Some thoughts regarding the Easyvibe's LFO

[mike_a]

Hi everybody,

For some time now I have been planning to build an Easyvibe.
I have read a lot of posts where people complained that the EV lacks sweep range etc.
I've studied the schematic and something came to mind in regards to a possible design flaw of the LFO.

They way it is now, and I'm refering to John Hollis' schematic, is that there are 2 pairs of LEDs(in series), connected in parallel.
The 4 LEDs are connected, in series, to a 220 ohm resistor and a 1k trimpot.

The problem I see is that if the LEDs forward voltage is not exactly the same (and it's not), the LEDs with the smaller forward vlotage will limit the ones with the larger forward voltage.
This can cause the LEDs with the larger forward voltage to not reach their maximum lighting capabilities.

If I would have had to design this part of the circuit, I would connect each LED to a separate resistor.
Then connect the four LEDs+resistor in parallel.

Or, with keeping to the original design, connect 2 LEDs in series with a resistor.
Well, I hope you get the idea....

I would really like to hear what you guys have to say about this.

Mike.

[R.G.]

John's design, like all good designs, is a compromise between ideal operation and complexity. It would probably be better (more "ideal") if all the LEDs were in series, forcing the forward current to be the same. That would require a much larger power supply as the LED voltage would be something like 5-8V all by itself.
One resistor per LED would be good, but that's more complex, and now the currents are mismatched.

If I had to change that design, I'd probably go with a four-output current mirror. Within close limits, a current mirror forces the current from an output to be the same as the current forced into a control node. Neat circuit. You pay for the neatness with complexity; it costs you at least one control transistor and one additional transistor per output. More perfect current mirrors use more transistors (what a surprise!). So a simple four output mirror is one control transistor and four output transistors. Possible, and easily feasible, but now there are five more transistors (although no resistors and caps) to get backwards, wrong device, and soldered incorrectly. Resistors are easier for beginners.

[mike_a]

Hi R.G, and thanks for your reply.

I believe that you misunderstood me slightly, which is understandable because I can barely understand what I wrote...
What I meant wasn't to force the same current through all of the LEDs.
I drew some pictures in order to clarify.

the first one is the original:

In this case, If, for example, the 2 left LEDs have a combined forward voltage of say 3.2 volts and the 2 right LEDs have a forward voltage of 2.8 volts - the right ones will force the voltage of 2.8 volts on the left and prevent them to reach their nominal voltage (and not light as bright as they should).

the second one is using a separate resistor for each LED:

This way each LED can operate in its own forward voltage, and the resistor will take the rest on itself.

The third is what I would do because it would have the best "benefit to cost" ratio:


This would involve adding only one resistor and solve the problem of co-restricting forward voltages.
Although I believe the second method would yield a larger sweep range.

Mike.

[soggybag]

I built an easy vibe a few years ago. I remember having trouble getting it work. The problem had something to do with the LEDs. It would not vibe unless all 6 of the LEDs were the same type, the four in the LDR section and the two in the voltage divider. I never quite understood what was going on here.

I'm thinking about building another. Maybe I'll give your idea a try.

I was also thinking about using only two LEDs in the LDR section. PLacing one LED in between two LDRs. The Phaser Fleur does this. Also the Univibe uses a single bulb surrounded by four LDRs. 

[mike_a]

Thanks for the input, soggybag.

I have another concern regarding the opamp's capability to drive the required current.
I saw another thread where someone (tone god if I'm not mistaken) suggested running the LEDs from 9v to the opamp's output,
rather than from the opamp's output to the ground.  This way the opamp is not the current source.
however, I don't know how it handles the reverse current in it's output so I'll have to check.

I will try to breadboard the changes and check.  I just need to find time for that....