A new Oakley Sequencer project?

[Synthbuilder]

Ok, so skip = "skip current step" and reset = "reset after current step"?

Yes. The latter seems to me the more logical way of doing it, rather than the usual select the step the reset happens. The usual way gives you a runt pulse on the reset step - a short blip of a pulse that happens just prior to the sequencer realising it needs to reset. I didn't like that so went with the reset after the selected step has done its bit. No runt pulses and everything stays in sync.

Perfect, this is the info that I was curious about, namely if it needed buffering or not.

You could probably get away with just using a diode and 1K resistor in series with the output, and using a 100K pull down to 0V. It'll probably work for most modules needing a gate style input. That said, I'll do a proper buffer module at some point if the module sells in any numbers.

And since both headers are 16-pins, I wasn't sure if the one provided more than just 8x stage outputs.

I was going to put a few more outputs on the EXP1 header but in the end it's just the 8 stage outputs, a bunch of 0V and the internal clock pulse.

Tony

[ericD13]

Is there any builder willing to build one ?
I could DIY but time .. and I don't trust me with surface mount

[Pav]

Hi Synthbuilder ..just assembling the pots and sockets on the pcb prior to fitting the panel and soldering.

trivial question maybe, but the build guide seems to suggest you put the washer and nut onto the socket after the panel is on, then solder.

this is how it is on Oakley 5U but most prebuilt modules in my eurorack do not have an external washer. (might explain why they come undone easily)..

so is it ok to put the washers on the sockets before the panel is on from a clearance /alignment perspective? ie when you put the star washers on the switches.

I assume i do not need both nuts that come with switches. yes?

[Synthbuilder]

... the build guide seems to suggest you put the washer and nut onto the socket after the panel is on, then solder.

Yes this is correct. Secure lightly first and then solder. For the 5U SockX boards I tend to recommend soldering the sockets before tightening them up, although the sockets should still be sitting in their places in the panel.

Washers on the outer surface should always be used where possible. Mostly to stop ugly scratching to the top visible surface of the panel. I have asked Thonk to see if we can get some nice thin flat washers (sometimes called dress washers) for their switches at some point in the future. On the 5U stuff the switches are tightened up from the rear so there is no chance of scuffing to the panel, but with the eurorack designs I have done they have to be tightened from the front.

most prebuilt modules in my eurorack do not have an external washer.

Could be due to panel thickness. With the 2.5mm panels I use in Euro there is enough space thankfully for the thin socket washers.

so is it ok to put the washers on the sockets before the panel is on from a clearance /alignment perspective?

No don't so this. The sockets fit up to the inner surface of the panel. The only things that need washers in the eurorack build on the inside are the switches. That it is because they sit loosely under the panel until tightened with the front nut.

I assume i do not need both nuts that come with switches. yes?

Yes. I have a ever increasingly pile of all the unused nuts from all the switches.

With the smaller Euro modules I have been fitting the switches and pots at the same time and this works well. In the sequencer build I recommend fitting and soldering the pots and switches separately. I may try doing them both at the same time on my next build of the sequencer to see if this is quicker or easier. Taking off that large panel is a bit of a pain.

Tony

[tIB]

Like the look of this - interesting design. I'll be following these 3u developments with a potential banana conversion brewing...

[Pav]

Thanks for the advice...just finished the pots,switches,and push buttons ..I'm just having a break before the LEDs.

I di resort to using masking tape to hold the switches in when turning the board over. Because my first attempt three or four did slip out of holes. I think this was because I didn't tighten the four pot nuts sufficiently heeding the warning to not over tighten. This gave me some additional assurance.

I also have an observation that the on-on switches gave me a little hassle when refitting panel because you need to balance the switch ?knob? In a center position or it snags. If I built a second I might choose to fit 3 on-off-on. Albeit a few pennies more but just to make panel fitting easy.

Edit - green LEDs it is . Decided not to mix red and green

Wish me luck with power up.

[Pav]

ok no luck on power testing main board.

no magic smoke , no excessive heating of ics.
analog 5v test passed.

digital 5v test fails ..i get 11v ish on the lower leg of r30 but upper leg 2.5v

i seem to have 2.7v and 1.0v on the two outer pins of the 78L05 voltage reg u14.

Question would the input voltage be dragged down by too much current draw from elsewhere ?

so far

Found a bridge on U9 removed it, retested, no joy but now U9 getting too hot so the AD712 is fried 4 squid down the drain.
Not sure its related to digi 5v test. May just put TL072acd back until i know whole unit is working

EDIT - pin4 not soldered to ground correctly - the AD712 is ok

5V Reg not getting right input voltage.
diagnosis continues after a break.

[Pav]

I think the AD712 may be ok ..pin4 reflowed and it no longer gets hot.
Maybe red herring.

The voltage regulator U14 is now hotter than id like after 5 minutes
ive used the magic glasses you recommended to me and closely examined the soldering of all other chips for poor soldering but nothing is jumping out.

time to change the regulator..perhaps.

edit - voltage regulator replaced, solved nothing and a sil socket took collateral damage from the soldering iron and fused 3 holes together

im officially stuck , and need some advice pls.

[Synthbuilder]

The voltage regulator U14 is now hotter than id like after 5 minutes.

Is this with just the main board or is the pot board connected too? If the pot board is connected then remove this and measure the digital supply again.

If the digital supply is reading anything less than 4.8V then you have a problem with the +5V supply taking too much current. The problem is probably a shorted output which wants to be a logic one (+5V) but is either shorted to 0V directly, or to another output which is at logic zero (0V). In this case the chip, or chips, with the shorted outputs are probably getting a little warm. Given the complexity of the circuit there are multiple places where a short like this can happen. However, it is most likely to occur near the chip pins, ie. you probably have a solder bridge on two adjacent IC pins somewhere.

The first thing to do is inspect the board again with the power off and under a decent light source. If all is well, turn it back on and check quickly for any over heating chips. All the 74HC chips should not be getting warm at all. If one is then the problem is probably around there. Do not think the chip is at fault as the problem is probably a short and not the chip. Chip failures are rare but chips can be damaged by a shorted output.

If you can't find the fault then check all the output pins of every 74HC chip. Use the schematic as a guide - the output pins are the ones on the right hand side of each of the devices on the schematic. The 74HC outputs should be either high, +5V or so, or low, 0V. Anything that is showing a voltage between 1V and 4V view with suspicion. Use a scope for speed rather than a meter.

Tony

[Synthbuilder]

Two other things come to mind.

Excess current from the +5V rail can be taken when an input on a 74HC chip is allowed to float, ie. the pin is not soldered down properly. Again that chip should be getting warm if this so, and should be easy to track down.

Finally, check that R30 is a 47R resistor. If it is 4K7 then that would give the impression of too much current being taken from the +5V rail.

Tony

[Pav]

Hi,
Confirming All testing so far is main board without pot board.
R30 is 47R checked with meter.

Ill work my way testing outputs as suggested

problem is im seeing all chips where vcc should be 5v are 1v as that is what vreg is measuring. so will outputs ever go high?

/EDIT - I find i have placed U17 the wrong way round - very rookie mistake.

passed digital 5v power check.
next plug in pot board.

[Synthbuilder]

problem is im seeing all chips where vcc should be 5v are 1v as that is what vreg is measuring. so will outputs ever go high?

Good point. A shorted pair of complementary outputs may still go to half supply voltage though. Although, once the supply voltage goes lower than 2V or so, who knows what will happen.

Glad you found it though.

Tony

[Pav]

I had to replace this...


then first test with potboard got this


[video][/video]

I appear to have no Gate B working , and a flickering A+B light suggesting a solder joint needs looking at ..

I havent got my head around whether preload is working or not or how to test..but flicking a step to load then pressing load changes the cv.

if i have steps 1-3 say triggering gate A ..and gate width full...does it become one continuous gate on (or leds appear to ) ? With Gate width shift Im seeing the short on/off pulse on all three steps selected.

Everything else looks like working.

[Pav]

Others had mentioned Seq switch via reset for a 16 step use. Actually the first thing I noticed I may miss is a means of applying skip to channel a or b. You only get both. Can some external module like clock divider help in this scenario?

[Synthbuilder]

Others had mentioned Seq switch via reset for a 16 step use. Actually the first thing I noticed I may miss is a means of applying skip to channel a or b. You only get both. Can some external module like clock divider help in this scenario?

The Oakley Sequencer, like the Moog 960 it was inspired by, is a single eight stage sequencer. There is only one shift register inside so channels A and B behave identically in terms of steps. It has a pair of pots per sequencer step to give the two channels. Each channel can be a different voltage output but can't operate independently from one another.

So, it can be used to control two different synths, but you could also think of channel B as being a way to control another voltage controlled parameter that is part of a single synth voice being controlled by channel A.

You can indeed force it to do 16 maximum steps by using an external sequential switch module to swap between channel's A and B every eight steps. However, nothing can make channel B skip and not A.

What you need is another sequencer.

At some point I am planning for a sequencer add on that will allow some additional cleverness but as yet I haven't decided what it will do.

Just bought a new camera and some video editing software so I hope to be able to do some videos soon of it in operation.

Tony

[Pav]

I just got hit today with a huge bill to get my car through its mot with new radiator and a window winding mechanism...so I will have to put on hold any second build .

Luckily I have a small backlog of Oakley Euro PCbs to keep me busy

[Pav]

gates fixed by U28 reflowed. It was intermittent working when touched with finger and slightly warm. heres another test video. I have been told outside the forum my test patch was not very melodic, so gates test is a tad melodic,
//edit - not much better but shows alternating gateleds flashing.
[video][/video]

..next to calibrate

[Pav]

calibration complete - knobs added and in case - and another wiggle - again oly mobile phone quality and short and sweet.


[video][/video]

[Pav]

Really enjoying patching with this sequencer..I recreated peals of church bells for instance yesterday. It helped to just enable four steps and a slow clock in to see how loading function works and sounds.

I also patched a button module ( addac 305 latches) between Lfo providing clock pulse and Shift In to provide stop and start function , enabling mix of manual and auto stepping. Also helped me in preload mode to stop stepping, change pattern, and adjust frequencies lower as needed then restarting.

For builders, only if putting in a skiff, you might wish to modify the template so text labels for the sockets goes underneath the sockets as you can then see them when leads plugged in when sitting down.

Thought this might help others.

[Synthbuilder]

Put some short audio examples of the sequencer in action on the website:

http://www.oakleysound.com/sequencer-e.htm

Tony

[Pav]

Nice. Tangerine Dream like. My speakers were bouncing.

[Pav]

Tony, Whilst sequencing the new DVCO ...I think i now have a couple of gremlins to resolve on the parallel loading side.

first, the load on shift has stopped working - that was working before.

second i have been suspicious whether the "leave" setting was working from the outset, or if i misunderstand it.
left should mean left in whatever state it is already in.

scenario im seeing:
I have reset on stage 4 to get 4 steps.
I load steps 1 and 3.
I see the leds ping pong between 1+3 and 2+4 and the cv is sum of 1+3 then sum of 2+4.
I set stage 3 to "leave" and press load.
It does a load off instead and i get 1,2,3,4 steps again.

So its not saving the "on state".

[Synthbuilder]

first, the load on shift has stopped working - that was working before.

Does the preload light come on when you press load and then go off when the load is pressed again?

So its not saving the "on state".

Does it do something different when you have the 'load on shift' switch in the up position? If so, it's probably working.

If you have 'load on shift' engaged then the state that is left alone is not the one it is currently on but the one it is about to get on the next shift command. In other words, in leave mode, that step ignores the load command and its state is determined by the previous step's state. So if you have an off in the previous active stage, and shift and load are activated, then that step will now be off.

Now this would seem to be at odds with the description of leave, since this implies that the state of that step be left alone. The latter does indeed happen but only when 'load on shift' is set to off. When 'load on shift' is activated we have both the shift and the internal load command taking place at the same time and shift takes precedent.

Tony

[Pav]

>>Does the preload light come on when you press load and then go off when the load is pressed again?
Yes - regardless of position pre/on shift switch.

I think it is just my misunderstanding again. Plus the switch is acting as if it was upside down
but that is not possible is it ?


I was expecting (Hoping) " load on switch" mode was to allow you to say set Step x load on switch and the next "shift" will action
the load without pressing the Load switch or external Load in pulse .
What i see now is that you always require the load ..and the next shift will always action it.

and that "preload mode" facilitates a preview (so to speak) of a stages LoadOn or Load Off cv state
by pressing Load button on and off without requiring a "shift".
And if you preview two stages LoadOn together, the "leave" setting allows you to then preview Load Off -of one of the two stages

Does that sound right ?

[Synthbuilder]

Plus the switch is upside down.

Unless a switch is marked as such then down is on.

Maybe in a future revision I'll put some links on the board to allow the positioning to be set either way. Overseas users are certainly finding this habit of mine rather confusing.

I was expecting (Hoping) " load on switch" mode was to allow you to say set Step x load on switch and the next "shift" will action
the load without pressing the Load switch or external Load in pulse.

If I understand you right I think this what the load switch and input is doing when the 'load on shift' switch is down. Think of the 'load on shift' switch as a mode switch that controls the action of the load switch and input.

I didn't really think about the 'load on shift' being a performance control but, actually, I have been using it as such. Particularly in regards to forcing the sequence to a set point in between shift pulses.

Does that sound right ?

To be honest I still haven't got my head around the leave mode. I tend to use it as a creator of happy accidents. Often I'm not sure what to expect when I hit load on the fly during a set. In truth the leave mode was just a freebie from using on-off-on toggle switches. If folk find a use for it, then I'm happy.

Tony