Using TO-220 ( power) BJT's/Mosfets in stompboxes

[Max999]

What would be the general tonal effect of substituting a small signal bjt/mosfet for a power type ( TO-220)? Let's say we use it in a ""neutral" circuit with  gain of about one, a buffer.

Why? I would like to increase the creepage between terminals, and they also look cool to me. Are all bigger bjts and mosfets automatically power bjt/mosfets or are there also big small signal ones?

[Gus]

think about beta and capacitance

[PRR]

How much "creepage" is in a pedal?

Why would they use a One-Ton package for a moped-load device?

Just do it. Gus is correct that strays may be significant. But that may not be bad.

[amptramp]

One thing that may become an issue is the interelectrode capacitance.  Not only is it higher than tube capacitances, it varies significantly with collector / drain voltage, so you may have a distortion mechanism built in.  In a grounded-source gain stage, the Miller capacitance which is the gate to drain capacitance times ( 1 + gain).  One way around this is cascode amplifiers which hold the drain at a near constant voltage.  This allows the input capacitance to be the gate-to-source capacitance plus the gate-to-drain capacitance and the fixed voltage means no variation in capacitance.

Large FET's have a higher transconductance than small ones but large bipolar transistors tend to have lower gain than small-signal devices unless you go with Darlington transistors.

Check out all the device specs and build it if you think you can tolerate the non-linearities.  You will either find it is what you want or you will find out why not many other people have done it before.

[R.G.]

A.T. beat me to it.
> creepage is a concept that was conceived for safety reasons in high voltage circuits, especially AC mains powered. Creepage is an alien concept to low voltage, low frequency circuits
> there is no advantage for having the pins further apart in low voltage circuits, at least to the circuit itself, and at low frequencies. Audio is practically DC compared to the frequencies where pin spacing would matter.\
> one ton truck vs moped is a good way to look at it; manufacturers only put power devices in packages with exposed thermal dissipation tabs, and TO-220 is one of these
> higher power devices tend to be lower gain devices, as A.T. said, because the makers optimize the parts they put in thermal packages for high power, not high gain; one exception is power darlingtons, but these have other problems.
> I kind of disagree with A.T. on interelectrode distances being an issue; MAYBE with super high impedance FET gate circuits; probably not with bipolars

If you want to do it just because it looks cool, go for it. Or if it just makes it easier to solder, which is one advantage that came to mind.

[bool]

...
Large FET's have a higher transconductance than small ones but large bipolar transistors tend to have lower gain than small-signal devices unless you go with Darlington transistors.
...

Power darlingtons usually have inbuilt B-E resistors.

[PRR]

> Power darlingtons usually have inbuilt B-E resistors.

Something to watch for, yes.

FWIW, I have build several "small audio" systems with TIP120, a 65 Watt TO220 package with 6K and 1K resistors in it. If this is stages 2 and 3 then your first stage needs >0.2mA which is often where you want to be. (FYI: Iceo of TIP120 is cited as 0.5mA max; it may be hard to run it much leaner but maybe only when it is hot.) If you use it as a single stage with an emitter resistor to set gain, the 6K will bootstrap up to much more.

[bool]

...
If you use it as a single stage with an emitter resistor to set gain, the 6K will bootstrap up to much more.
...

Yes but only if it isn't bypassed with a capacitor.
If there is a cap in parallel with the emitter res. the projected increase in input impedance will inversely follow the frequency band contour - the "input resistance" will be frequency dependant; biggest at frequencies below the pass-band and diminishing with the rising freq. (vice-versa in unlikely case you use an inductor).

[Max999]

Allright, so the use of power bjts and mosfets can sound interesting, but it will be non linear. Might be worth a tonal experiment.

About the creepage wish: I learned that the amount you want is dependent on the pollution factor of the pcb and the applied voltage.
I have not read that below a certain voltage tracking can not occur.. It would be nice to learn that it can not happen though.

Does anybody have information on this?

[R.G.]

Allright, so the use of power bjts and mosfets can sound interesting, but it will be non linear. Might be worth a tonal experiment.

Using different devices for different "side effect" changes to tone is a highly respected and fundamental approach to musical effects. Great thing to try.

About the creepage wish: I learned that the amount you want is dependent on the pollution factor of the pcb and the applied voltage.

You have been reading safety-related stuff and trying to apply it to low voltage effects circuits. The pollution factor is a concept used in safety testing specs for determining what creepage and clearance distance must be maintained for a given voltage level. the whole idea is to prevent dangerous voltages (either over 42V peak AC or DC, or lower voltages with a higher current/energy level) from breaking down the separation between circuit traces and shocking a user or starting a fire. The pollution factor attempts to allow for the boards being covered in dust, lint, high humidity, and so on.

It's a laudable goal; however, it's not applicable to pedals, as (1) the voltage are low to start with and (2) no pedal supply I know of can supply enough energy to make a "hazardous voltage" under the definitions of the safety regs. So this does not apply in general to pedal design. There will be unusual pedals where a high voltage is generated inside the pedal, like for a vacuum tube or some such.

I have not read that below a certain voltage tracking can not occur.. It would be nice to learn that it can not happen though.
Does anybody have information on this?

I puzzled over that statement for a while. If you mean "tracking from a voltage arc zapping across a gap", then yes, that is true. But again, this would be unlikely to about six decimal places in a pedal.

If that is what you meant.

[Kipper4]

Could they be used as a clipping diode?

Re this.

http://www.muzique.com/lab/zenmos.htm

[Rob Strand]

Could they be used as a clipping diode?

http://www.muzique.com/schem/shaka5.htm

[Kipper4]

Thanks.
A personal favourite. With mods.

[Max999]

Thank you RG. I have indeed been readfing safety information for high voltage designs. I know the low voltages and currents in pedals cannot harm humans, but I was thinking that a partial short could also influence the functionality without killing it. The partial short would be from pollution that gets a dose of condensation. FR4 does not have a very hight tracking index nr also , so I am always looking for things with bigger distances between the leads.

I wish I could pot circuits, the best next thing is the conformal coating.

Which creepage distance do you use here without conformal coating people?
The distance between opamps pads are about 0,6mm ( 0.023622 inch) usually. It always looks too small to me.

[Phoenix]

Under standard IPC-2221 (there are many other standards with different requirements, but this is a common one), creepage for uncoated traces is 5mm per kV on FR4, or just over 250Vpk per 0.1" (120V for 0.6mm). I think you're concerning yourself with a non-issue. Reference.

[Rob Strand]

but I was thinking that a partial short could also influence the functionality without killing it. The partial short would be from pollution that gets a dose of condensation.

Making sure a low voltage circuit works under adverse conditions goes outside of what the safety standards are about.   You have to address the specifics of the circuit.   

For example if you get water on a 0.1ohm shunt for a battery charger it won't care too much.  However if you have a sensitive measuring instrument with 1Gohm input impedance you might find the leakage on a *clean* PCB is too much and this is where you might use guard rings in your PCB design.   In some designs you can use ceramic stand-offs, in other cases  ground planes and ground pours may help.   It's all about using specific things for specific problems.

Effects pedals are in between these extremes and for the most part general condensation levels are not a problem.   Many years ago a lot of boards in cars and some industrial control gear had a lacquer coating over the bottom of the board and in some cases over the top of the board and parts.  That would help seal in the parts without having to use a thick conformal coating.

I tend to agree with Greg.

I think you're concerning yourself with a non-issue

[Max999]

Thanks Phoenix and Rob. In Phoenix' reference page there is a mention of ul-60950-1. this is the document I am using for determining creepage and clearance now. They say this is for " information technology equipment". The creepage tables in this document are much tighter then ICP 2221B.

Very confusing .. which one do I use ..

[PRR]

I think you're concerning yourself with a non-issue

[Rob Strand]

Very confusing .. which one do I use ..

There's no right answer.  They are all sort of achieving the same task.

Generally you have to comply with a standard, or standards, for some reason.  Usually your market forces you to meet certain standards.   It is a confusing business coming to terms with the details of standards.

If you want to sell products you have to comply with ul-60950-1.  That only applies to certain products.  For example, it won't apply to medical products which use another standard.   I don't think it applies to appliances like Toasters either.   Transformers have their own standards.  Also, you have to comply with the version of the standard that is active.

If you are building stuff for NASA they might require IPC-2221B (I don't know if that's what NASA use, it's just an example) or they might even demand both IPC-2221B and ul-60950-1 for some types of devices.

The motives behind ul-60950-1 and  IPC-2221B are different.   ul-60950-1  is strongly focused on product safety whereas IPC-2221B is more aimed towards reliability, manufacturability - basically avoiding problems.  It is interesting to note IPC-2221B actually says that safety standards override the IPC-2221B requirements/guidelines.

[Phoenix]

It's also worth noting that UL60950-1 doesn't require that you follow its guidelines, rather that the product pass HiPot testing, which any normal product will with much closer spacing than recommended, UNLESS it compromises safety. Something like a pedal powered by an external class II power supply would be challenging to MAKE unsafe.