Junebug VPP problems

[edeca]

I've touched up a few joints but no difference. My soldering isn't brilliant but everything else seems to work

VPP on the plastic header above the 18F2550 also reads correctly.

Any other ideas?

 

[blueroomelectronics]

Well if VPP is fine (and it appears it is) then test the VPP run (+5) & reset (Gnd) modes (also in troubleshooting)

 

[edeca]

Well if VPP is fine (and it appears it is) then test the VPP run (+5) & reset (Gnd) modes (also in troubleshooting)

Aren't they for PGC and PGD? Both appear fine, GND is around 0.09v and high is around 4.75v (the USB port on the machine I tested with is a little low).

I'll scour the circuit again later.. any other suggestions?

 

[edeca]

What should the resistance be between pin 20 on the 18F2550 and pin 23? Mine shows they are connected.. this shouldn't be possible because I don't think the 2N3904 (Q3) would conduct from collector to base.

Even if an NPN transistor did conduct from C to B.. there should be at least 120K resistance due to R1B, R1D and R7 (assuming the test voltage from the meter gets through L1 and D1). Right?

Please correct me if any of the above seems wrong. I could be completely out on the theory.

I ask because it appears there might be a little defect between the traces on the top layer from pin 20 and pin 23. This is barely visible between the 28 pin socket and the connector at the top of the board.

If there shouldn't be 0 resistance between the two pins, could this possibly be the cause of my problems? I can get out a macro lens and try to take a picture later if the resistance shouldn't be 0.

 

[futz]

What should the resistance be between pin 20 on the 18F2550 and pin 23?

Around 28 megohms on mine.

 

[edeca]

Around 28 megohms on mine.

0.6 on the 200hm: scale, 0 on all other scales. So our readings are different, any thoughts Bill?

 

[3v0]

Mine agrees with Futz.

 

[edeca]

Mine agrees with Futz.

Can you explain how you work out from the schematic? I've obviously got that bit wrong and it would be good to know why

However.. the fact that I get virtually no resistance suggests that there might be a short somewhere. It could be where I think there is a fault on the PCB, I'll try and get a picture.

 

[tunedwolf]

Can you explain how you work out from the schematic? I've obviously got that bit wrong and it would be good to know why

However.. the fact that I get virtually no resistance suggests that there might be a short somewhere. It could be where I think there is a fault on the PCB, I'll try and get a picture.

Remove U1 and measure again without power being applied, if the measured resistance is still a few ohms, I would agree with you in that you have a short on the board. If the resistance is no longer only a few ohms, measure the pins on U1 for a possible short

rgds

 

[3v0]

Can you explain how you work out from the schematic? I've obviously got that bit wrong and it would be good to know why

However.. the fact that I get virtually no resistance suggests that there might be a short somewhere. It could be where I think there is a fault on the PCB, I'll try and get a picture.

He He. You read that prior to me nuking the bit about the schematic. There are so many poeple here who know more about this then me that it makes no sense for me speculate. I am a CS type by training.

But here is what I was thinking
It should be infinite on an unpopulated board. No path between the two pins.

On a populated board with the power off I do not see any reasonable path between the two pins. I can see a possible path from pin 20 to the inductor D1 R7 Q3 collector to base, R18 to pin 23. But the resistance between the collector and the base of the 2n3904 (Q3) is for practical purposes infinite. It is a good spot to look for a short but even if it were you would have more the the .6 Ohm you are seeing because the other resistors in the path.

I do not put much faith in the logic. Any EE should know better. I will take my lumps

Now

 

[edeca]

Remove U1 and measure again without power being applied, if the measured resistance is still a few ohms, I would agree with you in that you have a short on the board. If the resistance is no longer only a few ohms, measure the pins on U1 for a possible short

I should have been clearer. I tested without U1 and without power.

See the attached pictures (sorry for the very narrow depth of field!). It does seem to me the tracks are joined, I'm unsure if that would cause the problems I'm having.

 

[ericgibbs]

hi,
Thats a copper track bridge for sure, use a small scalpel blade to cut the bridge.

 

[edeca]

hi,
Thats a copper track bridge for sure, use a small scalpel blade to cut the bridge.

I'm not sure the pictures do it justice, the gap between those tracks is very small and it's wedged between the two sockets! It will be very difficult to cut it without going through one of the tracks

I'll have to work up the courage to play with knives. I actually think there might be another (not quite so bad) on the bottom of the board too.

 

[blueroomelectronics]

Wow I'll have to check my PCB stock for that bridge, sorry about that I'll have to see what my PCB supplier can do about that. I do leave generous spacing between traces so it must have been something at the PCB shop.
A few cuts with an xacto will fix it, take your time and it should be fine. When I've done it a slow steady cut or two (and a steady hand) will do it. I'm sorry it caused you so much trouble.
Where do you think the other one may be?

 

[edeca]

Where do you think the other one may be?

Under SW5. I've attached a picture, it's cropped so nobody can see more of my awful soldering

I have scraped away the blue covering this time to get a good look.

 

[edeca]

Finally, a post to be happy with.

I made a bit of a mess with the cutting (scored the coating off 1 track) but I got rid of the bridge and didn't cut through either of the proper tracks.

See the photo below, which shows the LED flasher program from the Junebug manual working happily. There appear to be two LEDs on at once due to the slow exposure

However, the photo also shows a missing BS250. I have a replacement here but don't know which lead is which, I'll post a message over in the datasheet request forum if anybody can help. That's the last thing to fix before I start playing with it properly.

Thanks for all the answers, it's great to have it working.

 

[blueroomelectronics]

Glad to hear you found the problem and got it working, I've been examining the Junebug PCBs for bridges and it seems you've got the only one so far. There is a diode in the BS250 across the drain & source, an ohmmeter will spot it.
The new Junebug manual shows the orientation of both BS250 types.

 

[edeca]

The new Junebug manual shows the orientation of both BS250 types.

Thanks Bill. If you could confirm that the one I bought is compatible (180mA not 250mA) that would be great, I'll check for the diode.

 

[blueroomelectronics]

180ma is fine, keep the jumper in place unless you need VDD before VPP programming (only required for some very specialized PIC config modes)

 

[deputy]

Hi,
I seem to be having the same problem as Edeca had. I picked up an assembled Junebug from Creatron last summer (2008). Finally, tried it out recently and when I connect to the PicKit2 software, I get a VPP error. It is the same scenario as Edeca's first post. When I run PICkit2 v2.61 software I immediately get a VPP voltage level error. I can assign an ID to the Junebug, but cannot send a program to it or read what's on the chip. When I run the Troubleshoot option, I'm able to measure proper voltages with a voltmeter. In Troubleshoot Step 2, when I have /MCLR On and then click Test VPP, a short is detected.

Since I purchased it assembled, I'm hesitant to begin desoldering and pulling off components to look for copper cross-overs like Edeca found. There are no visible cross-overs on the bottom of the board at least.

Bill, do you have any particular places you recommend looking for issues?

Thanks, Tom