SPIN FV-1 unknown issue

[knutolai]

Hi folks!

So I've decided to try my hand at the SPIN FV-1, but as per usual when trying new stuff I've ran into some initial trouble.

The SPIN project circuit I've got running: (schematic)
http://i.imgur.com/w2QAfJS.png

I've built the circuit on a fabbed PCB linked in the image below. The circuit has pots for Dry/Wet-mixing, Master Volume, Parameter 1, 2 and 3 for the SPIN and a program selector built around a ATTINY85-p. The 3 LEDs form a binary number (000, 001, 010..) from 0-7 indicating the program currently running. The board has a DIP-switch for toggling between the internal and external programs. The EEPROM is a socketed DIP-package 24LC32A. I burn the .HEX file program-package onto the EEPROM on a separate board which mounts straight onto my PICKit2 (schematic in the link below). When this is done I mount the EEPROM in its socket on my SPIN project board.

My programmer board: (simple schematic)
http://i.imgur.com/4WunJuN.jpg

The problem:
I'm quite unable to get any audio output from the SPIN. Whether I'm running the external or internal programs (with or without a EEPROM in the socket). The OpAmp circuit performs as intended, all the soldering checks out (no short circuits or missing connections), and the voltages appear normal. I'm unable to check if the crystal is running as I don't possess an oscilloscope, only a DMM. I've included the less obvious voltage readings below:

SPIN pin 1/2 (Inputs) : 1.3v midpoint, Reacts to guitar input (voltage fluctuations)
SPIN pin 27 (Output) : 1.55v, does not react to guitar input
SPIN pin 10 (crystal X1) : 1v
SPIN pin 9 (crystal X2): 2.5v

Possible causes:
-As I don't have any past experience with crystals and the FV-1 chip so I'm uncertain how fragile they are to heating when soldering. I do a lot of soldering and consider myself quite skilled. The soldering was done by hand (used flux for the SPIN) at 375 degree celsius. Neither component suffered extensive heating. I handled the SPIN with great care. I think it's unlikely that the SPIN has been damaged by static electricity.

-The first time I wrote a .hex file to a EEPROM I had the PICKit2 set to 5V. I then proceeded to place it in it's socket (on my SPIN project board) and applied power to the circuit. In the "eTap2hw"-project it's stated that the PICKit2 should be set to 3.7v. After this I used a new EEPROM, wrote the .hex file at 3.7v, but got the same result. Could this have caused damage to the SPIN?

-The .spn projects I've written to the EEPROM has all been example files where output is sent to the right output (dacr). I think it's unlikely that the issue is a software-issue, as running the internal and external programs yield the same results.

Are there any tests I can perform to determine if the crystal or SPIN chip are damaged?

Is 22p too small a crystal capacitor? (it's within the range described in the FV-1 datasheet)

Should, for some reason, the EXT/INT resistor be larger?

Any and all help and observations are much appreciated! I'm eager to start writing patches and I'm sitting with a feeling there's something very obvious I've missed.

[Ice-9]

If you have any pictures of the assembled PCB that would help in diagnosing any possible problems here.

If you used flux for the FV-1 chip did you clean the residue after soldering, if you used water based flux this can cause problems with the FV-1 starting and does need a clean.
It's also possible that the crystal is at fault. Some crystals may need the capacitor size adjusted to start properly. 22p 'should' be fine but maybe try a 47p.

The EEPROM is fine to program at 5v when it is connected to the Pickit2 directly, the 3.7v reference on the eTap2 project refers to programming the EEPROM that is on the circuit board. As the FV-1 is a 3.3v chip, trying to program the EEPROM at 5V will put 5v on the FV-1 chip and possibly cause damage. I repeat though, if you are removing the EEPROM to program directly using the Pickit2 then 5v is ok in this situation.

Did you fully test the analogue part of your circuit was working before fitting the FV-1 chip ?

[GuitarPhil]

Can you hear anything when your Mix pot is set to full 'dry' signal? If not then you've got a bad connection or a dead op-amp!

However, the fact you have the input buffer connected directly to the top end of your mix pot is going to put a DC bias of 4.5V on that pot which may be a problem but maybe I'm talking rubbish!

If you can break the connection to the FV-1 input pins (e.g. lift one end of R5) and jumper the input op-amp output to the FV-1 R-out you should be able to tell if you audio path is working. If you had fitted the clip LED you could perhaps have checked if it flashes when you hit a big E chord hard.

I think the Int/Ext resistor value is ok.

Unfortunately probing the crystal pins will just stop the crystal from oscillating.

Phil.

[Ice-9]

To Add to what GuitarPhil has said.
1. your input op amp gain control pot is connected in the feedback loop with the wiper going to ground ??
2. your output op amp doesn't have a good bias voltage and as mentioned is not correctly DC de coupled. Have you checked that the analogue is actually working as it should be ?

[knutolai]

Picture of the board:
http://i.imgur.com/vVjdZ1U.jpg

Some things to note:
-I've created a connection between IN_L and IN_R on the SPIN with some excess solder. On the original PCB the trace only connected to IN_L. I was uncertain if IN_L only communicated with OUT_L in the internal programs so I soldered them together.
-The TO-92, diode and resistor in the upper left corner are part of the bypass LED circuit. It works as intended. I realize I shouldn't have excluded this from the schematic in the original post (wanted to make reading it more straightforward).
-The jumpers in the upper left and right corners are there to bypass two functions affiliated with the SPIN pots that I decided to discard. The voltages for the 3 pots all check out (adjustable from 0v to 3.3v).

If you used flux for the FV-1 chip did you clean the residue after soldering, if you used water based flux this can cause problems with the FV-1 starting and does need a clean.
It's also possible that the crystal is at fault. Some crystals may need the capacitor size adjusted to start properly. 22p 'should' be fine but maybe try a 47p.

I cleaned off the residue after soldering. I went over again today for good measure, but the result was the same. I've also replaced the 22p with a 47p cap. This didn't have an effect either. Still much thanks for the suggestions!

Did you fully test the analogue part of your circuit was working before fitting the FV-1 chip ?

I did all the soldering in one go and tested the circuit afterwards. I haven't had to go over any of the soldering again after that first session, apart from soldering the SPIN inputs together and swapping the 22p to a 47p cap. The analog part of the circuit works. I know the mix circuit isn't perfect with the biasing, but it seems to work as intended. At fully ccw I get clean sound. As I turn it cw the volume decreases. One thing I forgot to mention: when at the cw extreme (all wet) if I strum really hard I'm able to create these very gated, clipped grains of sound ("pops"/Impulse responses) as if there's some kind of harsh clipping going on. Any idea what this could mean?

I think the Int/Ext resistor value is ok.

Unfortunately probing the crystal pins will just stop the crystal from oscillating.

Thank you! Good to know.

To Add to what GuitarPhil has said.
1. your input op amp gain control pot is connected in the feedback loop with the wiper going to ground ??
2. your output op amp doesn't have a good bias voltage and as mentioned is not correctly DC de coupled. Have you checked that the analogue is actually working as it should be ?

That's bad drawing on my end. What you see is a ground symbol and a regular resistor drawn too close to each other. There's no input gain control. The schematic below should clear this up:
http://i.imgur.com/xenSRzz.png

While not perfect the output mixer gets it's bias voltage from the input buffer (pin 7). Something I could try is leaving the mix pot at the cw extreme, break the "pot lug 1"-to-"TL072 pin 7" connection and wire it directly to +5VA. This way the dry/wet mixer would be turned into a bias resistor for the output opamp, which would output only the sound from the SPIN (if the circuit had been working as intended). That procedure will have to wait until tomorrow however.

[Ice-9]

Something I could try is leaving the mix pot at the cw extreme, break the "pot lug 1"-to-"TL072 pin 7" connection and wire it directly to +5VA. This way the dry/wet mixer would be turned into a bias resistor for the output opamp, which would output only the sound from the SPIN (if the circuit had been working as intended). That procedure will have to wait until tomorrow however.

that would certainly bias the output op amp better for testing your circuit.

I can't view your picture.

[slacker]

It's probably not the problem but, if you can, try removing the Attiny and temporarily jumper the select pins to ground, just to rule that out.

[knutolai]

I can't view your picture.

Should be fixed now. Here it is too:
http://i.imgur.com/vVjdZ1U.jpg

Dont have a screwdriver handy atm, I could add a picture of the flip side tomorrow too.

It's probably not the problem but, if you can, try removing the Attiny and temporarily jumper the select pins to ground, just to rule that out.

Nice suggestion. I'll do that and report my findings in the post tomorrow. I'm not ruling any issue out!

[knutolai]

Here are the actions, results and observations from today:

I've turned the MIX pot into a bias resistor for the SPIN output (right output) so that the output OpAmp "sees" only the output from the SPIN chip. Went over the voltages, all power and bias voltages read as expected. When playing the guitar I read AC voltage activity on the SPIN input pins. The pedal now outputs total silence. I no longer get the gated, clipped grains of sound I described earlier. I suspect these sounds were due to the not-perfect dry/wet mixer circuit.

Next up was replacing the ATTINY85-p with ground connections for S0, S1 and S2. Again the voltages read as they should. 0v on all three pins. Still no sound. Tried both the internal and external program, with and without the EEPROM attached).

Lastly I've gone over all the soldering again, even though they all looked fine originally. Still just silence on the output.

Would I be getting the "correct" voltage readings on a burned out SPIN chip? Would the chip be hot?

I feel keen to give up this board and start fresh with a new PCB. Better buffer circuit. Perhaps have the SPIN mounted on its own daughterboard that connects to the main board via header pins. Any recommendations?

[Ice-9]

The FV-1 should not get hot. Were you asking this or IS it getting hot ?

I have not checked your FV-1 tracing as you have drawn it out in a different layout for pin numbering which while making it easier for you to draw the schematic, it makes it a more of a chore for me to read.

You could temp connect the clipping diode to pin 5 to see if this is lighting up.

The two things i would check at this tage is the crystal, it just might not be starting up, second I would remove th FV-1 and jumper pin 1,2 and 27 together, this would make certain your modified audio circuit is/isn't working.

Also make sure the 3.3v regulator is working.

One last thing, where did you buy the FV-1 and what is the date marking on the chip?.

[knutolai]

Currently on my phone at work. Won't be able to answer everything.

The SPIN is not hot. It's bought from Profusion who are listed as a distributor on the spinsemi site. I'm sorry about the confusing schematic. Thank you for helping me out so much!

I have two spare crystals (bought all three from taydaelectronics). I could try substituting it with one of them. The way I understand it the crystals are not polarized?

The regulator outputs 3.3v. Seems to be working fine. As the schematic shows the regulator also powered the attiny85-p which operated as intended (now swapped with ground connections for S0, S1 S2).

I'll provide the other data you requested and attempt the additional tests you suggested tomorrow. I guess getting the clipping diode to light would rule out the crystal as the issue? The chip would have to be running for the clipping diode to light up correct?

[Ice-9]

I guess getting the clipping diode to light would rule out the crystal as the issue? The chip would have to be running for the clipping diode to light up correct?

The opposite, I'm pretty sure that I have forgotten to solder in a crystal before and in this situation the clip led is permanently on. I will double check that result tomorrow as I am building some FV-1 based pedals tomorrow.

[knutolai]

The opposite, I'm pretty sure that I have forgotten to solder in a crystal before and in this situation the clip led is permanently on. I will double check that result tomorrow as I am building some FV-1 based pedals tomorrow.

Thank that would be great! So if this is the case I could use the clipping LED to diagnose:

-Constant light = SPIN working, crystal not oscillating
-Light only when inputting a signal above 3.3Vpp = SPIN working, crystal oscillating
-No light whatsoever = SPIN might be broken or for some reason not running

The SPIN has the following text:
SpinSemiconductor
FV-1
1545A

Looking forward to hear your findings in regards to the clipping LED! I'll have a go at it again tomorrow and report my findings. Fingers crossed!

[Ice-9]

The LED results don't necessarily mean any of that but I will check if I was right about it saying on if the crystal does not start. Usually if all is working fine the LED lights when power is applied and after a second or two goes off.

1545A I believe the code means the chip was manufactured week 45 of 2015. Profusion is the correct European supplier, I only asked as there had been a problem with someone having a non working chip on the spin forum and it seems to have come from china and the manufacture date was something like 2008 so it suspect a possible reject or fake chip if my memory is correct.

[knutolai]

The LED results don't necessarily mean any of that but I will check if I was right about it saying on if the crystal does not start. Usually if all is working fine the LED lights when power is applied and after a second or two goes off.

I see. I've connected the LED as described in the FV-1 datasheet "Typical Application"-schematic. I get a constant light from the moment I power up the circuit. I've swapped the crystal, but the LED behaves the same. Constant light. The voltage on the CLIP pin (pin 5) is 0,25V.

I've also measured the current consumption of the circuit. On startup its 70mA. This decreases gradualy and stabilizes at ~52mA. The reading was done without the EEPROM attached, and with the SPIN in INTERNAL mode.

I'll delay with the desoldering of the SPIN til I get your clipping LED findings.

EDIT: Minor detail. Forgot to read the voltage of pin 3 (MID). It reads as 1.61V. Is this close enough to 1.65V? I'm guessing it is.

[knutolai]

Another update. The FV-1 datasheet suggests that some crystals need a cap to ground on both legs of the crystal. I've now tried 15p, 22p and 47p (independently) to ground on X2, and another setup with 15p caps on both X1 and X2 to gnd. Still getting a constant light on the LED.

I also found this rather discouraging thread over at the spinsemi forum: (Similar clip diode issue, unresolved)
http://www.spinsemi.com/forum/viewtopic.php?t=564

My SPIN chips were sent in an anti-static bag though.

[Digital Larry]

I recall the cloudscapes board mystery and think it was ultimately found to be residual flux, although it might be worth asking to be sure.

[Ice-9]

I have tested the crystal/led idea today.

1. pedal works 100% fine to start with.

2. with removed crystal  the LED is always ON, so that can indicate a state of non startup.

3. with fully working pedal but a BLANK EEPROM (pedal in ext eeprom setting) LED is also always ON.

I would try another xtal (not the same make).

[knutolai]

Good to know! Could you recommend me a source for the crystal?

[knutolai]

Did a new try with a different board and I now have a working setup! Thank you for all the help and suggestions. It's still unclear what was the problem on the first board. The schematic related to the SPIN FV-1 is identical in the new setup. A quick summary for future readers with the same problem:

-Source your FV-1 and crystal from reputable distributors
-Take good care to clean off all solder flux
-When designing your PCB its a good idea to include the SPIN clipping LED circuit as a diagnostic tool