Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

16F876A possible latchup, suggestions welcomed

Status
Not open for further replies.

foxabilo

Member level 3
Joined
Jul 16, 2007
Messages
67
Helped
13
Reputation
26
Reaction score
1
Trophy points
1,288
Location
Derby, England
Activity points
1,886
โหลดccs mplab plugin

Hi,

Hmmmz, right well first off I hate using these 876A units because of latchup and the one I have in a new ICD2 design seems to have done just that, gets hot, draws alot of Amps and I am dammed if I can reset it.

Any suggstions on how to reset a latched 876A ?

Fox
 

16f876a filter

Sounds like you've got the chip in backwards or wired wrong. They are not prone to latchup.

PS your web site is down.
 

mclr reset 16f876a debug

hehe, thanks for the heads up on the site blueroom , I've just given it a check and it did take over a min to display, I think the host is working on something as my incoming email is also down :(.

Back to the problem, nope definably not wired up backwards or odd, it was working away doing all the things a pic should do, then I power cycled the board and noticed the bench meter showing 200mA load (the max load for the little 7805) and low and behold, yet another 876 with all pins except MCLR, being almost shorts to GND.

I have a bag here of about 30 of the dam things all in the 12F or 16F range all get hot and dont seem to reset, I reckon I need to start putting a 100-150Ohm resistor after the 7805 regulator, incase some sort of a power spike is getting though, no idea how though, its got a whole row of 100u, 10u 1u 0.1u 100n 10n 1n 100pf 10pf 1pf filter caps.

I have just decided to do some whacky tests, the 876F is now in the freezer just on the off chance that a low temp might reset it, I don't hold out much hope though :).

Still if anyone has a sure fire way of getting a pic that's latched to reset and be usable again I'd be greatfull.
 

16f876a designs

Curious, there are 3 versions of the 7805
78L05 100ma
78M05 500ma
7805 1A

Which one is 200ma?

Is it an ICD2 clone? which version, what firmware. I've sold hundreds of my ICD2 kits and they don't even get warm. The 7805 does when you draw more than 200ma from it but it's fine as it uses a heatsink.

The only 16F876A ICD2 clones I've seen are Eric Stolz versions. Most use the 16F877A as it can be upgraded to USB with a 18F4550. Same firmware though.

Erase the PIC, unless you've destroyed it.

PS very cool GPS stuff.
 

latchup pic

:) thanks blue, GPS and GSM/GPRS is my speciality.

The 7805 I am using is the 100mA model and yep, normally it never even gets much over ambient (in regard to the 200mA load I think thats a combination of analogue meters on my bench PSU, poor eyesight :D and a resolution of 50mA per division). Yes I am modifying Eric's design a little, I liked his back to basics approach. I fancied a little project and an IDC2 looked like a good idea, it is very interesting you mentioning the 18F4550 can be a direct drop in replacement for the 877 and far as I know the 876A and 877A both use the same bootloader and if I read what you are saying correctly then I could replace this 876A in Eric's design with an 18F4550 (my favourite PIC of all time) and have usb as well as serial.

Still I think the PIC latched when I disconnected the ground pin to the PIC while (at least I think) the rs232 MCLR 13v line was connected, probably a floating earth and 13 unregulated volts did something weird to the cmos logic and when I powered it back up with no current limiting other than 200mA-250mA worth
 

16f876a replacement

The 18F4550 does not drop in replace the 16F877A. Both are required for a USB ICD2 as the 18F4550 controls the 16F877A in PSP mode.

See the Inchworm+ & Unicorn pages on my site for info.

Without seeing your schematic it's impossible to see the problem. My ICD2 kit is based on Erics also, but I modified the VPP generation.
 

    foxabilo

    Points: 2
    Helpful Answer Positive Rating
16f876a serial programmer

I'm going to give your site a good look through, might place an order or two, excuse the messy schemo, but its just a "something to play with" project,

**broken link removed**

Vpp @~14v is supplied from the bench all others from that.[/img]
 

Well it's a typical stoltz design (op amp version), it could be your serial port, you did turn off the FIFOs and reboot.
 

Got a couple of tubes of 876A's, 877A,s and 4550's arriving this morning, so I'll let you know if the 100R on the Vdd line limiting it to ~50mA solved the latchups.

And yea, I have not reset the FIFO's as yet, that is something I am going to do now actually.
 

Oh it's the target PICs you're latching up. I thought your programmer was latching up.

The 100ohm should help and is recommended when using the opamp version. The 33 should have done the job though. I use the transistor version as it's already current limited.

Also try setting the VPP to 12V or so, anything over 8V generally works with modern PICs.
 

Ahh it is the programmer PIC that's latching :) but in general I have had a bad experience with every PIC but the 18F range, so clearly I am doing something in common with them all causing these permanent latches, I now believe it is down to a combination of several things, firstly my own poor disconnection policy when it comes to PICs after being programmed, i.e just pull it out rather than powering everything down first, and also the fact that until now none of my designs had the 100R current limiter on the Vdd line, that "should" in theory prevent the PIC from ever getting latched to a point where it damages the cmos logic.

Although I realise that a large voltage/current on any pin can cause latchups 99% of my designs are only ever small signal stuff so I can safely rule out large spikes coming from that direction, even so I try to match impedance wherever possible.

I'll let you know how the new 876A does when I solder it in and give it a try.
 

New 876A is working a charm, actually quite amazed worked first time from schematic to board, I don't get that very often. Right now to USB it up, I'll go hunt the forums for the 4550 stuff.

after that I think one of these U40 things is in order. Ohh any by the way blue, your right, I think everyone does just build programmers and debuggers and no actual projects to use them with :)
 

The U40 is basically an ICD1, it's only supported by the CCS C IDE. Your ICD2 clone even the serial one you built is a better programmer. The U40 uses a FT232 USB to RS232 adapter and offers no speed improvements over RS232.

If you want speed and USB look at the Inchworm+ & Unicorn schematics. That's a fast programmer/debugger and it works with MPLAB and CCS has an MPLAB plug in so it works too.
 

    foxabilo

    Points: 2
    Helpful Answer Positive Rating
Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top