5v. VIN is the input to the on-boad 5v regulator. If you bypass that by supplying 5v to the USB port, what comes out of VIN is current you are feeding backwards through the regulator. http://arduino.cc/en/uploads/Main/Ardui ... ematic.pdf

[AlWa]

The relays were only a representation to show how they are powered and controlled.

I am powering the whole setup with USB power only with a 1 ft USB cable to the Mega board. As to the Vin voltage, can you reverse supply it though the 5V regulator - is that even possible?

So I studied the schematic you supplied (mine was pre-R3) and let's see if I've got the Mega (not UNO) power layout right:

Upper right of the diagram shows Vin coming from the barrel jack and going to U1, presumably the on-board 5V regulator. 7-12V in, 5V out. HOWEVER, I see no Vin branch-out prior to the regulator - it must be there to go to the Vin output (or input) pin.

Upper left of the diagram shows Vin supplying a /2 voltage splitter whose output goes to a comparator with the 3.3V supply. This comparator disables the USB 5V supply if Vin > 6.6V by means of T1 (a FET?). The Mega 5V bus branch-out occurs after this so the Mega 5V bus is controlled by the external USB supply - no other regulation other than for the 3.3V bus.

The left of the diagram shows the USB connector and its 5V first going to a 500 ma fuse before going to the T1 on/off gate.

So I conclude that when USB connector powered, Mega total current is limited to 500 ma. A 1000 ma supply like the 501 is pointless in this situation. I think the Adafruit recommendation for the 501 misses the mark for a newbie like me. The only way around this is to break into the USB cable and supply 5V to the Mega 5V pin (thus bypassing the fuse).

So do you have a cable that will do this and save my 501 purchase? Which is the better 5V supply at 1000 ma: the on-board regulator (makes me go out and buy a 9V power adapter) or the 501 and the break out cable? Remember that my relay coils are 5V and that I want to supply my thermistors with "quiet" 5V.

[adafruit_support_bill]

can you reverse supply it though the 5V regulator - is that even possible?

Possible and advisable are two different things. It is not designed to work that way and (as you have discovered) it doesn't work particularly well.

Which is the better 5V supply at 1000 ma: the on-board regulator ...

The on-board regulator is not capable of supplying 1000ma. The best way to achieve what you are trying to do is either separate supplies for your relays or 100mA directly to the 5v pin.

I think the Adafruit recommendation for the 501 misses the mark for a newbie like me.

I don't see where we recommended this supply for your application. Based on the description of your project I would have suggested these:
http://www.adafruit.com/products/276
http://www.adafruit.com/products/368
and a direct connection to the 5v pin.

[AlWa]

Adafruit_support_Bill, I really appreciate the way you've been bird dogging this post.

Soooo. I decided to try going back to running the whole setup off the 5V pin supplied by the 501 USB power supply - no more vagaries of backward regulator supply of Vin. Got the following:

#501 USB powered
Relays off/4 relays on
Vcc 5.11 / 4.69 (Much better)

So I then said what the h%$l. I picked up somewhere that the USB fuse was a polyfuse (resetable)so I hooked up all 10 relays and got:

Relays off/10 relays on
Vcc 5.11 / 4.30
NO BLOWN FUSE!

Ran all relays on for 15 minutes without a problem. At this voltage, I figure the relay coils were drawing about 620 ma and the digital pins for the relays were drawing 150-200 ma more. The fuse spec must be way off or my Mega manufacturer changed something - it's a knockoff. I tried isolating the relay's 5V from everything else by using seperate 5V pins on the Mega and got .1V better. So now, on to hooking up the logger. Thanks for the help.

[adafruit_support_bill]

I rarely see a polyfuse on an Arduino shut-down. Usually the one in the USB port shuts down first.