Rampage DIY build problem - fall-time on channel A very long

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

Moderators: lisa, Kent, Joe.

Post Reply
TL072
Learning to Wiggle
Posts: 7
Joined: Sun Sep 06, 2015 3:42 pm

Rampage DIY build problem - fall-time on channel A very long

Post by TL072 » Sun Dec 01, 2019 11:29 am

Hi all,

I have been struggling to troubleshoot a DIY rampage for the last couple of months. It has led to much learning about analogue circuits but I am still not able to fix the problem. I re-flowed a lot of the circuit, switched the op-amps and I think checked most of the connections (there are lot).

The problem: channel A has a extremely long fall time. No matter what I do, I cannot seem to get it to be lower. There is some variation adjusting the controls but on the shortest time setting it is still several seconds long.

From reading about slope generators (kassutronics discussion of slope generators), the issue seems to be that the "variable resistor" created using 4 transistors (matched) is getting the incorrect input voltages (if I understand correctly: way too low). When I look at the link above, I can understand the circuit but when I look at Befaco's schematic (so nice they provide these btw - link) it is slightly different and I really struggle to understand what is going on, even after re-drawing the circuit several times. It seems to be similar to the original serge style one but not the same. Can anyone break it down for me? Even just a strong understanding of this may assist me to debug what is going on. Here is the picture:

Image

The nice thing about having one working channel is that I can measure the one more or less working circuit (there is something that I have not yet got to where the rise sticks a little... but that is for another time) and compare it with the non-working one. I have noticed that the output of the following op-amp is consistently -12v. The fall time is a combination of 3 signals. It would make sense if this -12v was pulling the rise time improperly long.

Image

If I understand the schematic correctly, this means that the voltage at pin 13 is fairly positive which seems wrong. I do not know how to measure that point because it is tapped off a virtual ground on IC5D.

The difficulty with this circuit is the whole thing is one big feedback network so trying to figure out what is causing what to occur is far from easy. If anyone can provide any suggestions I will be very grateful and even if someone can just explain what is happening in that rats-nest of matched transistors, I might be able to more easily figure out where the problem is... Many thanks!

jakobprogsch
Common Wiggler
Posts: 103
Joined: Tue Jul 31, 2018 7:48 pm
Location: Switzerland

Post by jakobprogsch » Sun Dec 01, 2019 6:00 pm

One way to understand the Befaco circuit is to think of it as two exponential converters:

Image
On the left are the usual expo converters as found in VCOs either as a current sink (NPN) or current source (PNP). Conveniently you can combine them as seen on the right. Whenever V is positive only the NPN part is active and the PNP are reverse biased and therefore don't contribute to the current. When V is negative the opposite is the case. The current linearly depends on V and exponentially on the CV voltages. You can use the grounded transistor bases as additional CV inputs with reverse polarity. You could also use this as a very primitive exponential VCA but it will suffer from some serious crossover distortion.

As far as the CV summing networks are concerned that is hard to guess. I would check that all the resistors have proper contact and are the expected value. The opamp being glued to a rail might suggest that the feedback resistor R21 is not properly connected?

Edit: Better picture to correspond to the actual Befaco circuit

TL072
Learning to Wiggle
Posts: 7
Joined: Sun Sep 06, 2015 3:42 pm

Post by TL072 » Wed Dec 04, 2019 4:17 pm

Many thanks jakob!

For some reason I did not get an email notice on your response so just seeing it now.

Your explanation is going to require some digesting for someone who is still learning but how you have broken the circuit apart in an image is immediately helpful and I think I more or less can see what is going on!

I have read about temperature compensated exponential converters in eurorack in the past (found the in depth article again if you are interested) but I guess what little I understood from that article did not lead me to see the underlying structure in the Befaco circuit. I also think that how you have broken apart the circuit may help me in the future with other tangles of components - so many thanks for that insight as well.

As for the R21, I will give it another check tonight. I have to believe I did re-flow / check it but that has been famous last words when troubleshooting other modules - so also thank you.

Edit: with your response to me you hit 100 posts. Congratulations!

Image

vaxdatex
Learning to Wiggle
Posts: 13
Joined: Fri Jul 05, 2019 9:56 am
Location: Seattle

Post by vaxdatex » Sun Dec 08, 2019 7:21 pm

I once went crazy on troubleshooting a Rampage module to the point I bought another one off Reverb to test and compare.

Did you swap the transistors to see if that makes a difference? Did you check all resistor values to be correct (this turned out to be the problem when I messed up of of my builds).

The Befaco boards need to be supplied with sufficient solder at all joints as some the vias may not be plated through properly. So make sure it really goes all the way through.

Have you sent an email to Befaco? They can at times take some time to answer but are usually helpful and nice folk.

Cheers
Vax

Oulipopo
Learning to Wiggle
Posts: 10
Joined: Mon Dec 09, 2019 7:18 am

Similar problem

Post by Oulipopo » Thu Dec 12, 2019 2:05 am

Hi!

I've been troubleshooting a similar problem for about a week now.

What is especially surprising to me, is the effect of turning the Shape A pot, as it has no effect on the shape of the contour, it just seems to further change the fall time of A.

What I've done so far:
- Double-checked the placements of all components
- Taken off the fader for rise A and cut the headers beneath it, originally the where pressing against the bottom of the fader
- Reflowed most, if not all, of the joints
- Shuffled around all ICs

What I haven't done:
- replacing the transistors ... I really hope there's a simpler solution

Control PCB - front
Image

Control PCB - back
Image

Main PCB - front
Image

Main PCB - back
Image

Out A - turning pot Shape A from fully CCW to fully CW
Image

All help or tips for how to continue would be much appreciated!


My apologies if this is thread-hijacking.

Edit: Added "shuffled around ICs" to the list of done items

Oulipopo
Learning to Wiggle
Posts: 10
Joined: Mon Dec 09, 2019 7:18 am

Re: Rampage DIY build problem - fall-time on channel A very long

Post by Oulipopo » Sun Jan 05, 2020 7:03 am

I've been trying to look further into the problems describe in my previous post, but haven't gotten anywhere. It seems to me like the "logarithmic portion" of A is dysfunctional, but I need help figuring out how to best isolate the issue.

My PCBs are still as shown in the last post.



It feels like most of my DIY modules have something off about them, so I'd really like to at least figure this one out. So all help would still be much appreciated.

User avatar
eewee
Common Wiggler
Posts: 225
Joined: Wed Aug 26, 2015 2:34 pm
Location: Belgium, close to Brussels

Re: Rampage DIY build problem - fall-time on channel A very long

Post by eewee » Sun Jan 05, 2020 8:26 am

Not sure this is going to help, but judging from your pictures, what I’d do:
  • clean the boards with IPA, T1 top leg solder side looks funny in the picture;
  • make sure that all solder flows through to the component side. It could be the lighting but I have the impression that not all of the holes are fully filled (R117, or the power connector for instance);
  • make sure you trim the leads sticking through on the mechanical side, so there is no risk of a contact with sliders, pots or jacks
Befaco builds can be quite dense, you’re not the only one suffering :confused:
Why, say she now, is I not glean that one may say of thing while thing is not.
(Alan Moore, Voice of the Fire)

Oulipopo
Learning to Wiggle
Posts: 10
Joined: Mon Dec 09, 2019 7:18 am

Re: Rampage DIY build problem - fall-time on channel A very long

Post by Oulipopo » Sun Jan 05, 2020 12:11 pm

Thanks for your reply, eewee!

I've gone through the tree points now, to no avail, unfortunately.

I've noticed something else, none of the two triangle symmetry trimmers does anything, not A, not B. Could I have messed them up trying to calibrate initially? (I stupidly turned them a lot when I initially saw that A wasn't symmetric.) Could such trimmers be ruined by turning them excessively?

User avatar
eewee
Common Wiggler
Posts: 225
Joined: Wed Aug 26, 2015 2:34 pm
Location: Belgium, close to Brussels

Re: Rampage DIY build problem - fall-time on channel A very long

Post by eewee » Sun Jan 05, 2020 12:22 pm

trimmer: can't help you there, sorry. I looked around, and there are as many people saying you *can* easily break them as those that say it is very hard. If you can manage to desolder them, you'll be able to measure whether they are still fine. Unless you've been going at them with a very heavy hand, my guess would be that they're still OK.
Alternatively, I'd contact Befaco, and ask them for some test measurements you could perform, so you can check those, and maybe find out where things go wrong.

Good luck!
Why, say she now, is I not glean that one may say of thing while thing is not.
(Alan Moore, Voice of the Fire)

Oulipopo
Learning to Wiggle
Posts: 10
Joined: Mon Dec 09, 2019 7:18 am

Re: Rampage DIY build problem - fall-time on channel A very long

Post by Oulipopo » Sun Jan 05, 2020 3:34 pm

Thanks again, eewee. Getting some directions as to how to continue was just what I needed now.

I'll try desoldering in a couple of days, and pushing another email on Befaco (I sent one as I posted the PCB pics in December).

User avatar
guest
Super Deluxe Wiggler
Posts: 3786
Joined: Mon Aug 19, 2013 11:49 am

Re: Rampage DIY build problem - fall-time on channel A very long

Post by guest » Sun Jan 05, 2020 7:52 pm

what is the voltage across R43? what are all the voltages on IC2?
openmusiclabs.com

Oulipopo
Learning to Wiggle
Posts: 10
Joined: Mon Dec 09, 2019 7:18 am

Re: Rampage DIY build problem - fall-time on channel A very long

Post by Oulipopo » Tue Jan 07, 2020 2:06 pm

The voltage across R43 seems to be 107 mV.

On IC2 I get the following voltages (all values in volts, relative to GND on the power connector)), running from pin1 to 14:
(p1) -5.0, 0, 0, 12, 0, 0, -5.6 (p7)
(p8) 0, 0, 0, -12, 0, 0, 0 (p14)

Do these values seem correct?

User avatar
guest
Super Deluxe Wiggler
Posts: 3786
Joined: Mon Aug 19, 2013 11:49 am

Re: Rampage DIY build problem - fall-time on channel A very long

Post by guest » Tue Jan 07, 2020 4:48 pm

with R43, one side is connected to ground. is that 107mV with that as the ground reference, or was the measurement made the other way around, so its -107mV? assuming the latter, it is then a problem in the summing amplifier sections. this is a sum of IC2 pin14 and pin1. put your meter on the wiper of the FALL POT, and see if this goes from 0 to +12V as you turn it. does the voltage on pin1 change as you turn it?
openmusiclabs.com

Oulipopo
Learning to Wiggle
Posts: 10
Joined: Mon Dec 09, 2019 7:18 am

Re: Rampage DIY build problem - fall-time on channel A very long

Post by Oulipopo » Wed Jan 08, 2020 1:42 pm

guest wrote:
Tue Jan 07, 2020 4:48 pm
with R43, one side is connected to ground. is that 107mV with that as the ground reference, or was the measurement made the other way around, so its -107mV?
Sorry, didn't notice that one terminal was at ground. It seems the other terminal is at a) a constant value of around 7 mV with the FALL POT all the way down, it's around -60 mV with the FALL POT in the middle, and around -200 mV with the FALL POT at the top, all regardsless of the RISE POT.
guest wrote:
Tue Jan 07, 2020 4:48 pm
assuming the latter, it is then a problem in the summing amplifier sections. this is a sum of IC2 pin14 and pin1. put your meter on the wiper of the FALL POT, and see if this goes from 0 to +12V as you turn it. does the voltage on pin1 change as you turn it?
Both the wiper on the FALL and on the RISE go from 0 to +12 V as I turn them upwards. The voltage on pin1 is 0.4 V (FALL at min) and -10 V (FALL at max).

Thanks for the help with debugging!

User avatar
guest
Super Deluxe Wiggler
Posts: 3786
Joined: Mon Aug 19, 2013 11:49 am

Re: Rampage DIY build problem - fall-time on channel A very long

Post by guest » Wed Jan 08, 2020 3:56 pm

ok, that looks good. can you measure the voltage on all 6 pins of U$1 (the dual PNP transistor) for fall pot full CCW and full CW? its a really small device, so it might be hard to get in there. if youre probes arent small enough, you can clip them to a small wire to probe with. its important to measure on the transistor, as it might be a bad solder joint on the transistor. if possible, can you take these measurements while the circuit is in the fall direction? you might also just try reflowing the solder joints on that transistor. and double check its orientation.
openmusiclabs.com

Oulipopo
Learning to Wiggle
Posts: 10
Joined: Mon Dec 09, 2019 7:18 am

Re: Rampage DIY build problem - fall-time on channel A very long

Post by Oulipopo » Wed Jan 15, 2020 3:01 pm

I'm not entirely sure that I did this correctly, but here are the values I got. They are identical for the SHAPE POT fully CCW and CW. I measured transistors T1 and T2. I set the FALL FADER at max value, the RISE at min, and had it running in cycles.

T1: Collector (top leg) 8.3 V
Base (middle) 12 mV
Emitter (bottom) 0 V

T2: Collector (bottom) 20 mV
Base (middle) 0.7 V
Emitter (top) 0V

User avatar
guest
Super Deluxe Wiggler
Posts: 3786
Joined: Mon Aug 19, 2013 11:49 am

Re: Rampage DIY build problem - fall-time on channel A very long

Post by guest » Wed Jan 15, 2020 5:55 pm

is that T1 and T2 in the flip flop section? i was referring to the super tiny 6-pin transistor package for the BCM857 in U$1. it would also be good to make sure this is the right part, in the right orientation.
openmusiclabs.com

Oulipopo
Learning to Wiggle
Posts: 10
Joined: Mon Dec 09, 2019 7:18 am

Re: Rampage DIY build problem - fall-time on channel A very long

Post by Oulipopo » Fri Jan 17, 2020 1:40 pm

guest wrote:
Wed Jan 15, 2020 5:55 pm
is that T1 and T2 in the flip flop section?
Well ... me being stupid and thinking this was going to be easy, of course it was. I see which transistor set you mean now.
guest wrote:
Wed Jan 15, 2020 5:55 pm
i was referring to the super tiny 6-pin transistor package for the BCM857 in U$1. it would also be good to make sure this is the right part, in the right orientation.
I now understand why you suggested getting some extra wires. Still, I cannot fathom how I'll get to them, so I think I'll need to make some extension cords to keep the boards a little separated. This will probably take me a few weeks, as I don't have the parts I'll need, so I'll post I reply when I have it done.

Thank you very much for all the help thus far, guest.

Oulipopo
Learning to Wiggle
Posts: 10
Joined: Mon Dec 09, 2019 7:18 am

Re: Rampage DIY build problem - fall-time on channel A very long

Post by Oulipopo » Wed Feb 12, 2020 2:57 pm

Okay ... so I've given it a go. I couldn't, however, find a way to measure on the IC directly, so I used the points on the back of the main pcb as shown in the image below. I checked for continuity between the point I measured and the leg the IC. All measurements were on longest fall, shortest rise for short range.

Image

For leg 1 and 4 on the IC (both collectors) I got the following:

Shape CCW
Image

Shape CW (high value is while falling)
Image

For leg 2 on the IC (base) I got the following:

Shape CCW
Image

Shape CW (high value is falling)
Image

For leg 3 on the IC (emitter) I got the following:

Shape CCW
Image

Shape CW
Image

For leg 6 on the IC (emitter) I got the following:

Shape CCW
Image

Shape CW
Image


I have no idea if this is legible to anyone, but thanks for looking and trying to help.

Post Reply

Return to “Music Tech DIY”