latigid on

midiphy SEQ v4+

475 posts in this topic

This looks sooo cool! Thank you! Any idea on depth of the modules?

Share this post


Link to post
Share on other sites

Thanks, @lukas412! :)

Just measured this quickly and took a photo, as i am sure others will have the same question:

fd925a69b74f11e839e57f6ea1d99671.jpg

The photo is of the new triple-PCB sandwich "A1 Expander" (8x CV out), which has the highest depth of these modules.

You should be good with a 50mm deep case, the module just fits nicely in my Moog 104HP skiff despite its "max depth: 48mm" specs :).

Of course all kudos for these modules fly out to Andy!

Have a great weekend and many greets!
Peter

Share this post


Link to post
Share on other sites

Amaaaaazing! I’m on vacation atm but I will probably order these boards as soon as I get back.

Is there any chance I could convince y’all to release the design files—or even just a super basic drilling guide—for the panels? I have the means to DIY my own panels and some kind of official measurements would make the job quite a bit easier.

Share this post


Link to post
Share on other sites

Maybe we should have a new thread for module things?

Regarding depth, note that it would be doable to squeeze things tighter with shorter headers or even direct soldering. Right-angled IDC headers would also help.

For panel files, at least at the beginning I prefer to keep them closed. We might offer different colours in the future (e.g. black panels). But if you really would like to do your own, just put a note in when you order and I can give you the dimensions. It wouldn't be difficult to measure either :).

 

Share this post


Link to post
Share on other sites

Up and running in its new home.  Big thanks to all those that played a role in making this so easily available to the masses!

 

IMG_2983.jpg

Share this post


Link to post
Share on other sites
28 minutes ago, TheMentat said:

Up and running in its new home.  Big thanks to all those that played a role in making this so easily available to the masses!

#21 !

and Holy Modular Batman! That's all DIY too !!?? Awesome!

Share this post


Link to post
Share on other sites
48 minutes ago, latigid on said:

#21 !

and Holy Modular Batman! That's all DIY too !!?? Awesome!

yeah... My Squarp Pyramid was the only non-DIY piece in the mix, so the SEQ had to take its place!  :decayed:

 

Share this post


Link to post
Share on other sites

I'm about to start soldering the Super Flux LEDs. I'm curious about the circuit driving the LEDs, specifically the supply voltage. I found what I think is the correct data sheet for the LEDs here: 

https://files.elv.com/Assets/Produkte/10/1094/109464/Downloads/109464_superflux_data.pdf

According to this sheet, the typical forward voltage of the green and blue elements is 3.4V, so the supply voltage on this PCB is probably 5V (since 3.3V wouldn't be enough). On the other hand, the red element has a forward voltage of 2.0V. If the LEDs were driven directly by the 74HC595 shift register, whose outputs are very close to the supply voltage, then the current through the LED would be (5 - 2)V / 47ohms = 63.8mA. This is higher than the 50mA maximum current stated on the data sheet for the red LED, and much higher than the 20mA typical current used for most of the LED's specifications. The same is true of the current through the green and blue LEDs, although the difference is not as dramatic: 34mA calculated current vs. the 30mA maximum stated on the sheet.

So, I assume that the LEDs are multiplexed, at the very least, but even so the amount that the red LED's current exceeds the stated maximum is somewhat disconcerting. So maybe the operating voltage is lower than the 5V expected output from the shift register? Is there another voltage drop somewhere?

Thank you for any comments. Of course all of these questions would be answered for me if the schematics were supplied with the kit.

 

Share this post


Link to post
Share on other sites

Not sure if this is a grumbling post about schematics not being supplied? You bought the PCBs, so you are welcome to the schematics if you send an email in reply to your order confirmation. Snippets of the schematics are supplied in part throughout the troubleshooting thread. This is generally a better way to provide targeted help to builders and I'm very responsive in that regard.

So, rant over! :-) If that wasn't your intention, then I'm sorry.

The LEDs are indeed multiplexed as is typical for MIDIbox projects and driven through bipolar transistors. The key datum is actually the peak forward current per chip, listed as 100mA. True, our multiplexing speed and fraction is slower and larger than the given conditions. All of the LEDs have been put through a burn-in test using the same circuit as the one in the SEQ, so we know that they are good. Also, these are superflux LEDs with much higher current-carrying capacity than your typical LEDs.

Have fun building!

Share this post


Link to post
Share on other sites
10 hours ago, latigid on said:

So, rant over! :-) If that wasn't your intention, then I'm sorry.

It wasn't my intention, no, although this is the first time I've heard of this procedure for asking for the schematics in response to an order confirmation, so I will do so.

My concern about the current is partly this: if the software fails for some reason and the multiplexing stops, it is possible that the LEDs receive the full current for some period of time until the power is shut off. Under that circumstance, I'm not sure that relying upon the 100mA current per chip is safe if the individual elements have a lower rating. 

I'm not concerned about burn-in testing showing whether or not an LED operates in a short-term test. I am concerned about the operational lifespan of the LEDs being affected by the current they are carrying, particularly since they're being operated above their stated limits.

Share this post


Link to post
Share on other sites

If you're concerned with the LED lifetime then simply increase the values of the RJ resistors to a current you are comfortable with. I would consider the probability of an LED matrix "freezing" in the way you describe as a low risk though.

 

Share this post


Link to post
Share on other sites
24 minutes ago, latigid on said:

If you're concerned with the LED lifetime then simply increase the values of the RJ resistors to a current you are comfortable with.

Yes, that is my plan. I started thinking about all of this because I'm considering a custom color scheme using a single element for the "A" light and two elements for the "B" light.

25 minutes ago, latigid on said:

 I would consider the probability of an LED matrix "freezing" in the way you describe as a low risk though.

Definitely, except that there is a good chance that I will someday make my own modifications to the firmware... :-)

Thanks for your help!

Share this post


Link to post
Share on other sites

No worries!

I know Peter and a few others ran a mixed colour scheme and it seemed to work out well, though the current drive through the LEDs might get a bit non-linear I think.

Stress test is currently running with the "worst-case" matrix fully on, so no multiplexing. Half an hour in and nothings burned out yet. What do you think is a fair test time?

Share this post


Link to post
Share on other sites

@studio nebula maybe for your peace of mind :) - before we launched the v4+, i tested my first batch of LeMEC boards for a full week, 7x24 hours using only 10R resistors all around! No single LED failed :). The 1/8th duty cycle probably made that happen - and at fairly quick duty cycle rates we probably are looking at average current rather than peak current, if we have a good heat sinking solution, which is given for those superfluxes. The average current would only something like 60mA divided by eight when you use a 47R for red. Nevertheless Andy since then has given me the title equivalent of "mad LED slayer" or such, could not decipher the word he used, hehe :). Note also that brightness did not increase much more when going down from 22R to 10R, the LED is just saturated and the long cycle-off time and the big internal superflux heatsink material will probably keep it from burning up.

I doubt that you can kill it with a 47R even with disabling the matrix. But i think you will be really safe with 100R resistors all around, but i would not worry too much, in doubt the superfluxes can be easily replaced, as they are accessible from the backside. But never had to do it yet.

Have a great new week start!
Many greets, Peter

Share this post


Link to post
Share on other sites

Thanks for your replies, guys. I'm a professional engineer so it's in my nature to first want to understand what I'm building, and second to improve it if possible. When designing hardware I try to be aware of anything which might affect the lifespan of the product. I am sure that the way you guys have the LEDs wired means that they will operate for many years--but possibly after many years these LEDs will no longer be in production and so finding replacements might not be so easy. Nobody knows. Hence I might feel better if I use 100R resistors. Of course the same is true for the mechanical components, which is exactly why you chose the good encoders and tactile switches. Nobody can make hardware that lasts forever, but we can do our best to make it last a long time. I started building my modular synthesizer 13 years ago and I have other synthesizers which are around twice that old. I plan to keep using them for decades.

You guys have a good week, too! Hopefully I will have some time for more soldering. :-)

Share this post


Link to post
Share on other sites

Four hours of stress testing just done. In MIOS Studio "set dout x 1" where 0<x<7 inclusive sets the sink sides active and  "set dout y 0" where 8<y<15 inclusive sets the source sides active. I tested a mixture of red, green and mixed LED colours.

No burn out or overheating of any parts. Colour and brightness after the test are identical to those of a non-stressed board (running in normal multiplexed mode for both).

This doesn't mean that you should run your matrix like this (and if you did you would lose control over LED and switch registration), but it shows that a "failure" with the LED drawing the full theoretical current will not instantly damage the LED. At least there will be enough time for you to notice and reset the unit :-).

 

Share this post


Link to post
Share on other sites

Hah! Thank you for doing that test. It is good to know that I will have time to recover from my own mistakes without damaging anything. :-)

Share this post


Link to post
Share on other sites
On 21.10.2019 at 6:33 AM, studio nebula said:

If the LEDs were driven directly by the 74HC595 shift register, whose outputs are very close to the supply voltage, then the current through the LED would be (5 - 2)V / 47ohms = 63.8mA. This is higher than the 50mA maximum current stated on the data sheet for the red LED, and much higher than the 20mA typical current used for most of the LED's specifications. The same is true of the current through the green and blue LEDs, although the difference is not as dramatic: 34mA calculated current vs. the 30mA maximum stated on the sheet.

This isn't correct, see also the SN74HC595DR datasheet: http://www.ti.com/lit/ds/symlink/sn74hc595.pdf

Output drive current is 6 mA, which means in other words: you've to add an internal resistance of ca. 800 Ohm to the calculation.

This could be doublechecked by clamping an output to ground, and measuring the current draw while the pin is set to logic-1.

Best Regards, Thorsten.

Share this post


Link to post
Share on other sites
Just now, TK. said:

This isn't correct, see also the SN74HC595DR datasheet: http://www.ti.com/lit/ds/symlink/sn74hc595.pdf

Output drive current is 6 mA, which means in other words: you've to add an internal resistance of ca. 800 Ohm to the calculation.

This could be doublechecked by clamping an output to ground, and measuring the current draw while the pin is set to logic-1.

Best Regards, Thorsten.

True as far as I understand the structure of the chips, though I think the maximum current drawn could be 150mA per chip IIRC? Normally if they are overloaded the supply voltage begins to sag.

In this case it's academic though, as we use bipolar driver transistors for both sides of the matrix. So the actual current could be somewhere near the ohmic value.

Share this post


Link to post
Share on other sites

Without a model of the IC an academic discussion isn't possible.
Simple measurements would help to understand the internal circuit.

Best Regards, Thorsten.

Share this post


Link to post
Share on other sites

Not sure if I missed this - but is the maximum length for the Seq->eurorack db25 cable the same as any db25 cable (15 feet) or should it be shorter to avoid issues?

Thanks!

Edited by synaptech

Share this post


Link to post
Share on other sites

Thanks latigid! I'll give the long one a shot once I get the euro modules ordered to see how it performs (hoping to have the seq centrally located in the studio). I'll fall back to the shorter one if it's not up to the task. :)

Share this post


Link to post
Share on other sites

I mean RS-422 is supposed to drive almost a mile, but I'm not sure if the MIDIbox would like that :).

 

1 person likes this

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!


Register a new account

Sign in

Already have an account? Sign in here.


Sign In Now