What voltages do you measure at the opamp pins? (Inputs, outputs and power supplies.)
To go from 20mA to 5V you would have to have a very bad (but not open!) connection in the opamp feedback loop, but you said you had the same thing on all gate outputs? Weird.
Blown opamp ICs? But both of them?
Were the power or chips ever plugged in backwards?
Are the two opamp chips on the right side of the board? (Grasping at straws here.)
Any bent pins on the opamps?
2 Likes
Is that 7805 in the right way around?
1 Like
Good questions! It’s a bit hard to measure the opamps, because they are on the second board below, but I will think of something, maybe I need to remove the panel before…
On the power, I have shrouded headers, so should not be a problem, the opamps are the right way in, but I will check them separately again…
1 Like
Yes, it is a larger one instead of the TO92 version, but I made sure it is the right way in. It gets a little warm, but I think that is expected because it has to shed 7V, right?
I have 5V at the correct pins of the atmega, ground also looks good. I can see some kind of signal at the midi in pin with my scope, not sure if it is the correct one though…
1 Like
One more thing, are you sure the connections between the two boards are good?
Hmm, if the big one gets warm, the little one would probably burn out.
1 Like
I will reflow those and everything else in the path…
But I measured the 20mV directly on the atmega pin, where I expect 5V… I will measure again with the lower board disconnected, but I think it is the same…
1 Like
does the status led flash at all??? as this is before the atmega pin. so that needs to be flashing. if that isn’t it may be before the pin
1 Like
The Status led does flash, that looks good.
Two things are strange: I do not see any output on the atmega (I check pin4 which is gate 1 when I play a sequence on the beatstep, btw: is the atmega listening on channel 1 by default?) the output voltage is always something between 20mV and 70mV.
The second thing is that from this 20–70mV the opamp somehow makes 5V for the output jacksocket!
1 Like
do you have an oscilloscope to hand? or something to see wether the signal is going to pin 2 of the atmega chip this is the rx pin which the midi ends up going to.
also check if pin 1 is getting any voltage, this is the reset pin which should be high at 5v to make the atmega run.
definitely something rather strange going on!
edit. the 7805 shouldn’t get hot. which is a tell tail to something.
as for the defaults its channel 1-4 outputs.
does the bottom trigger (which defaults to a clock out) does that flash when you press play on the beat step pro???
3 Likes
This is the signal directly out of the midi port (not connected to the board).
It is the same as on pins 2 and 3 of the optocoupler
EDIT: I just read that I can not measure the MIDI signal like that! I need to connect a resistor between the two MIDI lines! Sorry, will try that again!
This is the signal as it arrives at pin2 of the atmega:
Pin1 of the atmega (reset) has 5V.
The bottom trigger (clock) is always on, same as all the other gate LEDs…
1 Like
Okai, I removed ICs until the 5V regulator stopped getting warm! I think it stopped when I removed the optocoupler! I put in a new one and now I have a signal at pin4! hope it is my gate xD
I am a bit afraid to reconnect the boards, maybe there is something on the other board which pops everything now, but I will give it a try! 
The hexinverter chip is also still not inserted… one step at a time!
1 Like
Okai, I reconnected the boards and all the LEDs are off… progress I guess? xD
Now when I start a sequence on the beatstep pro, the clock LED flashes very quickly and the LED above the clock LED lights up when I play a C on sequencer 2, but only then! (only for the “standard” C, octaves do not work) This is very weird!!
EDIT:
Okai, I think that is a problem with my beatstep pro… If I use the keystep, at least the gates work as expected and I have high hopes for CV and everything else!
Thanks everyone, you’re THE BEST!
3 Likes
My initial 6N137 was bad. I tried a 138, which seemed to work. Then I popped in another 137 and everything worked. This is why I usually order extras of all but the most expensive components 
3 Likes
Wheeey!!! so in conclusion was it a bad optocoupler? I dont think I have experienced a bad optocoupler before interesting! but at least its on its way to functioning!!! usually when the optocoupler doesn’t work for me for some reason I just put a wire from the midi cable to the destination and just skip the optocoupler haha, most of the time it works fine without one! just sometimes the circuits need isolating
5 Likes
Not sure! The issue is coming back! Very strange! I already resoldered all the pins around the Atmega and optocoupler and this fixes it for the day, but then the next day it’s broken again! I did this twice already and both times I thought, now it it fixed!! But no! I still think it is a bad joint… Annoying!
1 Like
Hmmm. This one is giving me issues. I could use some help.
I’m at 10.6 at the test point. There is 5V at pin 1 of the atmega. When I plug it in, the bottom led lights up briefly. Then no matter what I try no other LEDs lit up. The chain is pins are connected. I tried swapping several ICs. I reflowed a bunch of spots, but the solders look good.
I might try my cheap DIY oscilloscope tomorrow.
Calling it a night.
Here are some photos. Any ideas or tips you have would be much appreciated.
2 Likes
Until the MIDI-in LED flashes when you send MIDI to the board, nothing else will work, so start there.
Check the optocoupler and the 74LS14 and surrounding components.
3 Likes
I’m thinking it’s one of those. I likely will order more of them from a different vendor.
1 Like
Mine did what sounds like the exact same thing. Bad optocoupler.
1 Like