Is a 2M2 resistor bad to use ?

[JanJansen]

Hi, i use a 2M2 resistor on the base of a transistor, voltage goes from 1 to 11 volts in the 2M2 resistor.
Is it bad to use a 2M2 resistor somehow ?, since it is not very common.
thanks

[HB9EVI]

that depends... for a low power/high hfe BJT it might work if Ic is not too high either; for a FET - depending on the input capacitance it's likely going to take too long to charge/discharge the gate to have proper control over the transistor. Of course 1V would be too low for a FET; depending on the required current, you need >5V

for more accurate infos, you have to be a bit more verbose

[Zero999]

Yes a more detailed description is require, or better still post a schematic. There's no way of knowing whether this is good or bad, without more information.

[JanJansen]

BC549C
Emitter goes to circuit as control voltage
Base is tied to 2M2 resistor to vatiable voltage ( 1 to 11 volt ).
Collector is shaped control voltage from envelop generator.

The thing is if i lower the resistor, the shape from the envelop wont be the same, lets say overloaded/clipped.
The voltage need to be this high because the envelope going thru the transistor is also to 10 or 11 volts.

[ejeffrey]

It isn't "bad" to use a 2M2 resistor.  The only question is whether it is appropriate for your circuit and application.

2 megaohm is not super high, but with all high impedance nodes it is important to consider leakage currents and capacitive coupling that bypass your 2 meg resistor.

[JS]

There might be a way around for using a lower value, or if you really need 1M+ input resistance you probably want to use a JFET instead of a BJT. Hard to help without a schematic, even a rough one on a napkin could do...

JS

[Siwastaja]

Nothing "wrong", they wouldn't be manufactured if they were useless, yet up to 4M7 seems very standard.

Look out for:

- Calculation / simulation models being incorrect (i.e., possibly simplifications for higher currents)
- PCB surface leakage due to initially flux contamination, later environmental contamination such as dust, dew etc.; at 2M2, this easily dominates over the resistor tolerance rating, i.e., a 10% error is well possible if you have 22M of leakage resistance. Use bigger package size, and specially designed footprint with increased pad distance. Also, using several in series helps. Note you need to keep the careful thinking after the resistor, too, i.e., near the base where it's connected to.

[NiHaoMike]

Are you trying to bias the transistor with a resistor to the supply rail? That's a bad idea for many reasons - most significantly high dependence on temperature and unit to unit variations.

[iMo]

@JanJansen: Is this the circuit you are writing about???
In the simulation the V2 base voltage goes form 0V to 11V step 1V.

Not sure I've understood your circuit's description properly
If yes: the value (2M2) in the base has to be set such the emitter of the transistor can supply the load without clipping - here I am using 100k as an example.

Anyhow, the circuit as depicted on the schematics.. is a crap..   

Better write us, plz, what you want to achieve (what should be the functionality of the circuit)..

[JS]

Is your circuit that one?
What's going wrong with your circuit, it isn't behaving as expected?

That's an strange circuit, first time I see someone trying to use it. Common base, that's fine, but the other way around, using the emitter as input and collector as output.

Maybe add a previous stage as buffer, like an emitter follower or an opamp and then run this same thing at a higher current, maybe a lower hFE transistor, like a small power one, BD139 comes to mind. That way you can keep your ~100k input impedance, but use smaller resistances all around, like 10k instead of 100k and 100k instead of 2M2.

In any case, the output of that circuit will change a lot with temperature and from device from device, as NiHaoMike said.

The next option is a redesign of the thing to get a similar response, knowing the sources impedances is critical for that. The roll off is what itches to get a different circuit to do this, which seems quite important for an envelop generator.

  Few options that might work, try with a lower hfe transistor with the circuit as it is, and lower the base resistor. Try scaling down both resistors, maybe not exactly together, again, this will depend on the device.

JS

[JanJansen]

Yes that is exact the circuit.
The control voltage is coming out a opamp, and the envelop generator comes from a darlington pair transistors.
I placed the resistor very close to the base.

Here is what i try to achieve : replace the accent potentiometer in this schematic :
http://www.freeinfosociety.com/electronics/schematics/audio/pictures/rolandtb303.gif

Maybe i should use R1 of 47K instead of 100K also.

[iMo]

That single transistor circuit is not suitable for operating as a "potentiometer" controlled via the base current.
Single transistor (bjt or jfet or fet) circuit will be highly nonlinear, working "somehow" in pretty narrow base/gate current/voltage regulation region.
You would need something more complex, imho..

[JanJansen]

It works good enough + i have 10 bit PWM controlling it, so can have any curve ( will use only 8 bit PWM btw ).
I use instead of 1 potentiometer 2 transistors, 1 for accent filter 1 for accent volume.
Its the most simple and best i can get, what else would you suggest ?
Live tweaking knobs is not my best skill, i rather edit with mouse.

I,m having a 5volt PWM coming out of a PIC16, then i throw that in a FET with second order filtering ( this gives 1 volt rail for the non rail to rail opamp ), then i throw that in a opamp to make the voltage twice as high.
So its from 2 to 12 volt, not 1 to 11 like i say before.

If your method is better and works on a 5 volt signal instantly, please post a schematic, thanks.

[HB9EVI]

Sounds all quite weird. Maybe it would help us all, to give us a general overview on your project - I'm just guessing, but I'm sorry, I really don't get what you're aiming at with your explanations.

[NiHaoMike]

If the signal you want to attenuate is of much lower frequency than your PWM, you could just use the PWM to gate it and then low pass filter it.

[JanJansen]

If the signal you want to attenuate is of much lower frequency than your PWM, you could just use the PWM to gate it and then low pass filter it.

I dont want the signal slowing down, and certainly needs a second order filter to remove PWM artifacts.

Sounds all quite weird. Maybe it would help us all, to give us a general overview on your project - I'm just guessing, but I'm sorry, I really don't get what you're aiming at with your explanations.

As i told before : i am working on this synthesizer to replace the accent potentiometer with a voltage controlled thing, i use a transistor for that.
http://www.freeinfosociety.com/electronics/schematics/audio/pictures/rolandtb303.gif

[HB9EVI]

If your aim is to not mess up the audio path with any kind of distortion or noise from the pwm, you should use a digital pot,  e.g. a MCP42010; an Mcu you are obviously already planning with, so why not go this way

[Zero999]

If your aim is to not mess up the audio path with any kind of distortion or noise from the pwm, you should use a digital pot,  e.g. a MCP42010; an Mcu you are obviously already planning with, so why not go this way

I agree. If it really needs to be voltage controlled, just use an ADC.

[JanJansen]

I dont wanto make a cyborg synthesizer with some ADC,
a digital potentiometer is nice for this one, i better use that on the reso,envmod and cutoff, still has limitations ( single supply digipot cant handle negative values in the audio, so have to remove cap somewhere wich changes the sound a tiny bit ).
For this one i go simple i have gain , reso , cutoff & envmod potentiometers, the rest is without potentiometer ( waveform,tune,decay,accent ).

Here is a better link, the other dont works :
http://www.freeinfosociety.com/electronics/schemview.php?id=760

I go with the 2M2 resistor no problem, look how expensive those digi pots are, voltage controlled is better anyways.