MUFF WIGGLER Forum Index
 FAQ & Terms of UseFAQ & Terms Of Use   Wiggler RadioMW Radio   Muff Wiggler TwitterTwitter   Support the site @ PatreonPatreon 
 SearchSearch   RegisterSign up   Log inLog in 
WIGGLING 'LITE' IN GUEST MODE

OBX voice card (with non-custom CEM3310 equivalent circuits)
MUFF WIGGLER Forum Index -> Music Tech DIY Goto page Previous  1, 2, 3 ... , 67, 68, 69  Next [all]
Author OBX voice card (with non-custom CEM3310 equivalent circuits)
Dimitree
the filter seems just like the one on the SEM, but the SEM uses jfet buffer, while this one opamp buffers.
Maybe this is the reason for the lockup of the filter?
jimfowler
If I had it to do over again I'd put the fine tune AND pan pots on a panel.
JanneI
Done and calibrated! nanners Guinness ftw!

oberling
JanneI wrote:
Done and calibrated! nanners Guinness ftw!

congratulations! And it's looking very nice with those knobs!
JanneI
Here's the mod I did for positive 15v regs. Bend 7815 middle leg upwards and place a 1n400x vertically.

jvq
JanneI, I like how you did the 100uF caps on the voice board. I used a different arrangement (both caps flat, pointing to the audio output Molex header) and while it fits, now I have the problem that I bend/damage the pan pots of the next voice when I pull a voice out. The 50k pan trimmer gets caught behind the 100uF cap. With you solution, the 100uF cap is out of the way of the pan trimmer.

What's the voltage regulator mod for? I looked twice on the previous pages but I see no other mention of it.
JanneI
jvq wrote:
What's the voltage regulator mod for? I looked twice on the previous pages but I see no other mention of it.


I'm no expert, but...sometimes while powering up the 7815/7915 combo, the 7815 puts out -0,6v instead of +15v..
jvq
Sounds familiar from reading the 7805 datasheet.

https://www.sparkfun.com/datasheets/Components/LM7805.pdf page 8 figure 5

Quote:
In many cases, a regulator powers a load that is not connected to ground but, instead, is connected to a voltage
source of opposite polarity (e.g., operational amplifiers, level-shifting circuits, etc.). In these cases, a clamp
diode should be connected to the regulator output as shown in Figure 6.


Only in your case it's not a clamp but a series diode that chews 0.6V off the +15V rail.

(I have no clue what I'm talking about, I just read data sheets and remember things.)
oberling
this is what I'm currently thinking about a 19" front panel.



that means a cpu motherboard is missing and I'll have to implement it. I plan on expanding my MIDI-CV-Unit with some oberheim style sample and hold circuits.
The 5.6V-problem might have vanished with modern components (100 Ohm inner restistance of a modern 4051 in combination with 100 kOhm potentiometers should not be that much range limiting to make the squashing noticable I think... but I'll have to experiment with that).

I'm also not quite sure about the filter expansion board yet. Having read the [/img]crowbx eurorack thread I don't see why there is 4 modes: a low pass mode, a high pass mode, a bandpass mode and a variable notch mode. I thought having 2 modes (the variable notch and bandpass) would suffice since the notch being variable should cover high pass (fully clock wise) and low pass (fully counter clock wise) already.
... Must investigate further
jimfowler
That looks really great. I'm keen to see how it turns out.
ultravox
oberling wrote:
this is what I'm currently thinking about a 19" front panel.




Hey, that's a great looking front panel. I'd be interested in getting one of those.

Good luck mate!
jvq
I'm close to finishing my build but I've hit a bump while tuning the last voice card. VCO1 tunes fine but VCO2 is stuck at 925 Hz no matter what pitch CV I send it, and it also doesn't move when I turn any of the tune pots on the control panel. Comparing it with VCO1 the waveform looks just fine on the scope, and I can also adjust the initial pulse width. It just seems that somewhere 'early on' something goes wrong with the pitch. Unfortunately it looks like the signals in that area are very weak, and my scope only goes to 1mV.

Anybody here got any hints of what I could look for? I already tried reflowing all the solder joints because that has bit me in the past. I used cheap multi-turn trimmers but everything else is supposed to be good quality. Edit: if I measure between the pins of the soldered-in trimmers I see comparable values in VCO1 and VCO2 so I doubt that this is caused by bad trimmers.

One thing I discovered by accident which may or may not be relevant: when the voice is disconnected and powered down, the resistance between the pin 2 of T4 trimmer (INIT FREQ) and pin 2 of A1 is about 80K (VCO1, which works). While the resistance between pin 2 of T1 and pin 2 of A2 is 160K (VCO2, the bad one).

I don't know if this difference is expected or not. I know that VCO1 and VCO2 are not identical, and that they are linked asymmetrically.

Edit 2: I suppose I should pull out one of the three voice cards that are working, and compare passive measurements with VCO2 on that one. That should at least answer the 80K-vs-160K question.
oberling
jvq wrote:
I'm close to finishing my build but I've hit a bump while tuning the last voice card. VCO1 tunes fine but VCO2 is stuck at 925 Hz no matter what pitch CV I send it, and it also doesn't move when I turn any of the tune pots on the control panel. Comparing it with VCO1 the waveform looks just fine on the scope, and I can also adjust the initial pulse width. It just seems that somewhere 'early on' something goes wrong with the pitch. Unfortunately it looks like the signals in that area are very weak, and my scope only goes to 1mV.

Anybody here got any hints of what I could look for? I already tried reflowing all the solder joints because that has bit me in the past. I used cheap multi-turn trimmers but everything else is supposed to be good quality. Edit: if I measure between the pins of the soldered-in trimmers I see comparable values in VCO1 and VCO2 so I doubt that this is caused by bad trimmers.

One thing I discovered by accident which may or may not be relevant: when the voice is disconnected and powered down, the resistance between the pin 2 of T4 trimmer (INIT FREQ) and pin 2 of A1 is about 80K (VCO1, which works). While the resistance between pin 2 of T1 and pin 2 of A2 is 160K (VCO2, the bad one).

I don't know if this difference is expected or not. I know that VCO1 and VCO2 are not identical, and that they are linked asymmetrically.

Edit 2: I suppose I should pull out one of the three voice cards that are working, and compare passive measurements with VCO2 on that one. That should at least answer the 80K-vs-160K question.

As a first step I would try to swap a known-working voice card including its slot-adapter with the quirky one to make sure the signal goes through to the slot-adapter. If that doesn't work (the former known working is now quirky and the former quirky one now can be calibrated) you have a problem on the carrier board. This is however less likely to be the case since on the carrierboard there is mainly the mixing circuit iirc.

Second I would try swapping the slot adapter boards with a known-working one since the portamento-circuit has a direct impact on the pitch cv. If that does the trick you'll have to examine each and every component and trace on the slot adapters portamento circuit to see where you lose your voltage.

If that all didn't do it I would be sitting down and measuring the components in C1_KEYCV path on the voice card according to the schematics and in comparison with a known-working voice when not powered on.
JanneI
Just guessing....If all above don't help..Could it be the 1k smd tempco? Bad solder joint? Can it be measured from the LS318P pin2 and ground (while unpowered)?
jvq
Thanks oberling and JanneI!

Quote:
As a first step I would try to swap a known-working voice card including its slot-adapter with the quirky one to make sure the signal goes through to the slot-adapter.


I was able to tune VCO1 of this voice just fine so I don't think there is a problem in the slot adapter or the carrier board. If there was such a problem, it would affect VCO1 too wouldn't it?

Quote:
measuring the components in C1_KEYCV path on the voice card according to the schematics and in comparison with a known-working voice when not powered on.


It's a very short path. I can see good values on the pitch CV coming in to the voice card. From there pitch CV goes to only three places: A1A and A2A (LT1013's of VCO1 and VCO2 respectively) and A7 (which eventually leads to the VCF frequency for key tracking).

Before pitch CV reaches either of the LT1013's it passes through a 100K 0.1% resistor. I might be reading this wrong but what I think this does is to bring the signal down from the volt level as used by V/Oct CV (e.g. 2V) to the microvolt level (2V/100,000 = 20uV ?). This is where my multimeter and oscilloscope become useless, neither can distinguish more than 1mV. In short pitch CV looks fine until it drops off the radar into microvolt-land.

Quote:
Can it be measured from the LS318P pin2 and ground (while unpowered)?


Yes I read 1K between pin 2 and ground, unpowered.

I think I'll also try swapping out the LT1013 of VCO2 for another op-amp. It's socketed so I might as well try.
jvq
More concretely, if I put my multimeter on the left side of R1 (pitch VCO2) and I play the keyboard I see voltage change as expected. If I put it on the right side I see some low voltage that does not change at all: microvolt-land. This happens with R2 (pitch VCO1) as well so I think that's just how it is supposed to be.
jvq
Swapping the LT1013's did not change anything. VCO1 good, VCO2 stuk at 925Hz. I'll try swapping carrier board positions and slot adapters next. As we all know pulling voice cards off the carrier board is a bit of an ordeal and I only had a few moments to try stuff yesterday.
oberling
Hi jvq, maybe your LS318 isn't working properly? If that one is socketed you could easily swap it with the one of VCO1.
Otherwise: Pin 8 is between 7,9V and 8,9V for me for a wide range of octaves on both LS318 of VCO1 and VCO1.

As you already did reheat all the solder joints I assume you also thoroughly compared all the color codings of all the resistors in that area with a known-working voice-card? And also checked the Transistors and for their labeling and orientation (especially the two J112 Q7 and Q11)...
jvq
I found the problem! The trace between R21 and R25 got damaged so I had an 'open' there. I discovered this by checking connectivity (unpowered) for the various pins of the LS318, and by comparing with a voice that was working. Easy fix with a piece of resistor leg. w00t

I haven't wired up all the jacks yet but the unit is playable now so I'm taking a break from building to learn how to actually get nice sounds out of an OB-X. No presets to fall back on here. smile



[/img]
jvq
Wondering how others did this: if I tune the VCF so that the absolute lowest cutoff frequency is nice and low, i.e. I can no longer hear the resonant bump but the skirt of the filter is still letting audible sound through, then I cannot 'open' the filter (maximum cutoff with full high frequency timbre) without the turning up the 'filter modulation' (EG depth) pot on the front panel.

I am starting to think this makes sense because the cutoff knob has 5V of range, which means 5 octaves of range for the cutoff, which is not enough to operate from 20-20,000 Hz which is about 10 octaves or 10V.

I think I will just tune my VCF's so that the low is nice and low. Is this normal for the OB-X? Or is this cutoff knob range of 5V specific to the CrowBX?
mattwa1sh
Hi guys--

This is a semi-noob question but here goes: my friend and fellow wiggler Jim Fowler built a crowbx and it's awesome. We're building mine next when the boards come in August, but we were wondering: is it possible to get more resonance from the filter? I know this wouldn't be the same as the actual OBX, but it would be a useful feature to extend the range of the filter into self-resonance. Or even just near-self-resonance.


Best,
Matt
jimfowler
According to CS80.com the latest 4-voice board sets are comprised of 6 boards as opposed to the previous 10. Apparently the slot adapters have been omitted. Can anybody shed some light on this? Reason I ask is that I'm planning on converting my 4-voice to an 8-voice and just want to ensure that the new voice boards will play well with the older boards (once I get an 8-voice backplane, that is) and that everything will fit. I've emailed Scott but no response as of yet.

- Jim
oberling
jimfowler wrote:
According to CS80.com the latest 4-voice board sets are comprised of 6 boards as opposed to the previous 10. Apparently the slot adapters have been omitted. Can anybody shed some light on this?

Maybe he managed to cram all the lag and panning circuitry from the slot adapter on the carrier board - that would eliminate the problem of matching the 7915 chips on each slot adapter for ~equal lag times.
However, omitting the slot adapter would also impose a problem connecting the strangly positioned output pins of the ob-x voice cards to the carrierboard of the crOwBX. Maybe there are now two possible connector solder pads for them on the voice cards?

jimfowler wrote:
Reason I ask is that I'm planning on converting my 4-voice to an 8-voice and just want to ensure that the new voice boards will play well with the older boards (once I get an 8-voice backplane, that is) and that everything will fit.

If my above asumptions are right this should not have changed any connections and thus work well with the older versions.
On the other hand: if he has now two different versions of the voice cards (that could be assumed because he explicitly heads the ob-x owners over to another - not yet existing - item in the catalog) we cannot assume everything works as expected...
As I also already ordered a 4 voice system and apparently on of the last 8 voice backplanes this would be of great interest also for me.

It would also be interesting to know if the multimode filter option and feg->osc2 addition are now already on the voice boards... That would be awesome! screaming goo yo
jean_schön
i'm having trouble with a voice card. A13 (LM13700) keeps dying. works for a while when i switch it for a new one, then no signal on the inputs. where should i begin to look for the cause? my other voice cards work fine...
jean_schön
jean_schön wrote:
i'm having trouble with a voice card. A13 (LM13700) keeps dying. works for a while when i switch it for a new one, then no signal on the inputs. where should i begin to look for the cause? my other voice cards work fine...


might not be that. the card works, then in the middle of calibration i try to reset trimmer T8 to the middle, and thats when the signal stops coming into A13. any ideas what might be happening?
MUFF WIGGLER Forum Index -> Music Tech DIY Goto page Previous  1, 2, 3 ... , 67, 68, 69  Next [all]
Page 68 of 69
Powered by phpBB © phpBB Group