Sounds like the left side usb power switch has half-blown, or the diode mod next to it needs some rework.
Either way they'll need a lab fix. If thats something you'd want to do, I could share instructions. Otherwise, best to contact Parallax support and see about returning it for diagnosis. I'll send a headsup to the engineer that'll handle it.
Send me the instructions anyway. I'll have a look and decide if I want to dig in that far or send it in.
The IC is U501 on the schematic, Diodes Inc : AP22802AW5-7
The other 2 parts will look different on your board, as there was a mod made at that place. And that might be what has failed.
Pretty small parts- you might need a decent magnifier to inspect and probe them
The Zener is D504 on the schematic, 3V zener voltage, ON Semi : MM3Z3V0ST1G
That Zener has a 10K pulldown resistor. And there's a tiny cap under there too.
As you get 2V or so, my hunch would be the Zener is not doing it's job, and the U501 is half-switching on when it shouldn't be. If you pulled a high current load when you measure 2V, you'd probably smoke the U501. But that seems not yet to have happened based on your readings, so I'd look for signs of an issue with the mod first.
My P2_ES board appears to be suffering the same issue.
Trying to power the board from the P2-USB socket drags 5V down to 0.5V.
Board works fine from PC-USB but needs boosting as 5V is loaded down to 4.7V
My P2_ES board appears to be suffering the same issue.
Trying to power the board from the P2-USB socket drags 5V down to 0.5V.
Board works fine from PC-USB but needs boosting as 5V is loaded down to 4.7V
This sounds like the same trouble Evanh's P2 Eval board developed.
Mine only behaved that way when a certain dead VIO pin on the prop2 was jumpered to the VIO switchmode supply. The switchmode started whining and drawing lots of current. Take that jumper off and current draw went back to normal.
Ozpropdev, do you know what is drawing such excessive current?
@ozpropdev
This seems similar, but could be slightly different to the issue DaveJenson reported above.
Has this problem just started? (Not sure when you got your board- if this was working before...?)
Are you able to read the FTDI settings, and confirm that CBUS is configured to "PWREN#" ?
If that setting is good, then either U501 (AP22802AW5-7) has blown, or the mod components have failed
I'd start by metering the connections around the mod to ensure there's no short between the zener anode and gnd (should be 10k).
One other thing... is there any difference in behaviour if you connect either supply only (from power off state), vs connecting one supply first, then adding the other? In terms of 5V level and the ON-LEDs behaviour.
Here's that part of the schematic, including the mod parts D504 and R507.
1. His board is still working fine when in normal "PC USB" (via ftdi) supply mode
2. When AUX power is added, the 5v as measured on header pin drops to around 72mV
3. FT_PROG shows same for his and mine, specifically PWREN# on C2 and C3 of FTDI
4. My anode reads 10040 ohms to ground, his 9560 ohms
5. Just powering from AUX side only results in 0V on the 5v header, rather than 72mV
6. We noticed, when powering from PC USB (ftdi chip side), the 5v header reads 72mV until enumeration takes place, at which time it jumps to 4.97v
Digging a bit deeper, I'm pretty sure its the Aux 22802 switch that has failed. Despite the slight differences on 10k nominal, the EN pin of the 22802 is still getting 2.82v on OzProp's board, vs 2.88v on mine, both well above the 1.5v threshold on the datasheet.
Looking at the other end of the zener shows 4.91 v from the AUX USB, and we know there isn't something permanently loading the 5v Common because its fine when PC USB supplies power.
I can't think of what else it can be other than that 22802 switch. Doesn't look too hard to replace, we just need to understand the 'stack up' of the mod, because I'd just replace everything in that corner
@Tubular ,
Perhaps the anode of the Zener mod is not making good connection with the EN pin of the USB switch leaving it floating. Is it worth resoldering that joint before replacing the entire switch or mod? Data sheet says not to leave EN pin floating. I'm wondering if that 22802 is actually off all the time. When powered by AUX-USB port is the 5V rail even able to supply any current at the 0.5V level indicated? Maybe that 0.5V is from some other leakage/diode protection path.
UPDATE: just re-read your post Tubular. If you've already probed directly at the U501 EN pin rather than the solder blob ignore what I have said. Looking at my own board the mod joint looks like a Y where the three components meet and I was thinking you could potentially probe part of it that didn't reach the EN pin if you didn't touch it in the right place.
We've been probing directly on top of that stack, but you're right, its worth resoldering that joint to be sure.
(edit: I checked under a scope, its a really solid soldering connection at that EN junction. No chance contact wasn't being made, but a good suggestion nonetheless)
1. His board is still working fine when in normal "PC USB" (via ftdi) supply mode
2. When AUX power is added, the 5v as measured on header pin drops to around 72mV
3. FT_PROG shows same for his and mine, specifically PWREN# on C2 and C3 of FTDI
4. My anode reads 10040 ohms to ground, his 9560 ohms
5. Just powering from AUX side only results in 0V on the 5v header, rather than 72mV
6. We noticed, when powering from PC USB (ftdi chip side), the 5v header reads 72mV until enumeration takes place, at which time it jumps to 4.97v
Thanks for doing all this Tubular.
3. Correct.
4. Probably 5% resistor. That's fine. Anything above 1K reasonable. The diode pMax is 300mW.
5. Interesting. The USB switch is fully off. Could be the output discharge of the USB switch is activating when you apply current to the OUT pin, pulling down the rail. That's only supposed to trigger for a couple msecs, but with something blown internally could be failing to perform as expected.
6. Correct. PWREN will only enable the USB switch after enumeration.
Here's the mod diagram you asked about.
AFAIK, the actual mod used 10k in-place of the 3k9 shown in this diagram. It's not critical, provided it keeps the diode power within specification when VIN is 6V.
The unmarked 0402 part that the 3k9 (10k) connects to is a cap. Anything 0402 size 1uF or more is fine here.
The red X on the R507 doesn't mean cut. Just remove that part.
Next batch of P2 chips are due in May. There will be an Eval board available pretty soon after those chips arrive. Look out for announcements here early May.
Next batch of P2 chips are due in May. There will be an Eval board available pretty soon after those chips arrive. Look out for announcements here early May.
Awesome, it’s now early May. Any announcements pending? :-)
No Getting Started guide yet, but the Product Guide does give some detail about each of its features. Commenting is enabled, and we invite your feedback.
How robust is the VIO supply on the P2-EVAL boards...?
I briefly touched an earthing post to V3239 for a split second and now my P2-EVAL board appears dead. Would the power regulator chosen for this not be short circuit protected? This was the latest type of revB EVAL board, not the original rev A design.
Update: board is back to life (after leaving it unplugged and to settle for a while, perhaps a polyfuse has saved it?).
I'd accidentally done this several times with my rev A board when attaching scope probes (didn't have the boot fully covering the ground clip) and the red led by the silkscreened '!' flashed rapidly, but the board never seemed to suffer any effects, though I haven't done the same so far with my rev B.
How robust is the VIO supply on the P2-EVAL boards...?
I briefly touched an earthing post to V3239 for a split second and now my P2-EVAL board appears dead. Would the power regulator chosen for this not be short circuit protected? This was the latest type of revB EVAL board, not the original rev A design.
Update: board is back to life (after leaving it unplugged and to settle for a while, perhaps a polyfuse has saved it?).
RevB Eval Boards:
The 1V8 regulator will latch off if a fault condition occurs that it can't withstand. (stuff like dead short for too long, or reverse power). Power-cycle the Eval board to restore operation. (Ie. unplug/reconnect the usb cable).
If a less serious fault (or momentary fault) occurs, the red warning LED will light and the regulator will usually recover without power-cycle. (Stuff like overcurrent)
Comments
Send me the instructions anyway. I'll have a look and decide if I want to dig in that far or send it in.
Schematic here: https://www.parallax.com/downloads/propeller-2-es-eval-board-schematic
The IC is U501 on the schematic, Diodes Inc : AP22802AW5-7
The other 2 parts will look different on your board, as there was a mod made at that place. And that might be what has failed.
Pretty small parts- you might need a decent magnifier to inspect and probe them
The Zener is D504 on the schematic, 3V zener voltage, ON Semi : MM3Z3V0ST1G
That Zener has a 10K pulldown resistor. And there's a tiny cap under there too.
As you get 2V or so, my hunch would be the Zener is not doing it's job, and the U501 is half-switching on when it shouldn't be. If you pulled a high current load when you measure 2V, you'd probably smoke the U501. But that seems not yet to have happened based on your readings, so I'd look for signs of an issue with the mod first.
Hope that helps a bit.
Thanks!
I think that is beyond my eyesight.
I will contact Support and proceed that way.
Trying to power the board from the P2-USB socket drags 5V down to 0.5V.
Board works fine from PC-USB but needs boosting as 5V is loaded down to 4.7V
This sounds like the same trouble Evanh's P2 Eval board developed.
Running at 180MHz with garryj's USB mass storage driver and a custom version of my P2 logic analyzer.
I removed all accessory boards (VGA & USB), made no difference.
I'll keep digging...
@ozpropdev
This seems similar, but could be slightly different to the issue DaveJenson reported above.
Has this problem just started? (Not sure when you got your board- if this was working before...?)
Are you able to read the FTDI settings, and confirm that CBUS is configured to "PWREN#" ?
If that setting is good, then either U501 (AP22802AW5-7) has blown, or the mod components have failed
I'd start by metering the connections around the mod to ensure there's no short between the zener anode and gnd (should be 10k).
One other thing... is there any difference in behaviour if you connect either supply only (from power off state), vs connecting one supply first, then adding the other? In terms of 5V level and the ON-LEDs behaviour.
Here's that part of the schematic, including the mod parts D504 and R507.
1. His board is still working fine when in normal "PC USB" (via ftdi) supply mode
2. When AUX power is added, the 5v as measured on header pin drops to around 72mV
3. FT_PROG shows same for his and mine, specifically PWREN# on C2 and C3 of FTDI
4. My anode reads 10040 ohms to ground, his 9560 ohms
5. Just powering from AUX side only results in 0V on the 5v header, rather than 72mV
6. We noticed, when powering from PC USB (ftdi chip side), the 5v header reads 72mV until enumeration takes place, at which time it jumps to 4.97v
Looking at the other end of the zener shows 4.91 v from the AUX USB, and we know there isn't something permanently loading the 5v Common because its fine when PC USB supplies power.
I can't think of what else it can be other than that 22802 switch. Doesn't look too hard to replace, we just need to understand the 'stack up' of the mod, because I'd just replace everything in that corner
Perhaps the anode of the Zener mod is not making good connection with the EN pin of the USB switch leaving it floating. Is it worth resoldering that joint before replacing the entire switch or mod? Data sheet says not to leave EN pin floating. I'm wondering if that 22802 is actually off all the time. When powered by AUX-USB port is the 5V rail even able to supply any current at the 0.5V level indicated? Maybe that 0.5V is from some other leakage/diode protection path.
UPDATE: just re-read your post Tubular. If you've already probed directly at the U501 EN pin rather than the solder blob ignore what I have said. Looking at my own board the mod joint looks like a Y where the three components meet and I was thinking you could potentially probe part of it that didn't reach the EN pin if you didn't touch it in the right place.
(edit: I checked under a scope, its a really solid soldering connection at that EN junction. No chance contact wasn't being made, but a good suggestion nonetheless)
Thanks for doing all this Tubular.
3. Correct.
4. Probably 5% resistor. That's fine. Anything above 1K reasonable. The diode pMax is 300mW.
5. Interesting. The USB switch is fully off. Could be the output discharge of the USB switch is activating when you apply current to the OUT pin, pulling down the rail. That's only supposed to trigger for a couple msecs, but with something blown internally could be failing to perform as expected.
6. Correct. PWREN will only enable the USB switch after enumeration.
Here's the mod diagram you asked about.
AFAIK, the actual mod used 10k in-place of the 3k9 shown in this diagram. It's not critical, provided it keeps the diode power within specification when VIN is 6V.
The unmarked 0402 part that the 3k9 (10k) connects to is a cap. Anything 0402 size 1uF or more is fine here.
The red X on the R507 doesn't mean cut. Just remove that part.
Is there any estimated timeline on a future p2 board? I'd love to play around with the p2...
Frank.
Bean
Thanks,
Michael
I briefly touched an earthing post to V3239 for a split second and now my P2-EVAL board appears dead. Would the power regulator chosen for this not be short circuit protected? This was the latest type of revB EVAL board, not the original rev A design.
Update: board is back to life (after leaving it unplugged and to settle for a while, perhaps a polyfuse has saved it?).
RevB Eval Boards:
The 1V8 regulator will latch off if a fault condition occurs that it can't withstand. (stuff like dead short for too long, or reverse power). Power-cycle the Eval board to restore operation. (Ie. unplug/reconnect the usb cable).
If a less serious fault (or momentary fault) occurs, the red warning LED will light and the regulator will usually recover without power-cycle. (Stuff like overcurrent)