Can't get the Peaks firmware update to work - tried all and everything (really!)

For more than three hours I tried to update the Peaks firmware via audio input (in my case: DMC-0.7beta). Unfortunately I can’t get the module to accept the update.
The module starts in update mode and when the audio file comes in it start blinking in a different sequence but somehow it does not seem to accept the data. When the audio file finishes it does not reboot and still has the old FW.

I tried all kind of audio levels. From very, very low linelevel up to +12db modularlevel (near distortion). I tried stereo cables, mono, split etc.
I modified the file in Audacity, raised the volume by 5db, lowered the volume, yadda, yadda, yadda,

I tried to send it from iPad, iPhone and notebook. I used VCA with a gainfactor of 2 etc…

I cannot imagine that the module is defect but after trying all and everything I can think of I get the slight suspicion that there is indeed something wrong with it (?).

Does anybody of you have a final idea?

If it hasn’t worked after trying all that, I suggest that you stop trying with the audio update in case it partially updates and then fails, and you are left with a non-functioning Peaks.

If you really want to install DMC on it, I would suggest getting an STLINK/v2 programmer and cable adaptors for it and flash the firmware directly. But otherwise just enjoy the factory firmware.

Of all the MI modules, Peaks seems to be the hardest to update via the audio bootloader - I also experience frequent failures using this method, even when using signal sources and levels that have worked previously.

Yes, thanks, I already thought about using a programmer for this. I’m trying for more than six hours now and send the file more than 80 times. Very annoying and frustrating. For me this is a defect but I will wait for a confirmation that this might be possible (even I don’t think so as I do have the problem for both of my Peaks modules).

I do already have two different programmers somewhere in my toolboxes. I just have to check if one of them will work.

Anyway, thanks for your help and your great work!

i had a hard time and i eventually found the sweet spot for the vol and it magically worked. Thinking hard of getting a programmer for it too! I wonder if it would be easier out a expert sleeper like audio output versus my macbook headphones.

I use output from my audio interface, which mostly works, but not always. Likewise via my ES-3 which has even higher signal levels - it works about half the time. Other MI modules update via these signal sources 100% of the time. Peaks is especially problematic. All my Peaks behave this way.

Hmm, this really starts to make me angry a little bit (even I’m still a big fan of MI and Olivier).

Found the programmers: PICkit 2, AVRISP mkII and a TipTop ZDSP card programmer.
Aaaarrgghh. Seems I have to order (a hopefully cheap) STLINK/v2 one. I will never do such a terrible and unseccessful day to myself again. :slight_smile:

Compared to the other modules, Peaks is unusual in that it doesn’t have CV inputs. It only has 2 trigger inputs. So the only thing the module can “sense” from the external world is whether the voltage on its input is above or below 0.6V.

You need a fairly high signal level to start with. The signal is already squarish, so amplifying on the software side won’t do anything. Hardware amplification would help - but the amplification should preserve the edges of the signal.

Just of curiosity, how does the update work without audio inputs? How are the bits encoded to trigger pulses?
edit the answer magically appeared above after posting

For Peaks, it’s dumb FSK. The module counts the number of samples (at 48kHz) between signal edges - 8 samples and it’s a one, 4 samples and it’s a zero, 16 samples and it’s a “welcome to the fantasy zone! get ready!”.

It could be that the playback device does not have 48kHz as its native sample rate, and that it tries to resample to 44.1kHz - in which case we can expect some ripples near the signal edges. Can’t think of anything else…

Y E S ! Success!

For the last half hour there was a voice in my head saying: “Crank it up! Crank it up! Crank it up!”

So what I did was putting the output from my iPhone into a VCA with full Gain and +7v CV on the CV-Input. Send the output to my Cwejman MX-4, full level in and out, +7V in the level input.
And… yes, now it worked. The Overload LED on the Mixer was illuminating the whole room. :slight_smile:

So the module really needs a hell of an input level.
P.S.: Just to make sure that it will work the next time as well I tried it three times. Every time it succeeded.

I have been updating my MI modules (4 x Peaks, 3 x Warps, 2 x Tides) from an iPad 1 using its audio playback level near maximum, through a Streams module in envelope mode at 12:00 with the filter fully cw the and input level near fully cw, and it doesn’t usually take more than 2 or 3 tries to find the sweet spot.

I happened upon that method while trying to update my Tides and almost giving up. I’m posting this in case others have a Streams module handy, and are trying to update from an iOS device.

Hi, been a longtime lurker but just signed up today just to say that I had problems too and I solved them routing the audio out from my iPad thru the audio in of my Doepfer A-119 External Input and routing the out to the trig1 in of Peaks. Gain knob should be just shy of overloading, at about two o’clock. Hope this helps someone, and hope I’ll be actively contributing to this community. (I love Braids)

1 Like

THANK-YOU Phonk -you are a life saver !! :slight_smile:
Was on the verge of sending my Peaks off to get looked at and then literally signed up today to this forum . Put the signal through my Doepher A-119 and played around until I had the LEDs flashing/ syncing to the audio signal ( from my MacBook ) and then tried it again after the 10 zillionth time of connecting it . It worked !

1 Like