OSSC stays dead when powered up

[zugspitzjockl]

Hi everybody,

I finished assembling my OSSC finally.

Did anybody have success putting the OSSC together using the latest batch of DIY components?

Mine doesn’t do anything, no LCD backlights, no LEDs.

Soldering should not be the problem, I am very experienced soldering parts like these and double checked everything using a 30x microscope.
Supply Rails all look good (they put out the correct voltage).

Maybe I get access to a Terasic USB Blaster to check if it is the flash.

Do you have any other suggestions?

BR
Felix

[marqs]

Sounds like either firmware does not get loaded or there’s a problem with the 27MHz oscillator.

[LazyEpic]

I’m having the exact same problem but my oscillator is putting out 27 Mhz on the nose, guessing they might have sent out chips without pre-flashed firmware?

Guessing that means I have to use the JTAG to flash it?

EDIT:
My kit is also from the last batch of DIY kits

EDIT2:
Gone over every via for continuity according to trace file, all pins for bridges etc, all the regulators for values as well as output when connected, checked every solder-joint, frequency outputs and can find nothing that seems to be off, it just wont react to anything so guessing it’s missing the firmware which is an issue as I don’t have a JTAG programmer at hand.

[AetherSmith]

@zugspitzjockl Are you sure things are okay with the power? At least on my build, the LCD backlight would still light as long as it was getting power, regardless of whether it was getting any commands from the FPGA.

@LazyEpic Is there any behavior from the LEDs?

I was thinking that the chip from my kit was missing its firmware for a while, but it turned out to be an almost imperceptible short to ground on one of the QFPs.

[LazyEpic]

@AetherSmyth There is no led activity, no LCD activity, no backlight. I’ve double and triple checked the power outputs, checked most vias and pins etc for shorts but will go over them again just to triple check them as well. But as we are two people with the exact same issue from the same batch my first guess would be a problem with the batch.

[zugspitzjockl]

Hi,
Thanks for your replies and suggestions.

At the moment I don’t have time to further check my ossc.

What I have already done now:
– re-check all soldering connections
– measured all voltages at the FPGA (did not have my fine Probe tips at hand, so i checked the voltages at the capacitors)
– measured frequency at the oscillator with the multimeter (oscilloscope and fine Probe tips are in the cellar)
– measured all resistors around the Ossi for correct values
– bought a terasic USB blaster, always got an error saying something about JTAG chain, but I never programmed a FPGA before, so maybe I am to stupid to use it

What is the correct behavior of the conf_done pins? Is there anything I can check with a scope?

[LazyEpic]

@ZUGSPITZJOCKL what exactly does the USB blaster say about the JTAG chain? Thinking about getting one to trouble shoot.

EDIT:
What’s your voltages on your JTAG header? Are you powering the board while trying to connect? (you should be*)

Edit2:
Ordered a USB blaster, should be here by Wednesday at the latest so I can double check your findings.

[zugspitzjockl]

I don’t remember the exact message from the USB blaster, but it was always identical, did not change when the board was powered or unpowered.
Also, the light at the USB blaster was lit up when I connected everything. After the error message, the light went out until I replugged the blaster at the USB side.

In the OSSC, there was nothing lighting up at any time, except the LCD backlight when I touched the transistor on the bottom side, the backlight was faintly lighting up. I guess the FPGA tri-states its outputs.

[marqs]

FPGA should be controllable via JTAG regardless of configuration pin connections (except NCONFIG and nCE which need to be pulled high and low respectively). Is your VCCINT voltage within +-5% of 1.2V? There’s been a couple cases with measured VCCINT ~1.35V (possibly due to bad regulator or soldering) instantly killing FPGA.

[LazyEpic]

I wont be getting my USB blaster until tuesdag/wednesday but mu VCCINT measures 1.196 at the FPGA pin so that should be ok, will be trying to go over every pin etc to test for bridges with continuity but while not using a microscope like @zugspitzjockl I do use 6x loupe glasses with ample lighting and cant see anything that looks “off”.

[zugspitzjockl]

VCCINT measures 1.201 Volts.

I will try JTAG connection again and post the error messages.

At the moment I think the FPGA is bad.
What else can I check on the board before I unsolder it?

The FPGA had many bent legs when it arrived, but I straightened them using the microscope. Maybe something went wrong mechanically inside the FPGA, or it overheated during soldering but I don’t think so.

Thanks for your help!

[LazyEpic]

My FPGA had two or three pins that where a little crooked when it arrived that I had to just move a little but nothing severe, we are talking at most half a pin pitch on each pin. What strikes me is that we seem to be the only ones with the issue of not even leds lighting up etc and we seem to have parts from the same batch, granted I’m guessing the DIY scene is fairly small but still raises concerns.

[zugspitzjockl]

So, I was so curious about whats going on here that I dug out the OSSC again.

First i checked nStatus, nConfig, nCE and CONF_DONE, they were all 0V when powered up.

I read about someone not properly soldering the exposed pad which led to proplems with nConfig always being low.
http://www.alteraforum.com/forum/showthread.php?t=49934
The exposed pad is not for heat dissipation btw:
https://www.altera.com/support/support-resources/knowledge-base/solutions/rd12042013_126.html

I resoldered the exposed pad and all (!) pins using my newly arrived Rosin Flux

Then after switching on – the Display lit up and both LEDs are also lit up. But there is no text on the Display and it does not seem to respond to the remote control.

CONF_DONE is still low, nConfig is at 3,3V and nStatus is at 1,685V (high?!?)
This document says that it is stuck during loading configuration:
https://www.altera.com/content/dam/altera-www/global/en_US/pdfs/literature/hb/cfg/cfg_cf51001.pdf

When connecting JTAG, Quartus detects a different chip than expected (JTAG ID code 0x020F20DD expected, found JTAG ID code 0x020F10DD).

The label on the chip I got in the advanced DIY kit says EP4CE10E22C8N, the BOM says EP4CE15E22C8N…

According to table 1-1 on page 3 of this document
https://www.altera.com/en_US/pdfs/literature/hb/cyclone-iv/cyiv-51001.pdf
The CE10 is almost short of everything compared to the CE10.

Was this FPGA sent out accidentally or is it possibly to run the OSSC fw properly on this one?

[LazyEpic]

Also reseated everything today and checked the connections from before, nothing wrong from what I could tell. I got the same Cyclone version EP4CE10E22C8N and not the CE15 version that is listed in the kit.. The wrong chip has been sent out right? …. wish I read that post before I reseated everything.

EDIT:
I don’t get any light on my however

EDIT2:
Checked the pinout, seems they are different on the CE10 vs the CE15 version

PIN    CE15          CE10
1      VCCD_PLL3      IO (B1)
2      GNDA3          IO (B1)
3      VCCA3          IO (B1)
61     IO (B4)        VCCINT

The most obvious ones
https://www.altera.com/content/dam/altera-www/global/en_US/pdfs/literature/dp/cyclone-iv/ep4ce10.pdf

[BuckoA51]

Well that sucks, must have happened when we switched supplier since the original supplier was out of stock.

Can anyone that has an EP4CE10E22C8N mail in with their order number and I’ll send out a EP4CE15E22C8N as soon as they come in.

Sorry for all the inconvenience.

[Author]

Posts