Picofly AIO Thread

[abal1000x]

I feel I've missed something. Why we adding resistors to the mosfets

To prevent the mosfet into open state, when the gate floating.

 

[Adran_Marit]

To prevent the mosfet into open state, when the gate floating.

 

[cgtchy0412]

The cons is if the gate cables somehow disconnected (for example by accident), then the resistor also disconnected, means the mosfet is floating.

Thats just mechanical issue its easily ruled out by good soldering, also the unit itself is already fragile so any sane person would not subject it to harsh condition mechanically.
As it is already hard to solder wires on mosfet, and now solder smd resistor on to it also, Hope no more Switch will meet the grave.
Electrically safe is at least good enough.

 

[Interceptor777]

Ok, I installed today with 2.73 FW on a lite switch, classic installation, 1 mosfet, I didn't add resistor on mosfet, I didn't solder sda, scl either.
Everything starts instantly.
So, can I avoid adding?

What kind of wire are you using for the mosfet?

 

[Danook28]

Ok, I installed today with 2.73 FW on a lite switch, classic installation, 1 mosfet, I didn't add resistor on mosfet, I didn't solder sda, scl either.
Everything starts instantly.
So, can I avoid adding?

Befor you install pico in console. Led was blue the yellow 4 time blinking???

 

[hippy dave]

It was discussed a bit further back (edit: oh wait other thread probably), but basically if something got disconnected from/not connected to the mosfet by accident, it could let rip when it's not supposed to and fry something. The extra resistor prevents this.

 

[Adran_Marit]

It was discussed a bit further back (edit: oh wait other thread probably), but basically if something got disconnected from/not connected to the mosfet by accident, it could let rip when it's not supposed to and fry something. The extra resistor prevents this.

cheers, it's quite easy to miss stuff when things update quickly

 

[QuiTim]

cheers, it's quite easy to miss stuff when things update quickly

So, when Pico sends voltage to G(ate) on mosfet the D(rain) and S(source/ground) create a short and current flows D->S.
Once this process (glitching) is done we need to remove voltage from G in order to "close" the mosfet.
This is (i guess) done by Pico via FW but the concern is that G is very susceptible to outside factors and it could get charged "on it's own" from neighbouring flowing current (eddy current?!?!?).
Adding a resistor between G and S would prevent this from happening and we would sleep better
One easier way of doing this was adding this resistor in Pico board as suggested by abal1000x with the downside of this approach being that if the CPU wire in pico gets disconnected then it is of no use (and it could be some time before you notice this if you don't power off you switch very often). Cheers

 

[Furious92]

Wondering if I am able to update or should update picofly v2.4. Toolbox shows that firmware, just don't want to try to flash the update if it will brick it. Literally have had 0 issues so far with the install, boots instantly, so wasn't sure if I should update or not.

 

[cgtchy0412]

I just ordered these mosfet to test. All are Alpha Omega.
AON7400A @ .4$
AON6414A @ .5$
AON36312 @ .8$
AON6512 @ 1$
AON6792 @ 1.1$
Will update next week for the test result.

Just to fulfill my promise.

Not Work/ Not Strong Enough to Glitch
AON7400A, It give false positive so cannot work reliably, one unit returned from customer with systomp can only boot to stock.
AON6414A

Works, Reliable But large package 5x6, Single Mosfet Install
AON S36312
AON 6512
AON 6792
Acctually as it is large the soldering is simpler, but for facedown install you need to cut the shield bridge between apu and ram, or better if you can solder it faceup and just slipped it under the bridge.

Works, Reliable, Perfect 3x3 package, Single Mosfet Install
FDMC 7672

All test were done on Lite&V2
Im now currently using AON S36312 for instalations.

 

[linuxares]

Wondering if I am able to update or should update picofly v2.4. Toolbox shows that firmware, just don't want to try to flash the update if it will brick it. Literally have had 0 issues so far with the install, boots instantly, so wasn't sure if I should update or not.

There is a classic called:"if shit work, don't fix it"

Edit: my Cisco teacher told us that

 

[abal1000x]

Just to fulfill my promise.

Not Work/ Not Strong Enough to Glitch
AON7400A, It give false positive so cannot work reliably, one unit returned from customer with systomp can only boot to stock.
AON6414A

Works, Reliable But large package 5x6, Single Mosfet Install
AON S36312
AON 6512
AON 6792
Acctually as it is large the soldering is simpler, but for facedown install you need to cut the shield bridge between apu and ram, or better if you can solder it faceup and just slipped it under the bridge.

Works, Reliable, Perfect 3x3 package, Single Mosfet Install
FDMC 7672

All test were done on Lite&V2
Im now currently using AON S36312 for instalations.

I am gonna write this to my note, for further analysis.
Thanks for sharing.

 

[Adran_Marit]

There is a classic called:"if shit work, don't fix it"

Edit: my Cisco teacher told us that

same fam

 

[agus]

the normal temperature of the switch after cfw is how many degrees Celsius ?? .. in album

 

[ofuroinu]

On a chinese site they are selling like is as a pack of USD7
Should I get V2 or V1 flex for switch V2 patched?

 

[marhalloweenvt]

Finally, I can get out of my sick bed and back to my Personal R&D Headquarter (my beloved bedroom). During my sick leave, there are lots of thing has changed/updated with Picofly scene.
Good news is I have completed testing cable for Picofly (fab it with JLCPCB and share them among my friends who are able to do hardware mod with their machine). Cable works perfectly. Beside JLCPCB, I already try with oshpark and nearly 2 months, I haven't received them yet

Bad news is I can only support Mariko variant due to not having V1 around. But I am willing to adjust if someone can give me dimension.
Some pictures:

P/s: Tidy something up and will upload them on github
--------
Note1: According to another thread, I will add 1k ohm resistor between Source and Gate pin.

 

[Adran_Marit]

Finally, I can get out of my sick bed and back to my Personal R&D Headquarter (my beloved bedroom). During my sick leave, there are lots of thing has changed/updated with Picofly scene.
Good news is I have completed testing cable for Picofly (fab it with JLCPCB and share them among my friends who are able to do hardware mod with their machine). Cable works perfectly. Beside JLCPCB, I already try with oshpark and nearly 2 months, I haven't received them yet
View attachment 373305
Bad news is I can only support Mariko variant due to not having V1 around. But I am willing to adjust if someone can give me dimension.
Some pictures:
View attachment 373303
View attachment 373304

P/s: Tidy something up and will upload them on github
--------
Note1: According to another thread, I will add 1k ohm resistor between Source and Gate pin.

good work, will be good to get the flexes in combo with the mcfly

 

[abal1000x]

Just to fulfill my promise.

Not Work/ Not Strong Enough to Glitch
AON7400A, It give false positive so cannot work reliably, one unit returned from customer with systomp can only boot to stock.
AON6414A

Works, Reliable But large package 5x6, Single Mosfet Install
AON S36312
AON 6512
AON 6792
Acctually as it is large the soldering is simpler, but for facedown install you need to cut the shield bridge between apu and ram, or better if you can solder it faceup and just slipped it under the bridge.

Works, Reliable, Perfect 3x3 package, Single Mosfet Install
FDMC 7672

All test were done on Lite&V2
Im now currently using AON S36312 for instalations.

Reading this the mystery is gone.

You said before that the 7400 is the best and 6512 is 15sec glitch. Thats confusing me, i postpone my analysis, seems something wrong with my understanding. Now since 6512 is reliable, so the prediction is correct, the smaller rds is better. And the effect of 2 mosfet is smalling the rds (parallel resistance), so become more reliable.

Unreliable
AON7400 16mohm
AON6414 10mohm

Reliable
AON6312 2.5mohm
AON6512 2.4mohm
AON6792 2.5mohm

 

[QuiTim]

On a chinese site they are selling like is as a pack of USD7
Should I get V2 or V1 flex for switch V2 patched?

View attachment 373299
View attachment 373300

Get the V2 and if possible get the one with USB attached so you can update to latest version before installing

 

[cgtchy0412]

Reading this the mystery is gone.

You said before that the 7400 is the best and 6512 is 15sec glitch. Thats confusing me, i postpone my analysis, seems something wrong with my understanding. Now since 6512 is reliable, so the prediction is correct, the smaller rds is better. And the effect of 2 mosfet is smalling the rds (parallel resistance), so become more reliable.

Unreliable
AON7400 16mohm
AON6414 10mohm

Reliable
AON6312 2.5mohm
AON6512 2.4mohm
AON6792 2.5mohm

But thats not explaining why hwfly can still glitch with that high rds, or maybe in the pcb itself have some low voltage circuit for glitching with that high rds.
Im not yet use aon7400 with low voltage wiring.. but maybe it would work.
Any other hwfly board feature that we miss? last time its a pulldown resistor.. so any other?