Supercomp Problems, with OTA voltages...

[$uperpuma]

I have a supercomp that is giving me this problem: when I strum, I get a great crunch sound that promptly fades (not the sound I'm supposed to get since I'm running a clean sound into it). at some settings I get a blooming tremolo effect...that then dies off... no compression and certainly no sustain.

ca3080E voltages:

1 0v   
2 .6v
3 .6v
4 0v
5 .69v
6 .61v
7 9.85
8 0v

any ideas?

[R.G.]

Is there a bias network for the inputs (pins 2 ans 3) and the output (pin 6)? The exact voltages of these don't matter greatly, but the inputs need to be more than three diodes above ground, and they're only one diode drop above ground. The output can biased anywhere from V- to V+, but it's difficult to couple to other stuff if it's at one diode drop above ground. I'm not familiar with the schemo for this, and you do not provide a link to it (as I asked for in What to do when it doesn't work) so I'm only guessing based on knowledge of how 3080s work. With a link to the schematic I might be able to help more.

1 0v   
2 .6v
3 .6v
4 0v
5 .69v
6 .61v
7 9.85
8 0v

any ideas?
[/quote]

[$uperpuma]

I haven't found a schematic yet... its said to be a Dynacomp with an attack time control... tracing the board is a little tougher with everything mounted to it..
here's jack's version of the dynacomp schematic:
http://www.muzique.com/schem/dynacomp.gif

[ayayay!]

but the inputs need to be more than three diodes above ground, and they're only one diode drop above ground.

Wha?  Can you school me on this R.G.?  And can you tell that just from what he posted, or is that just a general statment regarding the 3080? 

[R.G.]

It's a general statement about the 3080. It's based on knowing the internal schematic of the 3080.

The 3080 is unusual. There are NO resistors in side. There are only transistors and metal interconnections.

Specifically, there is an input differential amplifier made of two NPNs, and then there are four two-transistor current mirrors, two bottom-side (NPN) mirrors and two top-side (PNP) mirrors. The inputs to the diffamp have to be one diode drop above their emitters. The emitters are tied to the output side of one NPN mirror. The input to that mirror is pin 5, the gain control pin. The mirror needs at least one diode drop to work at all, and two is much smarter if you want any input range at all, so you need around two diode drops on the mirror and then one to the input base, for three diode drops. The mirror input pin, pin 5, can ONLY work at one Vbe above ground, so it will always be 0.4 to 0.65V. If you put more than 1ma in it, the device self destructs because that tells the output mirrors to conduct too much current and it overheats and dies.

It's probably simplest to put them in the middle of the power supply, but not mandatory. That's because it's only the difference in input voltages that matters, and that can only be 25mV for linearity. The common mode DC voltage on the two input pins has no effect whatsoever on the DC voltage of the output. The DIFFERENCE between the two inputs can drive the output fully from top to bottom.

The output pin is the junction of two current mirror outputs, and is best thought of as a bidirectional current source. It really has no voltage except what is impressed on it by external resistors. The most common thing is to tie the output pin to some DC voltage roughly in the middle of the power supply through a resistor. The voltage tied to sets the DC level and the value of the resistance sets the voltage gain because the output current produces a voltage by ohm's law.

OTAs are neat, if somewhat strange devices. In the improved OTAs like the LM13700, CA3280 and NE5570, the current mirrors are replaced with improved Wilson or Widlar current mirrors and so the Iabc pin sits at two diode drops; otherwise, they work like the 3080 but with better specs and some added features.

[StephenGiles]

Hooorayyyyyyyyyyyyyyyyyyyyyy........"problems"!!!!!

[Paul Perry (Frostwave)]

Could be a cap missing or not properly connected?

[TELEFUNKON]

Make that: NE5517 instead of the brushless motor control  , R.G.

[$uperpuma]

Could be a cap missing or not properly connected?

I read a post by Jay Doyle that said the rectifier cap sometimes goes out and described the tremolo effect... but I only get that at certain settings