3 Legged Dog + CMOS questions

puretube · July 15, 2004, 05:38:43 PM

edited...

Michael Allen · July 16, 2004, 01:09:56 AM

No I only have 2 outputs off of a 500ma transformer on this supply.

I'll try redoing the perf this weekend.

Athin · July 18, 2004, 04:17:41 PM

hi, I get a similar thing with a high-gain [1000 times] microphone preamp. I did a 'mod' to increase the gain [added another op-amp gain [10x] stage to the output]. The problem is that when I turn the gain pot to 60-70% the signal starts going on and off at first there is more 'on' but as I increase the gain there is more 'off'. I think it might have more to do with the max dissipation of the op-amp. Would getting 2 or even 4 op-amps [like 4558's] in parallel help?

puretube · July 18, 2004, 07:04:16 PM

no, it`s strictly a power supply/grounding routing/layout decoupling thing,
known since the 40`s of last century.....

puretube · July 22, 2004, 01:34:38 PM

free information block:

Michael Allen · July 22, 2004, 02:12:42 PM

Wow, I'd really like to see that.

I still can't get this thing to stop humming. I'm trying to make a pcb now, but can't get a laser printer access right now.

RDV · July 22, 2004, 02:56:47 PM

All my effects hum with those unregulated 9v(actually over 14) supplys. I have to use my Boss PSA-120 or equivalent.

RDV

puretube · July 22, 2004, 04:44:45 PM

that sort of motorboating isn`t neccessarily "hum",
but can be caused by miniscule powerplant voltage
fluctuation, which gets differentially amplified by
supply-voltage sensitive high impedance circuits
in different stages ("bad PSRR").

A regulated supply is best for both worlds...

Michael Allen · July 22, 2004, 08:02:05 PM

I've built a GEO power supply that I use, but even with a battery I get this hum.

And it is "hum"!!! It's loud in the background. It sounds like ground loop! so i'm sure it's a physical grounding issue on my perf.

puretube · July 23, 2004, 01:00:51 AM

err, but that "sounds" different to me than your firstly posted "fade in and out" pulsation, which effects all other pedals?

Nasse · July 23, 2004, 10:58:35 AM

QuoteI was wondering if anyone has any links to desinging with CMOS chips? Or information on using them as amplifiers?

Here is a formula for calculating the gain of a cmos inverter arranged as inverting amplifier stage:

A = - ((R2/R1)/(1+1/Ao+R2/AoR1))

C1, C2 = DC blocking caps, usually used at input and output circuitry (not included in the formula)
R1 = input resistor
R2 = feedback resistor
Ao = gain of open inverter stage (no negative feedback)

Ao is strongly depending supply voltage, frequecy and tolerances, and varies between manufacturers and individual chips, tolerance may be more than 20%. My data sheet suggests nominal typical values for Ao at different supply voltages, like

12 V: Ao = 50
9 V: Ao = 70
5 V: Ao = 100

I quess this is "nice to know" info and distortion stages are best designed by trial and error on breadboard

Michael Allen · July 23, 2004, 01:32:05 PM

Ton, I had some weird fading in and out, but I put a cap across the supply rails and it seemed to resolve that issue. Now I'm just left with some loud hum. I'm sure it's all due to my layout. I have a picture I'll upload later today that shows the layout. Maybe you could take a look and immediately spot a problem.....

puretube · July 23, 2004, 01:35:29 PM

free information block:

Michael Allen · July 23, 2004, 04:18:43 PM

So that equation is:

Where gain is negative? What is this Ao exactly?

Looking at different CMOS designs I'm foggy as to what the input resistor is? Is it a resistor from input to ground? or between signal and input?

P.S. Do you happen to have a preliminary schem for that circuit?

puretube · July 23, 2004, 04:38:30 PM

it`s the serial resistor (the latter one you mentioned)...

Nasse · July 23, 2004, 04:54:12 PM

In other words, hope this helps

R1 = input series resistor
R2 = feedback resistor
Ao = gain of open inverter stage (no feedback resistor)

I think Tube Sound Fuzz circuit stages are somewhat good example circuit for that equation. But that formula and suggested values are what were typical at year 1970, it was CD4001 circuit coupled as inverter. I think they nowadays use protection diodes and extra circuitry for static electric surge protection, so Ao values might be others. I quess the equation might be from some old RCA, Motorola or National Semiconductor application paper, or some other scientist, and I dont know if it is true. But it is better starting point than nothing.

Michael Allen · July 23, 2004, 07:49:19 PM

So If one wants to establish a stage with gain of say 50, how does he determine the two resistor values. I assume one of them is not variable?

Also, some designs don't have that input resistor. There is a lot that I don't understand and do not know, but I can't find anyplace to read about this. Does anyone have any more info or links to using CMOS. There's plenty I've found on Opamps, but I can't find any for CMOS.

Thanks guys!

puretube · July 24, 2004, 01:08:40 AM

free information block:

gez · July 24, 2004, 04:01:10 AM

Just to add to the soup, you can also use 4007s in these circuits

They have more of a crunchier vibe to them (Marshall?) than 4049s, but the intermodulation distortion is more pronounced. The above circuit is about as smooth as I've been able to get them (close to a 4049 in this respect, but still not a good).

As a rule of thumb, treat each stage as you would an inverting op-amp (though the above circuit partially decouples AC feedback), the difference being that gain is so low in each stage that the feedback resistor actually presents some input impedance (this is why you can get away without using an input resistor).

There's a good explanation of how these things work as linear amps in 'The Art of Electronics'. Plus plenty of info in the archives.

Nasse · July 24, 2004, 04:06:33 AM

My old magazine article refers to following sources (year 1973)

RCA COS/MOS Digital Databook + datasheets

Motorola: Motorola Complementary MOS I/CÂ´s, Application note AN-538 + datasheets

National Semiconductor datasheets

News:

3 Legged Dog + CMOS questions