Rampage and Ampmix output -5 to 5 volts instead of 0 to 5

From circuitbending to homebrew stompboxes & synths, keep the DIY spirit alive!

Moderators: Kent, luketeaford, Joe.

Post Reply
User avatar
Nagasaki45
Common Wiggler
Posts: 230
Joined: Thu May 11, 2017 1:21 pm
Location: London

Rampage and Ampmix output -5 to 5 volts instead of 0 to 5

Post by Nagasaki45 » Fri Aug 11, 2017 4:30 am

Hi wigglers!

I need some help in debugging a wierd behaviour with the Befaco Rampage and RYO Ampmix. When I connect any of the rising / failing or B>A outputs to any of the Ampmix channels, and set the gain to x1, I sometimes get -5 to 5 and sometimes 0 to 5 volts output. The output depends on what I'm connecting it to. If I connect the Ampmix channel the mixer or the 1V/oct on my 0-coast I get -5 to 5 volts, but if I connect it to the pitch input of the BIA I get 0 to 5 volts. If this is not wierd enough, when I connect the channel to the decay input of the BIA I get -5 to 5 volts again.

I made a video to demonstrate the issue properly.

Both the Rampage and the Ampmix are my builds, so might be faulty. But this fault is too consistent, and happens only with these outputs of the Rampage. It doesn't happen with the trigger outputs of my BeatStep Pro for example.

Any help to understand the source of the problem and how to fix it will be highly appreciated!

Thanks a lot,
Tom

User avatar
Nagasaki45
Common Wiggler
Posts: 230
Joined: Thu May 11, 2017 1:21 pm
Location: London

Post by Nagasaki45 » Thu Aug 17, 2017 4:04 pm

No one? really?

If the issues is not clear enough please ask and I will be more than happy to elaborate. BTW, I contacted Befaco and RYO a few days ago and didn't get any response yet. I will update here if they reply soon.

User avatar
mskala
Super Deluxe Wiggler
Posts: 2750
Joined: Thu Mar 27, 2014 7:33 am
Location: Toronto
Contact:

Post by mskala » Thu Aug 17, 2017 8:50 pm

Are you able to measure the actual voltages? The video doesn't convince me that it's really -5 to +5; only that your low voltage seems to be negative and your high voltage seems to be positive. If the Ampmix is adding a negative offset, that could explain a lot of the effect you're seeing. If the BIA's pitch input is going into an ADC that only measures down to 0V, that could explain it not seeing a difference between zero and negative voltage.

The first thing to do in debugging anything like this is isolate the problem: if you have some other source of signals instead of the Rampage, does everything behave as you expect? If you have some other mixer instead of the Ampmix (or just patch directly from the Rampage without the Ampmix), does everything behave as you expect? If you feed a known negative voltage from some other source into the BIA, does the pitch input respond to that?

At the moment my guesses are that your Ampmix is adding a negative offset (either by design, since it seems meant to add offsets, or because of a construction error) and your BIA is not meant to accept negative pitch inputs and treats them the same as zero.
North Coast Synthesis Ltd.
Twitch stream (Mondays, 3pm Eastern)

If you tell me that your goal is systemic change toward radical acceptance, and I see that you treat those you perceive as lesser-than with the same kind of scorn and derision that pushed me toward this insular little subculture where I feel comfortable [. . .] then you’ve successfully convinced me that your acceptance is not radical and the change you want not systemic. - "When Nerds Collide"

User avatar
Nagasaki45
Common Wiggler
Posts: 230
Joined: Thu May 11, 2017 1:21 pm
Location: London

Post by Nagasaki45 » Fri Aug 18, 2017 3:20 pm

Thank you so much for your response @mskala!

To answer your questions:

1. Are you able to measure the actual voltages? Not really. I don't have any scope / voltmeter. However, I do think that when the Ampmix is set to unity gain and get low input from one of the Rampage gate outs it outputs -5 volts. I did the following experiment: connecting one of the Rampge rising outs to an Ampmix channel, set it to unity gain, connect the output to the decay of the BIA and set the decay to fully open. When I trigger the BIA I get the same decay as when I close the decay knob completely, suggesting that the output from the Ampmix cancels the value of the decay knob, which is 5v.

2. If you have some other source of signals instead of the Rampage, does everything behave as you expect? Yes, the issue happens only with the Rampage gate outputs: the rising, falling, and B>A outputs. I don't encoutner the issue with the BeatStep Pro gates, nor with any other CV I tried (Rampage other outputs or slope / contour from the 0-coast).

3. If you have some other mixer instead of the Ampmix (or just patch directly from the Rampage without the Ampmix), does everything behave as you expect? Yes, with the 0-coast mixer, or without a mixer at all, everything behave as expected.

4. If you feed a known negative voltage from some other source into the BIA, does the pitch input respond to that? No it doesn't! These are good news, as it make the issue a bit more consistent. I checked it now by putting negative voltages (from the 0-coast mixer) to the BIA pitch input and it ignores them.

First, I hope that the issue is a bit more clear now. Second, it is also somewhat more consistent: when the Rampage gates are connected to the Ampmix, and the channel is set to unity gain, the output is -5v to 5v instead of the expected 0v to 5v, regardless of what the Ampmix is connected to. Again, it only happens with these Rampage sources and not with anything else.

Any idea what it might be?

Thanks!
Tom

User avatar
mskala
Super Deluxe Wiggler
Posts: 2750
Joined: Thu Mar 27, 2014 7:33 am
Location: Toronto
Contact:

Post by mskala » Fri Aug 18, 2017 3:52 pm

Okay, I think the answer is in the Rampage and Ampmix schematics, which I just looked up. Here's a screenshot of the B>A output; all the rising and falling outputs seem to have the same basic configuration.

Image

Note there's a series diode in there. So when the op amp's output goes high, it will drive the output to the op amp's maximum output voltage minus one diode drop, with as much current as the protective resistor will allow. But when the op amp's output goes low, the reverse-biased diode prevents the op amp from sinking any current, and the output looks like just the protective resistor, the LED's current-limiting resistor, and the high-impedance backside of the LED. At low voltages, that's an open circuit. Bottom line: this output does not generate 5V and 0V or 5V and -5V; it generates 5V and "disconnected."

Most inputs patched into "disconnected" will read it as 0V, but now look at the Ampmix's input from the Ampmix schematic:

Image

The input jack goes directly into an op amp input through a protection resistor; there is no resistor to ground or anything. So when the Rampage's output goes to "disconnected," it's as if there were nothing patched into the Ampmix's input. Most op amps are not specified for how they behave with nothing connected to an input; it's as likely to do what you're seeing as anything else. Note this is different from having nothing plugged into the jack, because then it would be normalled... it's more like having a cable plugged into the Ampmix but the other end not plugged into anything.

So when the Rampage goes "low," the Ampmix gets no signal and it's more or less random how it will read that. And the problem is specific to the combination of these two modules, because most outputs provide drive at all voltages, not only in the high state, and most inputs do something more predictable when they receive no drive. I would call this a case of design flaws on both sides. I don't think it's any mistake you made in assembling the modules - they're both behaving as those circuits would normally behave, and it's just unfortunate that the result is kind of weird.

So-called passive OR combiners with diodes built in create the same kind of "drive/no drive" situation and this kind of problem is one of the disadvantages to using those.

Running the gate signal through a buffered multiple would probably make its behaviour more predictable, if you really need to connect this output to this input.
Last edited by mskala on Fri Aug 18, 2017 4:11 pm, edited 1 time in total.
North Coast Synthesis Ltd.
Twitch stream (Mondays, 3pm Eastern)

If you tell me that your goal is systemic change toward radical acceptance, and I see that you treat those you perceive as lesser-than with the same kind of scorn and derision that pushed me toward this insular little subculture where I feel comfortable [. . .] then you’ve successfully convinced me that your acceptance is not radical and the change you want not systemic. - "When Nerds Collide"

User avatar
mskala
Super Deluxe Wiggler
Posts: 2750
Joined: Thu Mar 27, 2014 7:33 am
Location: Toronto
Contact:

Post by mskala » Fri Aug 18, 2017 3:57 pm

Also if somebody patches a strong low voltage into the Rampage output while that output is low, they're likely to destroy the LED with reverse current. I would not design an output vulnerable to that, quite apart from the other issues with the design.
North Coast Synthesis Ltd.
Twitch stream (Mondays, 3pm Eastern)

If you tell me that your goal is systemic change toward radical acceptance, and I see that you treat those you perceive as lesser-than with the same kind of scorn and derision that pushed me toward this insular little subculture where I feel comfortable [. . .] then you’ve successfully convinced me that your acceptance is not radical and the change you want not systemic. - "When Nerds Collide"

User avatar
Nagasaki45
Common Wiggler
Posts: 230
Joined: Thu May 11, 2017 1:21 pm
Location: London

Post by Nagasaki45 » Fri Aug 18, 2017 5:07 pm

Thank you so much, again, @mskala.

I don't fully understand the schematics but understand the general idea, and it seems to explain the issue perfectly.

First, it's good to know that this is not a flaw on my side.
Second, I'm sending your answer to Befaco and RYO to see if they can "officially" confirm that.
Third, if I would like to fix this (which I'm not sure I acctually want), I guess that a fix on the Ampmix side is easier. One option will be, as you implied, to add a resistor between the jack and the ground. Am I correct? Another option is to change the op-amp to one that does something like that internally. I have no idea if such an op-amp exist. If you (or anyone else) do know I will appreciate some direction / feedback. For reference the kit op-amp is TL084. I will also ask RYO the same question.

And again, thanks a lot!
Tom

User avatar
mskala
Super Deluxe Wiggler
Posts: 2750
Joined: Thu Mar 27, 2014 7:33 am
Location: Toronto
Contact:

Post by mskala » Fri Aug 18, 2017 5:33 pm

Yeah, I think the easiest fix would be to add a resistor to ground on the input, the output, or both. I would use a 100kOhm resistor and attach it to the jack tip connection - not to the op amp input because then it would form a divider with the existing 22kOhm resistor and you'd lose some voltage. Doing this will decrease the input impedance a lot (that's the point), which could conceivably result in signal loss if you have very-high-impedance sources, which you probably don't. If that's a concern, you could go to a higher resistance (even as much as 10MOhm should be okay with the very high input impedance of the TL084), but 100kOhm is standard in Eurorack.

Adding this wouldn't protect the LED on the Rampage - I think handling that would be more complicated - but it shouldn't really be a problem as long as you don't patch that output into other outputs that have strong negative voltages.

To be fair to Befaco, they probably had in mind that you might want to OR-combine these outputs with a passive multiple; it's not completely insane to include the series diode to make that possible. But adding a resistor to ground wouldn't spoil that.
North Coast Synthesis Ltd.
Twitch stream (Mondays, 3pm Eastern)

If you tell me that your goal is systemic change toward radical acceptance, and I see that you treat those you perceive as lesser-than with the same kind of scorn and derision that pushed me toward this insular little subculture where I feel comfortable [. . .] then you’ve successfully convinced me that your acceptance is not radical and the change you want not systemic. - "When Nerds Collide"

User avatar
Nagasaki45
Common Wiggler
Posts: 230
Joined: Thu May 11, 2017 1:21 pm
Location: London

Post by Nagasaki45 » Sun Aug 20, 2017 3:41 am

Here's the response I got from Andreas from RYO:
Hi Tom!

If you add a ground resistor to the Ampmix input you will loose one of it's functions that is to ignore the output resistor of the module connected to the input. It's intentionally designed this way to counteract the voltage drop of non-buffered outputs.

If you are OK with the voltage drop that in the vast majority of cases doesn't really matter you can add a grounding resistor to the Ampmix inputs. If you add a very high value resistor like 4.7 megaohm the voltage drop will be insignificant.

Best,
Andreas
RYO
Last edited by Nagasaki45 on Tue Aug 22, 2017 4:18 am, edited 1 time in total.

User avatar
mskala
Super Deluxe Wiggler
Posts: 2750
Joined: Thu Mar 27, 2014 7:33 am
Location: Toronto
Contact:

Post by mskala » Sun Aug 20, 2017 8:00 am

Your call, but bear in mind that anything designed to produce a critically precise voltage (i.e. pitch CV) will be designed to drive a 100kOhm input impedance or (better) with an in-the-loop resistor so as not to care about the input impedance; so the Ampmix's attempts to "ignore the output resistor" are equally likely to create an error in the opposite direction.
North Coast Synthesis Ltd.
Twitch stream (Mondays, 3pm Eastern)

If you tell me that your goal is systemic change toward radical acceptance, and I see that you treat those you perceive as lesser-than with the same kind of scorn and derision that pushed me toward this insular little subculture where I feel comfortable [. . .] then you’ve successfully convinced me that your acceptance is not radical and the change you want not systemic. - "When Nerds Collide"

User avatar
Nagasaki45
Common Wiggler
Posts: 230
Joined: Thu May 11, 2017 1:21 pm
Location: London

Post by Nagasaki45 » Tue Aug 22, 2017 4:25 am

And here's Befaco's reply:
Hi All!!

Looks like both modules have floating I/Os.
You can fix it in both cases putting what is call a pull down resistor, as you say from the in to the ground or from the out to the ground. 1M will do the job and will not affect the signals.

Hope this help
So, RYO advised against changing the inputs, and Befaco recommended changing the outputs. I guess that I will go with adding a 1M pull down resistor on the Rampage side if I decide to fix that.

User avatar
Nagasaki45
Common Wiggler
Posts: 230
Joined: Thu May 11, 2017 1:21 pm
Location: London

Post by Nagasaki45 » Mon Mar 12, 2018 6:21 pm

For the record, I recently discovered that the CV inputs on the the 4ms DLD are not pulled down to ground either. So, when a gate from the rampage is connected to, let's say, the time input, no gate produces negative values.

I think I must mod my rampage...

User avatar
Nagasaki45
Common Wiggler
Posts: 230
Joined: Thu May 11, 2017 1:21 pm
Location: London

Post by Nagasaki45 » Wed Mar 28, 2018 12:21 pm

Hi all,

Today I fixed the issue by soldering 1Mohm pull down resistors to the gate outputs. Here are the results:

Image
Image

Work like a charm!

I would highly recommend doing this modification. I mix gates to create melodies, and attenuate gates to sequence parameters on the basimilus. Until now I couldn't do this with the gate outs of the rampage, but now it's possible.

Thanks for everyone that helped finding the issue and proposed solutions.

Tom

User avatar
Faustgeist
Wiggling with Experience
Posts: 499
Joined: Thu Apr 16, 2015 1:48 pm
Location: Columbia, SC

Post by Faustgeist » Thu May 31, 2018 8:12 am

Very interesting indeed!
Seeking the esoteric and funky in 5U/COTK/MOTM and MODCAN B, send me a message if you are selling gear and thanks!

User avatar
Natemasterflex
Learning to Wiggle
Posts: 25
Joined: Sat Aug 12, 2017 9:17 pm
Location: Washington State

Post by Natemasterflex » Wed Jun 13, 2018 11:46 pm

Thanks for sharing this!

Post Reply

Return to “Music Tech DIY”