I recently received 4 full sized USB Host Shield boards from two different vendors. None will work with my current R3 boards. These R3 boards work fine with my previously bought USB Host Shields. My previous host shields are a mix of Versions 1 and 2. All were vendor configured for 5 volt systems.
These new boards all had the voltage and other pads un-soldered. I soldered the voltage pads for 5 volt operation. All voltages were tested and confirmed correct. All shields are mounted directly on the UNOs. My earlier shields have the same voltage pads soldered as I soldered on the new ones. However none of them work with my previously programmed R3 boards.
After testing four of these from different batches from two different vendors I suspect these are not all defective. These all look like ver 2 of the circuits@home schematic, same as many of my earlier shields.
Am I missing something? I'm old but still of sound mind, I believe. I'm a retired EE with some years of microprocessor design and programming. And yes, I can solder and read a schematic.
Suggestions? Ideas? Experience with this issue? Any help will be appreciated.
When you insert the Shield into the Uno, verify all male pins of the Shield mate with the female pins of the Uno. Some male pins have "missed" the mating.
I wish it were that simple. As I had said I've been designing and programming for many years. Two of the four boards worked a few times, About 4 out if a dozen tests. So all pins were cleaned with 800 paper and sockets were cleaned with contact cleaner. Then all pins were checked for continuity between boards and all voltages were verified. That's why I considered cleaning all but after cleaning, no tests were successful.
The white one is from Aliexpress, the blue/green from Amazon, two of each tested. The one labeled Keys is one of many I have been using for years. They all appear identical to me. Can it be that they really are all defective? Thoughts? Thanks, Aaron
Also... usually the Uno has the female pins and the shield have the male pins. Are the male pins on the Uno(s)? (I guess the male pins are on the underside)
Shields often use "passthrough headers". In this case, there are male headers on the bottom of the shield and female headers on the top. This design allows stacking multiple shields.
Just a random thought...
I had once a shield (not this one) that had too short pins ICSP connector pins, making no electrical contact at all.
I had to replace those six pins with longer ones.
On the shield i have that is a 74HC125D, it is not uncommon for 7400 series chips to not have markings.
As said i have a shield like that from Ali, but i have not had the time to do anything with it, though i'd be at least willing to try and see if i can get it to do something useful.
The male part of the through headers look a tad more flimsy than normal male header pins, but i think that could hardly be the cause. Also on mine there is residual flux where they are soldered on, and on the USB hub as well,
Somehow i also doubt that the PCBs have any differences except maybe for the silk layer, but one can never be sure unless by testing every connection.
I remember looking up which solder pads should be connected, and i have bridged the same ones as in the pictures.
I have always been a bit confused about some of the other solder pads, which seem to connect the SPI pins, and which should be connected i guess, but maybe those are just for use with the Mega ?
Anyway, if there is a code that i can upload to my genuine UNO and try and see if mine works, that could be interesting. I have the shield which is pictured on the Right with the B607 silk printing on it.
If you have a 74HC125 next to the ICSP header, it may eventually fail. The chip is functioning as a 5V to 3.3V logic level converter but it does not have 5V tolerant inputs. The 74AHC125 chip does. Two of the dead boards have 74HC125 chips. See message #12 for a link to person who revived a dead board by replacing the 74HC125 with the correct chip.
The other solder pads can be used to bridge the signals around the logic level converters (remove them first) so the board can be used with 3.3V boards such as Due and Zero.
Thanks for your comments and thanks especially for pointing out msg #12. I missed the full implications of it.
It took me a little while to realize that one could use logic gates as voltage level shifters. I retired over 20 years ago and all my EE experience in logic was TTL. No shifting needed.
Further investigation brings up some interesting contradictions: A number of sources agree with you that the 74HC125 is not suitable for a 5v to 3.3v shift as the input pins may not survive.
However the data sheet claims that max VCC is 6.0 and max pin input is VCC for the 74HC125 and 5.5v for the 74HCT125
And tho the 74**AHC125 and 74A**HCT125 shout out "Overvoltage tolerant inputs to 5.5 V," that's lower than the rated 6.0v for the 74HC125. Interesting!
There are a lot of vendors currently offering this same design, (circuits@home) host on both Amazon and Aliexpress, most with photos of the product. On many of them I could read the chip and it was a 74HC125. Not 74**AHC125 as in the schematic. In fact, I saw none with the 74A**HC125.
So that chip may not explain why I have 4 out of 4 dead host boards in this latest purchase. I do expect some duds but my experience for Chinese failues has always been less that 10%.
Sigh and alas! Thanks again for the help and education. Aaron
But look at the link in message #12 because there is a schematic of the
circuits@home board which the cloners copy. The schematic shows one 74HC125 and
one 74AHC125. The clone boards have two 74HC125. I am not an EE so I will stop
here.
But if Vcc is only 3.3v ? That would mean the inputs of the 74HC125 are rated for only 3.3v while the 74AHC125 accepts 5.5v, and that is the point that is being made.
Now i got my DMM out just to verify what the 74HC125 next to the ICSP header is powered with, and it's connected to the 3.3v pin, so i guess i would see if i can find a couple of 74AHC125 's somewhere.
Interesting point! It could be interpreted either way. Input voltage limit could be the same as the max VCC or the same as the VCC you are using. I didn't think of this. Obviously the safe bet is to go with the AHC version when going down to 3.3v from 5v.
If anyone can test a non working shield and confirm that the 5v to 3.3v chip is dead I'd like to hear about it.
However in my case it's an academic exercise. At my age, neither my eyes nor my fingers are able to test or solder or un-solder the tiny SMT chips.
And as I mentioned, all the current vendors are are using only the 74HC125 chips. So is there another option for those of us needing a 5v in, 5v out USB Host? I haven't found one.
