Running Pedals at 18v Question

[Andrekp]

So I have a DIY Marvel Drive using the layout at Tagboard Effects.  I am currently running it 9v.  I'm thinking about trying it at 18v, but have a couple of questions that may be generally applicable as well:

1).  This pedal requires biasing 3 of the 5 JFets to a particular drain voltage in the area of 4-5V.  Is this effectively just to set up a relationship between the drain voltage and the power supply voltage?  In other words it's not the particular voltage that matters, but the fraction of the total - MEANING: if I have the JFets biased from running at 9v, and I plug it into 18v, I don't need to re-bias the JFets because they are already set up with drain voltages relative to WHATEVER the supply rail might be?

2). All my caps have the appropriate voltage rating headroom, so to speak, for running at 18v, with one exception:  There is a 330u cap in the source circuit to ground on one of the jFets.  I did not have a 330u so I paralleled a 220u and a 100u to get within range.  the 220u is fine for voltage, but the 100u is only a 16v.  I know the simple thing is to just cut out the 100u, since is probably not needed anyway.  I also know that the voltage rating for the caps becomes the same as the lowest rated, so that leg is effectively only rated for 16v.  BUT: does is matter in this spot.  Does the source to ground leg of a circuit like this ever see anything like the supply voltage?  If the drain is biased to 4v on this JFet, doesn't that mean that the source will be AT MOST 4v?  So I could just leave it as is?  (different story, obviously, if these were power supply filters, or whatever).

Thoughts?

Thanks
-Andre

[R.G.]

So I have a DIY Marvel Drive using the layout at Tagboard Effects.  I am currently running it 9v.  I'm thinking about trying it at 18v,

Can you articulate exactly why you want to do this?

[Andrekp]

Because I'd like to see what it sounds like...  The same reason I built it in the first place.  My understanding is that these things CAN run at 18v and that they have a bit of a different character at 18v. 

But I also think my two questions are more informative (or the answers would be) in general.  Do JFETs that have to be biased in a circuit have to be re-biased when you change the supply voltage?  Do electro caps in circuit ALL have to be rated for higher voltage, or can ones like these be ignored because they will never see but a small portion of the supply voltage?

[R.G.]

Because I'd like to see what it sounds like...  The same reason I built it in the first place.  My understanding is that these things CAN run at 18v and that they have a bit of a different character at 18v. 

OK, just checking. If you know and accept that it may not be better and may be worse, and just want to see what happens, accepting that it can break things, that's a fine and dandy reason for trying.

But I also think my two questions are more informative (or the answers would be) in general.  Do JFETs that have to be biased in a circuit have to be re-biased when you change the supply voltage? 

In general, yes. It will be unusual if they don't need to be rebiased, although it is possible. Given that rebiasing JFET circuits changes the way they sound, it is also possible that rebiasing to some different operating point all by itself would make the circuit sound - well, it will certainly be different. Might even be in a good way.

Do electro caps in circuit ALL have to be rated for higher voltage, or can ones like these be ignored because they will never see but a small portion of the supply voltage?

Tricky question. "Never" is a long time, and funny things happen at turn on and turn off. If you're sure they will "never" see more than that fraction of the power supply under any conditions, sure, use lower voltages. Circuits from the 1960s and 1970s did this a lot. Many of them survived.

It's also different depending on whether you're doing one for your own experimentation, and can deal with the fallout of it not quite working, or on the other hand going to sell the thing. I design pedals for commercial production in my day job. I've quit ever specifying electros less than 25V for anything, simply because of the number of warranty repairs that came with a note that "Hey, I read on the internet that this thing sounded magic if you run it from 18V, and it did for a few minutes, then it died."

I personally won't go to lower voltages than the power supply, but there is some reason to expect that it might work OK on most situations. That is, it's a question of your objectives, not what is deemed OK by an internet forum.

[ElectricDruid]

+1 agree with RG. There's really no need to use electrolytics that are under-rated. I don't think I've even *bought* anything rated 16V for years - why would I when it makes more sense to get a 25V or 50V part that I could use anywhere?

That said, you'll probably get away with it. If it's a one-off unit, go for it. The worst that can happen is it screws up and you have to fix it. If it's a production design, that's a major headache and a whole different story.

T.

[vigilante397]

why would I when it makes more sense to get a 25V or 50V part that I could use anywhere?

Especially when it's literally pennies (generally one or two) of difference in cost.

[thermionix]

16v source to ground cap should be more than fine.

"Never" is a long time

I thought "forever" was a long time.  Does that make "never" no time at all?

[antonis]

16v source to ground cap should be more than fine.

On 18V PS..?? 

You probably mean JET's Source..

[Frank_NH]

I used my JFET spreadsheet to do a quick calculation for a J201 JFET common source stage.  If you are operating at a 9V source and have the drain at 4.5 V, and then change the source to 18 V and want to have the drain at 9V, then the drain resistor needs to increase by the same factor as the voltage.  So, for example, for a 1K source resistor, J201 JFET, at 9V source, 4.5V drain, the drain resistor would be about 16K.  For 18V source/9V drain, it goes to 32K.  So you will need to adjust your drain resistance when you change to 18V, and hopefully your trim pots have enough range to accommodate.  Hope this helps.   

[vigilante397]

I thought "forever" was a long time.  Does that make "never" no time at all?

Forever = for the duration of ever

Never = not ever

Ever is a long time

[Andrekp]

OK.

So if I have a circuit that, from ground to the power rail is:

GRND => 2K7 => J201 => Bias trimmer => 9v

I adjust the bias trimmer so that the J201's drain is at 4.5v and get the results I want.

My understanding of what we have here is essentially a voltage divider.  4.5v must be being dropped across the 2.7K/J201 combination, with 4.5v being dropped across the trimmer.  The trim I have done in this example, put 1/2 the voltage across each half of the sub-circuit and biases the drain of the J201 at 1/2*9v.  That much is clear.

What I wonder about, and is the root of my question, is whether the IMPORTANT part here (speaking generally) is that the JFet is biased to some particular voltage, or whether it is that it is biased to 1/2 the power rail.  In other words, in a given circuit, where proper bias is the drain being at 4.5v with a 9v rail, can it be assumed that in an 18 volt circuit, proper bias would be, similarly, drain at 9v, with an 18v power rail?

I presume that this fact is generally true since there are lots of pedals that DON'T have to be tweaked in any way when you switch from 9v power to 18v power.   The voltages go up at points in the circuit, but they remain relative in the same way, so provided component limits are not exceeded, the pedal still works as expected, albeit with more headroom, or whatever the extra voltage might bring.

So IF that is true generally (and I'm trying to learn here if it's not), why would I have to change any resistors in my circuit just because I switched to 18v?  Sure, the voltage drop would increase across each component, but the relative drops would remain the same, because the relative resistance would remain the same.  It seems to me that if I have 4.5v at a given point with a 9v rail, that I'd expect 9v with an 18v rail? 

Is this just incorrect?  Or is there something about biasing JFets that I have to take into account as well?  (I note again that some of my presumptions stem from the fact that I never see anyone talking about making changes to circuits or biasing when they talk about running things at different voltages - other that worrying about whether the caps are OK at that voltage - this led me to believe it just wasn't an issue, so long as the relative values, and therefore voltage drops,  remain the same.)

That was what was involved in my first question.  My second question was related:

Given the same circuit, but now with a cap in parallel with the 2k7:

GRND => 2K7 & 330u cap (in parallel) => J201 => Bias trimmer => 9v

Wouldn't the voltage drop across the cap ALWAYS equal (Drain Voltage) - (Voltage across the J201), or equal to the voltage across the 2k7?   Given that we have set up the drain voltage to be 4.5v wouldn't the voltage the cap is subject to always be less than 4.5v.  Isn't that Electricity 101.  Given that is true, wouldn't it ALSO follow that if I increased the rail voltage to 18v, without adjusting anything else, the the voltage drop across the 2K7, and thus across the cap, would only be some amount less that 9v? (both cases assuming SOME voltage drop across the JFet)

In my ACTUAL circuit, this led me to question whether it would actually be safe to run a 16v cap, IN THAT POSITION, with an 18v rail, since it at most has 4.5 volts across it at 9v, leading to 9v across at 18v rail?

This all seems pretty basic to me.  The potential fly in the ointment is the behavior of the JFet, if it's not what I am assuming. 

Question 1 can be distilled to: If a designer determines that with a 9v rail, the best operation of the JFet is 4.5 volts at the drain, is the best operation of the JFet a function of particular voltage (it wants to see 4.5v on the drain, regardless of power rail voltage), or the relative voltage (it wants to see approximately 1/2 power rail voltage, whatever it is).

Question 2 can be: Would some action made by the JFet under normal function ever cause the cap, in that position, to exceed its rated voltage when the circuit runs at 18v?  Can it be essentially safe to run a 16v cap in an 18v circuit IF that cap is somewhere that only ever sees less that 9v?  Even if the voltage drop across the JFet became 0, would the cap ever exceed it's rating?

I think I am being more clear now, if not I apologize.

[vigilante397]

My understanding of what we have here is essentially a voltage divider.  4.5v must be being dropped across the 2.7K/J201 combination, with 4.5v being dropped across the trimmer.  The trim I have done in this example, put 1/2 the voltage across each half of the sub-circuit and biases the drain of the J201 at 1/2*9v.  That much is clear.

What I wonder about, and is the root of my question, is whether the IMPORTANT part here (speaking generally) is that the JFet is biased to some particular voltage, or whether it is that it is biased to 1/2 the power rail.  In other words, in a given circuit, where proper bias is the drain being at 4.5v with a 9v rail, can it be assumed that in an 18 volt circuit, proper bias would be, similarly, drain at 9v, with an 18v power rail?

Excellent deduction. Yes, 1/2 of your power rail is a fairly standard JFET bias point. Many (I would wager most) will bias to that, whereas some may bias to that initially as a starting point then tweak from there using their ears instead of a multimeter to see if there is a "sweet spot" outside of 1/2 VCC.

As question #2 regarding cap rating, you are correct. If there is a spot you can reasonably assume will not see more than its rated voltage then you can reasonably put a lower rated cap there. Personally the main reason I (and possibly others) don't usually do this is that it is much easier to buy one rating (like 25 or 50V) and put them everywhere than to worry about what can go where without causing problems. But if you only have 16V rated caps you can safely put them in a bias path that will see 1/2 VCC. HOWEVER if this pedal is being built for someone else there is the obvious possibility of damage if they "want to see how it sounds" with the bias voltage all the way up.

[Andrekp]

My understanding of what we have here is essentially a voltage divider.  4.5v must be being dropped across the 2.7K/J201 combination, with 4.5v being dropped across the trimmer.  The trim I have done in this example, put 1/2 the voltage across each half of the sub-circuit and biases the drain of the J201 at 1/2*9v.  That much is clear.

What I wonder about, and is the root of my question, is whether the IMPORTANT part here (speaking generally) is that the JFet is biased to some particular voltage, or whether it is that it is biased to 1/2 the power rail.  In other words, in a given circuit, where proper bias is the drain being at 4.5v with a 9v rail, can it be assumed that in an 18 volt circuit, proper bias would be, similarly, drain at 9v, with an 18v power rail?

Excellent deduction. Yes, 1/2 of your power rail is a fairly standard JFET bias point. Many (I would wager most) will bias to that, whereas some may bias to that initially as a starting point then tweak from there using their ears instead of a multimeter to see if there is a "sweet spot" outside of 1/2 VCC.

As question #2 regarding cap rating, you are correct. If there is a spot you can reasonably assume will not see more than its rated voltage then you can reasonably put a lower rated cap there. Personally the main reason I (and possibly others) don't usually do this is that it is much easier to buy one rating (like 25 or 50V) and put them everywhere than to worry about what can go where without causing problems. But if you only have 16V rated caps you can safely put them in a bias path that will see 1/2 VCC. HOWEVER if this pedal is being built for someone else there is the obvious possibility of damage if they "want to see how it sounds" with the bias voltage all the way up.

Thank you for answering my question.  I figured that was the case.


Sent from my iPad using Tapatalk

[reddesert]

So IF that is true generally (and I'm trying to learn here if it's not), why would I have to change any resistors in my circuit just because I switched to 18v?  Sure, the voltage drop would increase across each component, but the relative drops would remain the same, because the relative resistance would remain the same.  It seems to me that if I have 4.5v at a given point with a 9v rail, that I'd expect 9v with an 18v rail? 

Is this just incorrect?  Or is there something about biasing JFets that I have to take into account as well?  ...

The JFET is not an ohmic device - that is, it isn't guaranteed to have a linear relation of current to voltage drop. Resistors are ohmic, but semiconductors often aren't.

If you change the supply voltage across something that is ohmic, like a voltage divider, volume pot, etc, then it works the same way in a relative sense and your logic is OK.

Because the voltage drop across the JFET is not guaranteed to double when you double the supply voltage, it might bias differently when you run it at 18V rather than 9V. Will this matter? It depends. There is no ironclad law that you have to bias a JFET drain at half the supply voltage. Clearly you want to bias the drain somewhere near the middle of the supply to have headroom - ie if you run off a 9V supply, bias at 4.5 V, then you can run a +/-2 V signal through without getting close to the supply rails. But maybe it would sound better biased at 5.5 V - that's a matter of taste.

Because JFETs are quite variable, it is really hard for anyone to tell you whether the circuit will work well if you simply double the supply voltage. Having a trimpot on the drain resistor would be prudent if you're concerned about it.