Uni Vibe LFO Help (Achieving more depth) - Development videos

[Kevin Mitchell]

I've been tinkering with a vibe idea with opamp phasing and a classic uni vibe like LFO. Very similar to the Micro Vibe or the Chicken Salad. But like most of my vibe builds, I still have the same idea I can't seem to achieve tinkering on my own.

So I ask;
How would I go about increasing the depth of the oscillation bringing the bulb brightness down even further? Increasing the total depth travel. I've been tinkering with values and different NPNs for a while now without any insight of luck.

Information on the LFO is presented here (about 1/3 through) but does not help me where I need. Also - I'm having trouble holding any depth at higher speeds. We'll get into that when the time comes. I'm using a very small incandescent bulb "Desc: 5.0V .115A". So far it's perfect - just missing that depth.

I'm using the classic Uni Vibe LFO with some minor tweeks. Omitted the 220k resistors and changed some values around. Nothing major. Running on ~9V.

Thanks guys. I'll keep at it. Fingers crossed for ideas 

[R.G.]

There are two fundamental problems with the Univibe depth at higher speeds. It's to do with the bulb's time constant and the LDR's response time.

The filament of your bulb, microscopic as it looks, has mass and takes some time to heat up when you put current into it. This can be "helped" by putting a big pulse of energy into it to start it heating. But it also has a time constant on cooling down. Where the heating time is determined by the rate energy goes into the filament and the filament's mass, the cooling time constant depend on how fast it can radiate the heat away and conduct it into the inert(ish) gas in the bulb. Cooling is generally slower than heating because you can't make it cool any faster than it radiates away energy, even if you abruptly turn the current in the filament off completely.

So as the LFO speeds up, the bulb reaches a point where its temperature can't fall as fast as the LFO waveform that's trying to drive/release it, and so the bulb variations from brighter to dimmer start averaging out, and you don't get as much sweep.

Much the same thing happens with the LDRs. LDRs are all compromises. The material and pattern of conductor on the exposed surface is tinkered to reach some compromise of resistance range, response speed, and memory. LDRs can be ... very... very ... slow... depending on the compromises. Univibes need fast ones. LDRs can have a narrow resistance range. Univibes need wide range of resistance changes. LDRs have memory, where they tend to change mostly to a new state at their nominal response speed, then slowly adapt to the new condition. That's why the LDR datasheets talk about "dark adapted" and "light adapted" measurements.  So the LDRs are slowing the change of resistance (which is what you need) down even more than the bulb is slowing down the light changes.

This can be tinkered with, but probably not "solved" with lights and LDRs. The light speed up/down can be solved by using LEDs. But the bulb heating/cooling time constant is recognized by some people as a unique feature of the univibe. The LDRs can be replaced by whatever you can find, and you might find a better compromise. A real "solution" probably means junking the LFO, lamp, and LDRs and hooking in some electronic resistance change mechanism, then tinkering with it to fake the part of the univibe sound you like and not abuse it too badly if you like the side effects.

On the other hand, you might be able to swap bulbs and LDRs and reach a compromise you like better. I'm sure we'll get some folks in here to tell you their favorite recipes.

[Kevin Mitchell]

Thank you very much R.G.! That's great information allot of people can use. Although I really wish a solution was a bit more obvious.

Here's a drawing of what's breadboarded for the LFO;


Is there no simple way to increase the amount the 2 transistors drag the lamp down?

Also: I don't feel that the LDRs play much role in what I'm aiming for. The swoosh I'm after is there when I manipulate the 500 ohm trimmer. The LFO just doesn't reach that depth on it's own which is what I'm after. I can get the high swing easily - but not the low.

[Kevin Mitchell]

Here's a quick video for example.

EDIT: Replaced with new video.

You can tell when I manually manipulate the lamp for example of the notches it can - but will not reach with the present configuration.

[R.G.]

Is there no simple way to increase the amount the 2 transistors drag the lamp down?

Time for defining terms. When you say "drag the lamp down", do you mean how dark they make it, or how bright? Is up bright and down dark, or vice versa?

Also: I don't feel that the LDRs play much role in what I'm aiming for. The swoosh I'm after is there when I manipulate the 500 ohm trimmer. The LFO just doesn't reach that depth on it's own which is what I'm after. I can get the high swing easily - but not the low.

Sadly, that's one way to describe slow thermal response on the lamps. The trimmer changes the lamp current and the bulb has time to cool down.

If you're after getting a wider swing of light out of the lamp and into the LDRs and you want this to persist at high LFO speeds, you're fighting some built-in time constants. It is also possible that the LFO is losing output swing at higher speeds. I've seen some of them do that.

Jut to eliminate that possibility, you might try replacing the two 3904s in the "darlington" for the LFO with a monolithic darlington like an MPSA13 or MPSA14, or even making the first NPN be an N-channel JFET and rebiasing the LFO.

But to eliminate on the bench, I'd scope the LFO, find out what its output size is over the speed range, and then feed the driver transistor with a signal generator just to force the LFO signal size out of the equation.

If it turns out that you're fighting the thermal time constant of the bulb, your choices are narrower - change the bulb for a bulb with a shorter cooling time constant, or ditch it for something faster, LEDs being the obvious first choice.

[Kevin Mitchell]

I've updated the video above maybe that would help you understand what I'm after!

I meant I could get the lamp to shine bright enough but in doing so - can not reach the low light level in it's depth travel which is what I'm trying to fix up. So I'm trying to get it to oscillate deeper/longer/wider. Get the bulb to shine very bright and very dark and once that's cleared up I can tinker with the speed problems which you may also notice in the video.

Sorry for any confusion 

I have some darlingtons stashed somewhere I'll keep at it.

EDIT:
I don't think it has to do with the rate of the bulb cooling. You can tell from the video that I can manually get the depth out of it. I believe it has to do with the strength of the oscillation "how much the bulb is pulled down". I feel this needs to be and could be increased but I don't know how to go about doing so. I've been at it with my minimal knowledge of electronics so I really thank you for your patience and cooperation.

[R.G.]

I don't think it has to do with the rate of the bulb cooling. You can tell from the video that I can manually get the depth out of it.

It's only my guess from very incomplete info. But I have seen exactly this issue before, that's why it's so familiar.

I believe it has to do with the strength of the oscillation "how much the bulb is pulled down". I feel this needs to be and could be increased but I don't know how to go about doing so. I've been at it with my minimal knowledge of electronics so I really thank you for your patience and cooperation.

And that is why I suggested eliminating the LFO as a cause for changing LFO signal size with frequency. Subbing in a signal generator and watching the signal to the bulb driver and possibly the collector of the driver transistor on a scope will instantly tell you whether it's signal size from the LFO varying or lack of frequency response from the bulb.

[Kevin Mitchell]

And that is why I suggested eliminating the LFO as a cause for changing LFO signal size with frequency. Subbing in a signal generator and watching the signal to the bulb driver and possibly the collector of the driver transistor on a scope will instantly tell you whether it's signal size from the LFO varying or lack of frequency response from the bulb.

I certainly believe it's not producing a full sine wave as it should.

I do not have a scope lying around I could use. Although I've been studying so many LFOs lately I would find one quite useful. Maybe next Christmas 

[deadastronaut]

hi kevin, have you considered an IC lfo /led/ldr combo..rather than bulb etc..

i have messed with these lfo's/leds a fair bit, and can get the led to go practically off, to bright.

which is what i think you mean?...

check out the tremshifter lfo.. might get you a better depth..

[Kevin Mitchell]

I've been trying to keep my focus on a classic vibe LFO mostly for that authentic fast high, slow low - "the thermal problem" but if I'm going to run into these problems it may be best to work with something more stable/versatile.

I have these yellow LEDs that match the orange glow of the bulb perfectly. I'll check out the tremshifter and also mess around with the easy vibe's schmitt trigger LFO.

I've also had the thought of recreating RM's VooDoo Vibe LFO with a newer/better waveform IC. That may be the direction I take.

[R.G.]

LDRs have a spectral sensitivity that is a reasonable match to the human eye in most cases - which means it peaks in the visible green. The orange/yellow light of an incandescent bulb wasn't all that great a match to the sensitivity of the LDRs in the first place, but that's all that was available.

[Kevin Mitchell]

Okay. You are indeed THE man, deadastronaut!

That LFO is perfect for my Vibe/Trem project. It's exactly what I've been after to achieve an "exaggerated univibe" lol. After a little tweaking here's a little sample of it in action. Don't mind the dark room and recording quality

Fiddling with Speed, Depth and between a square and triangle waveform - vibe only.

Needs a little more work - a trimmer for the high end of the LEDs and symmetry control if possible but other than that this thing has allot of potential! Loving the simplicity of it and especially the waveform switch. Very great!

Could we pull any other wave forms from this thing? 

[strassercaster]

great question  idea and post.  i am a phase univibe fanatic. I wasnt happy with my harbinger one univibe at first so i bought the sealed ldrs from small bear they are 3.75 each and worth every cent. i bought 8 and semi matched them . i had the brightness intesity too high i found that the standing them and bending them so they were almost touching the bulb helped a lot for depth andbeing chewy sounding. i used the reccomended 7371 small bear bulb. it made a huge differnce literally night and day . i also learned it was more important to go super dim than super bright. imho. the sealed top of the line from small bear rise and fall quickly with a large range.i am glad i bought 8 because three were all over the place. if i would have bought 5-6 it probably would have sounded as good. What bulb are you using now? How did you make it ? perf,stripboard,homade pcb,madbean pcb?I. spent a few nights on this samething a while back. in the end i was extremely happy and Me and three others picked it in blind tone tests over the micro vibe and dunlap uv-1. it was almost identical to the Mojovibe just as chewy but not as fast. No matter , i do hendrix and trower and write so none of that needs crazy speed i am very hapoy ha ha. let us know ill be checking back in. GOod luck. these are a challenge and have some quirks . i just started a stripboard layout that will fit into a wah. shell .i cant find a dual 100kc 24mm pot anywhere. i have a 16mm but i am afraid it wont last in the wah. thanks and good luck also i experimented after i did the direct light mounting it didnt even need the light shield but i used it anyway.

[deadastronaut]

cool, glad its working out for ya!..

now you just need some night vision glasses..


edit:

just  thought, i guess your on breadboard, you might like to take a look at the faze filter too..

its another switch but, you get 2 speed settings, fast to uber slow,

and manual which turns the lfo off...so it acts as a manual filter too.

while your tinkering...

[Kevin Mitchell]

I've been meaning to pick up a load of JFETS for various reasons - mostly to try the phase 90 which I assume the Faze Filter is based off of? Sadly I do not have any around at this time. Unless you were strictly talking about it's LFO?

Last night I plugged in another 4 stages with random cap values to see what would happen. Best described as "lush as f***" It's amazing. The extra stages throw more phasing depth (level) into the mix creating the illusive swishy creamy univibe swell I couldn't nail from any other vibes I've built.

strassercaster - That's allot of info and trouble about positioning LDRs. I've found those LDRs to be very costly and not really worth it even for a "boutique" build. If you read what R.G. has been saying here you may understand a bit more why. Even I have purchased a couple sets trying to nail that classic vibe sound.

[deadastronaut]

yeah pretty much the lfo speeds/and off i was talking about..

sounds cool...

[strassercaster]

I've been meaning to pick up a load of JFETS for various reasons - mostly to try the phase 90 which I assume the Faze Filter is based off of? Sadly I do not have any around at this time. Unless you were strictly talking about it's LFO?

Last night I plugged in another 4 stages with random cap values to see what would happen. Best described as "lush as f***" It's amazing. The extra stages throw more phasing depth (level) into the mix creating the illusive swishy creamy univibe swell I couldn't nail from any other vibes I've built.

strassercaster - That's allot of info and trouble about positioning LDRs. I've found those LDRs to be very costly and not really worth it even for a "boutique" build. If you read what R.G. has been saying here you may understand a bit more why. Even I have purchased a couple sets trying to nail that classic vibe sound.

Understood. i guess i should ask did you find some that were good that didnt cost so much somewhere. I really like what your working on . i am a univibe/phase fanatic ha ha

[Kevin Mitchell]

I'm back with a little eye candy!

I've scoped the waves of the 3 LFOs mentioned in this thread.


Maybe there's an error on the breadboard but Mr Hollis mentioned the EasyVibe LFO being sine - as a Uni-Vibe is but here you can tell it very much resembles a triangle wave. Also I'm not sure why or how, but omitting the diodes on the voltage bias - to LED opamp(or depth pot) does nothing the the waveform. What do you guys think? I'll try again another time.

Can we get a sine that pretty with a dual opamp and less parts? The green LED is swinging nicely on the Uni-Vibe LFO 

-I would like to note - If I intended to mimic/emulate how a vibe lamp/lfo would swing but with an LED, wouldn't the tremshifter LFO be a good candidate? Slow rise (bulb heating up) then a quick fall If only it was more sine and hung onto the high end a bit longer. Like a sine/ramp hybrid waveform. Any thoughts?