Saturday, January 31, 2009

What is a voltage divider? Pt3

Part 1
of this post showed a voltage divider made with two resistors connected end to end:

It's a useful circuit, but it's action is fixed - determined be the value of the resistors chosen for the circuit. A potentiometer (commonly called a "pot") is a variable resistor and it can easily be wired to form a variable voltage divider.

A pot generally has three pins. The outer two pins connect to the opposites ends of the pot's resistive track. The resistance from one end of a pot to the other never changes. In this example it's 500 ohms.

The pot's center pin is connected to the wiper. The wiper moves along the resistive track as the pot is turned. Because the wiper simply splits total resistance into two parts, the sum of those two resistances is always equal to the total resistance from end to end. In this case 250 ohms + 250 ohms = 500 ohms.

To the circuit this looks like two 250 ohm resistors connected in series. Putting volts across the ends of the pot as we did in part 1 causes 1.5 volts to appear from each end of the pot to the wiper.

So what happens when the pot is turned?

The resistance from end to end stays the same but the way that resistance is divided by the wiper changes:

Here the position of the wiper makes the pot function like a 100 ohm and a 400 ohm resistor connected in series. Following the rules from part 1 again, 3 volts in will produce 2.4 volts out.

If the pot is turned toward the other end of it's rotation and the resistances are reversed:

Now 3 volts at the input yields only 0.6 volts at the output:

This is just what is happening when you turn down the volume knob on your guitar or amplifier.

The Epiphone Valve Junior has both a fixed voltage divider for reducing the gain between stages and a variable voltage divider (the volume pot) wired in parallel. In the next post I'll take a look at that circuit as a real world example.

Friday, January 30, 2009

What is a voltage divider? Pt 2

To help visualize the concept from part 1 of this post, here's a quick circuit mock up of the circuit diagrammed there. On the left is a 10 ohm resistor with the left hand meter's probes connected across it. Connected in series to this resistor is a 90 ohm one with the right hand meters leads connected across it.

Below the meters are reading the actual values of their respective resistors (10.4 ohms to the left and 91.4 ohms to the right). This works out to pretty much the 10% and 90% of total resistance as shown in part 1.

Now the power supply is connected to the circuit and turned on.

The left hand resistor is seeing o.233 volts and the right hand one is seeing 2.112 volts. The sum of the voltage drops across the two resistors is 2.345 volts.

o.233 divided by 2.345 is .101 or about 10%
divided by 2.345 is .901 or about 90%

The voltage is raised to about 4.5 volts on the power supply.

Now the left hand resistor is seeing 0.439 volts and the right hand one is seeing 3.99 volts. The sum of those voltage drops 4.429 volts.

0.439 divided by 4.429 is .099 or about 10%
divided by 4.429 is .903 or about 90%

The voltage is raised again to about 10 volts on the power supply.

Now the left hand resistor is seeing 1.010 volts and the right hand one is seeing 9.05 volts. The sum of those voltage drops 10.06 volts.

1.01 divided by 10.06 is .1 or about 10%
divided by 10.06 is .899 or about 90%

So no matter how the voltage changes the ratio between the voltages across the resistors stays the same.

With an output connected across the 90 ohm resistor (as shown in part 1), any voltage input to the circuit will produce 90% of that voltage at the output. Now you can see how the voltage divider is effective for the constantly changing AC signal generated by your guitar as well as for a simple DC circuit.

What is a voltage divider? Pt1

You'll find voltage dividers in a whole host of places throughout your amp, your pedals and even in your guitar itself.

So what is it? And what does it do?

There are other applications, but one common use is as attenuators to reduce the voltage between stages. Once you understand the basic concept it's use in more complicated instances (like the divider formed with your plate load resistors) will be much easier to grasp.

The simplest sort is formed with two resistors hooked up in series:

You'll see that the total resistance from the top of the first resistor (R1) to the bottom of the second (R2) is just the sum of the two resistors (Rtotal).

Rtotal makes up 100% of the total resistance of the circuit. The 10 ohm resistor R1 makes up one tenth of the total resistance - so .1 or 10%. R2 makes up the remainder, or 90%.

So let's put a voltage across the two resistors, say 3 Volts:

The full voltage, 3 volts, could be measured across the two resistors. If you add the voltages that appear across each of the resistors individually they must equal the total voltage of 3 volts. How much voltage appears across each resistor?

Take resistor R1. It makes up 10% of the total resistance of the circuit. Take 10% of the total voltage (that's 3 times .1) and you'll get the voltage drop across R1. That's .3 volts.

Do the same for R2 (3 times .9) and you'll get 2.7 volts.

Add those two and you'll get 3 volts. So things check out.

Now because we're looking at voltage not current we can change the resistances and get the same results as long as the ratio between R1 and R2 stays the same.

Here's the same circuit with R1=50 ohms and R2=450 ohms:

When you do the math:

R1 divided by Rtotal = 500/50 = .1 or 10%
R2 divided by Rtotal =500/450 = .9 or 90%

You can see this is the same as in the previous example.

So why does all this matter? Let's imagine the resistors in a more complete circuit:

Here imagine that the input is coming from one amplifier stage the output is feeding the next stage. If we look again at how the voltages divide across the resistors:

You'll see that with 3 volts of input you a .3 volt drop and the output voltage is reduced to 2.7 volts. Changing the values of the resistors can change the amount of drop. If you use a variable resistor (a potentiometer or "pot") you can vary the drop by turning the knob on the pot. This is exactly how the volume knobs on you guitar and amplifier work. I'll cover that in more detail in part 3.

Friday, January 23, 2009

Fender Champion 600 - Discharging the Filter Caps Pt 2

Here's a more detailed look at discharging the filter caps in a Fender Chapion 600.

First I'll get this out of the way:


With the amp turned off connect the black lead of the multimeter to ground at the speaker jack. This will be the bare wire connected to the sleeve connection on the jack. You can use any ground connection really, but this on is very easy to get a clip onto. Connect the red lead to the top of R11. Set the meter to read DC Volts. Now flip the amp on.

You'll see the voltage on the meter rise quickly to around 420 volts. If you don't see any voltage reading then you've connected things incorrectly or you meter is not set properly. Don't proceed until you've figured out why you have no reading, and be very careful when when moving the probes as there is now very high voltage in the circuit.

When you have the meter connected properly and are getting a 400+ VDC reading shut off and unplug the amp.

When you shut the amp off you will see the voltage dropping slowly on the meter. You are now ready to connect the clip end of the discharge probe to ground. I find the easiest method is to clip it to the sleeve of the input jack.

Mote that the voltage is still dropping slowly even though the probe hasn't been fully connected. This is because the capacitor is bleeding it's charge though the circuit and through the meter itself. This is exactly what we want the discharge probe to do - but faster.

Now touch the other end of the discharge probe to a source of B+ voltage. Here I'm about to make contact with the red lead of the output transformer primary:

Now you will see the voltage drop quite rapidly. This picture covers the change in just a few seconds (click on it for a close up):

Finally all the voltage will be drained out and you're ready to work on you amp.

Note: Turning the amp on is not really not part of the discharging process. The reason I include it is because it insures that you are getting a good voltage reading and that you can watch that reading drop. That way when your meter shows no voltage you know it's because there is no voltage left in the caps not because the meter isn't connected properly.

Second note: If your speaker is still connected to the amp during this procedure you will notice that the voltage drops fairly quickly. This is because the connected speaker is doing the job of the discharge probe. This is actually a perfectly reasonable way to avoid the discharge probe all together. Discharging with the probe is an excellent habit to get into though because not all amps bleed off their caps voltage so quickly on their own. If you decide to forgo the probe, make sure you are absolutely confident in your voltage measurements in order to stay safe.

Fender Champion 600 - Discharging the Filter Caps Pt 1

Here's the procedure for discharging the filter caps in a Fender Champion 600.

The video is pretty small so I'll put up another post with pictures so the detail is a bit clearer.

I'll cover the construction of the probe in a later post.


Discharging Filter Caps - The Basics Pt 2

Here's hows how a discharge probe is used to bleed voltage off of a capacitor. This cap is completely removed from the circuit for illustrational purposes, but the procedure is the same when a cap is installed in an amplifier. The meter is tough to read here but I hope it's good enough to get the general idea.

The resistor at the clip end of the probe is a 56K 3 Watt one covered in two layers of shrink tubing for safety. The resistor will limit the current from a 450 Volt supply to under 10 ma but it is still a very good idea not touch the alligator clip end of the probe while connecting the other end to the B+. I'll cover the probe construction in a later post.


Discharging Filter Caps - The Basics Pt 1

Here are the basic reasons and methods for discharging filter caps. The meter is a little tough to read here but I think it's clear enough to get the basic idea. This is just for illustrational purposes, it's NOT the the method I recommend. I'll cover that in part 2.


Thursday, January 22, 2009

Fender Champion 600 Free Installation of the Mercury Magnetics Mod

Doing the Champion 600 mods in my previous posts left me with a good bit of curiosity about the specifics of the Mercury Magnetics kit. If anyone in the Boston area has a Champion 600 with a Mercury Magentics mod kit they need to have installed, drop me an email. For the opportunity to get in there and see just what's going on I'd be glad to do the put it free of charge.

Just putting it out there in case anyone's interested...

Monday, January 19, 2009

How to Remove Tubes / Another Way PC Board Fail

The first half of this video shows how to remove preamp tube from your amplifier.
The second half shows a rather extreme example of what can be happening to the circuit board in the process.

To be fair, most pc board mounted sockets are better supported than this one from an Ampeg V4B. The same principle applies though. It doesn't take much motion to make a hairline crack in on of the solder joints on the other side of the board. The worst thing is that these tiny cracks are likely to cause intermittent failures. So while the amp may work most of the time you won't be able to trust it to gig with or bring a recording session.

Saturday, January 17, 2009

Tube Amps and the Freezing Cold

I suppose this one's from the "do as I say not as I do" file. This is what can happen when you leave your amp in the trunk in the freezing cold. I was playing a gig that ended in the middle of a snow storm and I left my amp (a custom thing built on a Teisco Checkmate 50 chassis) in the trunk for a few days.

In the morning the trunk got warm enough for water to condense on the cold transformers and then got cold enough for it to freeze there.

Of course one would worry about rust and moisture damage, but the real danger here is not the frost and condensation on the transformers. It's the danger of the amp being switched on with that condensed water shorting the circuit inside the amp.

This can happen as a result of condensation inside the chassis or by water dripping there from the outside. Switching an amp on in this state is a great recipe for arcing and damage inside the amp.

If you are gigging in cold weather it's a very good idea to keep your amp covered in the trunk or van and ideally keep it covered while it comes up to room temp to avoid condensation. If you bring a cold amp into a warm humid room it will come up to room temperature in a fairly short period, but the water that condenses on the cold transformers and chassis can remain for hours or more. The above picture is of an unusual case. Remember, the condensation generally occurs after the amp has been brought inside. It may be perfectly dry coming in from the van. After 10 minutes in the indoors it can become dripping wet.

Now of course this has been happening to amps for decades and they haven't all gone up in smoke. But the danger is there. If you really want to play it safe you could build a simple series outlet current limiter and keep it in you gig box. Use that when you warm the amp up (and dry it off) after a particularly cold trip.

Friday, January 16, 2009

Fender Champion 600 Fat Boost Mod Resistor Values

In the Fat Switch Mod the circuit sees 3 different resistances for the mid resistor in the tone stack: 15K (stock), 30K (Frondelli Mod fat boost value), and 47K (for a little extra boost). The actual resistors on the switch are quite different values. The 47K value is there but the other two are 68K and 22K. Why not the 15K and 30K that the circuit needs to "see" for the mod?

In order to use a simpler switch I approached the mod a bit differently. I decided to replace the standard mid resistor with a 47K one. This sets the max mid resistor value. The fat switch then selects one of two resistors and connects it in parallel with the 47K resistor, lowering the effective resistance. In the center position both of the additional resistors are disconnected so the total resistance remains 47K.

This should be clear from a schematic drawing:

Too find the effective resistance when one of those resistors is switched in use the formula for finding the total resistance of any number of resistors connected in parallel. Incidentally, this is the same formula you would use when connecting speakers in parallel:

Since we only have two resistors connected at any time, it's a bit simpler. All we need is R1 and R2. Here's how the formula for the 47K resistor in parallel with the switched in 68K resistor is solved in detail:

That 28K value is plenty close to the Frondelli Mod value of 30K. If you're wondering how close, take a look at the graph at the bottom of the Fat Switch Mod post. You'll see from comparing the curves for the three fat boost resistor values that that 2K difference doesn't matter much.

Here's the same equation for the 22K resistor in parallel with the 47K one:

Solve that equation and you'll see that the 22K in parallel with the 47K results in 15K - the same effective value as the original R19. So switching in the 22K resistor puts the tone stack back to stock.

Wednesday, January 14, 2009

Reading Capacitor Values Part 1

Reading capacitors values from schematics or even parts lists can be confusing. There are a few basic rules to keep in mind that will make it a bit less daunting. First off the basic unit of capacitance is the farad. A farad is so large that you will never encounter a capacitor measured in farads. Most capacitors in tube amps are measured in microfarads. It takes one million microfarads to equal one farad. If a unit for a capacitor in a tube amp schematic is not specified the value is almost certainly microfarads (abbreviated uF). The rest will for the most part be picofarads (pF). I'll take some examples from the Fender Bassman 5F6A. Below is the complete schematic.

Here's an excerpt showing the 5F6A tone stack. The cap show in red is labeled .02-400. The first number indicates that the capacitance is .02 microfarads, the second number indicates the voltage rating, 400 volts in this case.
The treble cap is a much smaller value, .00025 microfarads. If you're looking though a parts catalog you are not likely to find a .00025 microfarad cap. That's because a cap of this size will more often be expressed in picofarads (abbreviated pF). There are on million picofarads in one micofarads. To find the value of a cap rated in micofarads in picofarads simply multiply by 1,000,000.

In this case .00025 microfarads x 1,000,000 = 250 pF.

Looking at the splitter of the phase inverter, there is a cap bridging the output of the inverter:

This cap is shown as at 47MMF. What does this mean? I've said already that the caps in tube amps are going to be fairly exclusively rated in microfarads (uF) or picofarads (pF). What does MMF mean?

In short, it is the same as picofards (pF). To understand why takes a bit of explanation. The abbreviation for "micro" is taken from the lower case of the Greek letter "mu".

Now you can see that the upper case "mu" looks just like an "M".
But what about the lower case"mu"?

Frequently you're not using the symbol for "mu" especially if you're typing. Convention has "mu" written as "u" so microfards is written uF. But in a way it's logical to think of the lower case "mu" as "m". This is why you'll sometimes see microfarads written as mF. Technically mF means millifarads (one thousandth of a farad) but since caps are not generally rated in millifarads you can safeley assume that if you see mF it is microfarads that is indicated.

MF techically means megafarad - one million farads. This is a value so ridiculously large that a cap this large could not even begin to be manufactured. If you see MF you can also assume microfarads (uF).

So the MMF from above can be considered to be mmF or micromicrofarads. "Micro" means "one millionth" or "divided by 1,000,000). A picofarad is a microfarad divided by 1,000,000. So saying micromicrofarad is really the same as saying picofarad.

So to sum up:

The majority of tube amp caps will be indicated in microfarads (uF, mF, MF) and picofarads (pf, mmf, MMF). Occasional you will see nanofarads (nF). A nanofarad is simply one thousanth of one microfarad or one thousand picofarads.

If a unit for a capacitor in a tube amp schematic is not specified the value is almost certainly microfarads.

Thursday, January 8, 2009

Fender Champion 600 Cathode Bypass Mod Pt 3

Here's the response curve for the switched cathode bypass caps in part 2 of this post. The first graph is for the stock tone stack:

From top to bottom the curves show:

6V6, 1st stage (12AU7) and 2nd stage (12AX7) engaged (stock).
6V6 and 1st stage (12AU7) engaged.
6V6 bypass cap engaged.
All bypass caps disconnected.

The curves in the next graph follow the same pattern but this time the Tone Stack Bypass is engaged. There is a great deal more gain with the tone stack bypassed. If you click to enlarge the graphs and look at the right hand scale you'll see that the peak response in the modified amp is about 10dB higher than in the stock one.

This amp has the 12DW7 mod installed so the 1st stage has less gain and correspondingly less increase in gain with the cap engaged. In a stock amp the jump from the aqua curve to the green curve would be increased to about the same as the jump between the green curve and the dark purple curve on top.

Tuesday, January 6, 2009

Fender Champion 600 cathode bypass mod Pt 2

When I wrote part 1 of this post I'd been planning to lower the cathode bypass capacitor values in order to clean up the up the bass response. What I ended up doing was removing the cathode bypass capacitors all together in order to get a bit of compression for a fuller low volume clean tone. The effect is a more roundness and a good deal less volume - very nice for a bedroom level clean sound. This meant I needed to find some other ways to control the low end (see the Input Voicing and Presence Plus mods).

For this mod I put in a three pole four position rotary switch that adds the cathode bypass caps in one by one. With the rotary switch the gain of the amp goes up with each setting until it reaches the stock (fully bypassed) position:

This could easily have been three separate switches, but I thought the single control made the operation a bit clearer and didn't clutter up the chassis as much. Since I'm limiting bass response in a couple other ways I ended up keeping the caps at their stock values. The first and second stages could easily have values from part 1 substituted if you have a need for greater bass reduction.

Here is:


And here's the internal view of the switch with the ground wires in place:

This is how the ground wires hook up:

The wire with the black arrow connects to the lower side of the C4 space on the pc board.

The wire with the blue arrow connects to the lower side of the C10 space on the pc board.

The wire with the red arrow connects to the lower side of the C3 space on the pc board.

Here is the rest of the switch wiring. Notice that the leads for the large blue cap (which serves as C4) jump three pins on the switch.

The stage 2 wire jumper two pins. This wire attaches to the shrinkwrapped end of C10 (indicated by the red arrow).

The wire connecting stage 1 to the switch connects to only one pin. It connects to the shrinkwrapped end of C3 (again, indicated by the red arrow).

If you've removed these caps and are reinstalling them, make sure you observe the proper polarity when you put them back in. The indented end of the cap lines up with the indent in the white outline on the pc board.

For convenience of installation, I replaced the original C4 with a axial lead cap of the same value. It's the large blue cap in the photo below. The negative side of the cap connects to the switch and the positive side is connected to the high side of the pc board connection for C4. The negative lead of the cap holds it pretty well in place but there's a dab of silicone underneath just for good measure.

This switch functions kind of like a staged clean master volume. If you're interested in getting distortion out of your preamp circuit, this mod could be rearranged to serve as a sort of dirty master volume instead. If you're interested let me know and I'll post the details.

Monday, January 5, 2009

Fender Champion 600 Input Voicing Mod Pt 2

Here's a graph showing the frequency response for the Input Voicing Mod. The measurement is taken at the amplifier output with a 4 ohm resistive load.

The top curve is the stock "High" input and the bottom curve is the stock "Low" input. The middle curve shows the response of the "Low" input with the mod installed. Notice that the modded "Low" input is within 1dB of the "High" input for the high frequencies but the lows are rolled off significantly.

I didn't include the modded "High" input response on the graph because it is so close to the unmodded "High" response that it's almost indistinguishable. I would swear I hear a subtle difference between the modded and unmodded "High" input. This may be because the tone generator used for the frequency sweep is non inductive and doesn't react to the input impedance the way a pickup does. Or I may be crazy. If you do this mod yourself, let me know what you hear.

Saturday, January 3, 2009

Fender Champion 600 Tone Stack Bypass/ Fat Switch Mod

This pic shows the shows a few of the mods I've made to this Champion 600. The two small red switches on the bottom of the chassis are for my Presence Plus and Input Voicing modifications. This particular post concerns the tone stack mod associated with the two bigger switches to the upper right.

This mod takes the coupling cap mod from the Tone Stack Test and hardwires it to a switch so the amp can use the stock tonestack or bypass the tonestack with a single coupling cap to eliminate the mid cut in the stock circuit. The right had switch below handles that job. The left switch acts as a three position "fat" switch.



Here's the finished mod from above:

The combination of resistors on the "fat" switch are selected to combine for three settings:
15K (stock), 30K (Frondelli Mod fat boost value), and 47K (for a little extra boost). The boosts effect the mids most dramatically, but they provide extra gain across the whole spectrum too.

All three resistors connect to the same pin on the right hand switch - the one just to the left of the center pin with the black wire.

The switches I used here each have one more set of contacts than are needed to make the mod. I frequently do this when I'm experimenting so that if I decide to add something to the switch later on I don't need to disassemble the circuit and solder in a new one.

The resistors in the fat boost circuit replace R19 on the pc board. Here's how the wiring runs:

Here's the coupling cap, prepared for installation:

And here it is soldered in place. Since only one end is really fixed and the other will be supporting a wire I put a bead of silicone underneath to make sure it stays in place.

Here's how the wires are run to the switch. Point "A" on the switch runs to point "A" on the board. Same for "B" of course.

Here's the frequency response for the four settings measured at the amp output with a 4 ohm resistive load:

Click on the graph for a high res version. From the bottom to the top the curves are for Stock, Fondelli Mod Fat Boost, Extra Fat Boost and Tone Stack Bypass. The tone stack bypass curve is up about 16dB up at the mid cut frequency!

A final note...if you try this one yourself, be careful to place the switches low enough so they clear the cabinet when you put the chassis back in. It a tight fit.