Problem with presence and resonance

[defaced]

A wee bit of background. I'm working on taking a Peavey Deuce VT and making it an all tube power amp. I've gutted the thing, added a phase inverter and one preamp stage - all of which works fine. I started working on the NFB loop with presence and resonance controls a few days ago and I can not get this thing to work. I've read everything I can find on presence and resonance controls, looked a numerous schematics, and I can not figure out where the problem is. Regardless of the position of either of the pots, the output does not sound any different. I have also tried various connection and grounding configurations based on various things I've read and none of them have resolved the problem.

I'm using the Peavey JSX phase inverter and NFB circuit as my guide. I've attached the schematic, the NFB is at the bottom left of the first page, the PI is on the 8th page.

If anyone has any input, I'm all ears.

[cbarrow7625]

You could just have an issue with a gain difference in the two output tube circuits. The resonance & presence controls rely on having enough open loop gain that you are throwing some away to reduce the overall gain of the amp. Both the presence & resonance controls are frequency dependent circuits that "bring back" into the circuit some of that gain that you would otherwise be throwing away.

You may need to do something to increase open loop gain. You may also want to consider reducing the 68K feedback resistor and/or increaseing the value of the 4.7k tail resistor to adjust the voltage divider ratio for the feedback.

Disconnect the feedback altogether & measure open loop vs. closed loop gain (voltage) and see if there is a difference. If the difference is small, there is your problem.

Good Luck. The wiring looks OK from what I can tell.

Chris

[loudthud]

A couple of things I noticed. First off, the Deuce uses the 4 ohm tap for feedback and the JSX uses the 8 ohm tap. Try the 8 ohm tap. Second, the cap C65 is shown on the JSX schematic with it's polarity reversed. The + side should go the Phase inverter.

To work properly, the power amp needs about 10 times as much gain with the feedback disconnected.

[defaced]

Thank you both for your suggestions. I have the resonance control working, but the presence is still dead. This is with the feedback resistor at 250K and the tail resistor at 130K.

I noted the polarity of that cap (it's different than what's on the 5150 II schematic I've also been using as reference), as well as the difference in the taps. I've been using the 8 ohm tap for most of my trials. Swapping the cap's polarity did nothing, the tap I haven't played with in a few days, and since I'm using the 8 ohm tap, I'll just leave it as is.

The other thing I've played with is what ground everything goes to. On the JSX, the speaker goes to earth ground, while the presence goes to signal ground - nothing there either.

Between what two point should I be taking a voltage reading so I can calculate gain? I'm not sure I'm taking the readings at the correct location because they don't make since.

[cbarrow7625]

Take the voltage redings at the input to the phase inverter & the output tap where the feedback is connected.

Have you tried just switching in / soldering in the presence capacitor? Maybe your pot is just bad. Could be a bad cap too.

I think you might want to scale down your feedback / tail resistors by a factor of 10 (25K / 13K). 130K for that resistor in the PI is going to mess up the PI operation quite a bit. Even better, scale them down by a factor of 20(12.5K / 6.5K - or somewhere close). That is going to get you a lot closer to the preferred tail resistor value of 4.7k. The scaling factor shouldn't make much difference since the feedback is coming from a very low impedance source (8 ohms). The smaller the feedback resistor, the more dramatic a difference your deep control should make.

Also, with a smaller feedback resistor you will likely have to use a larger cap in the presence circuit. The combination of the series feedback resistor & the presence cap make a low pass circuit for the feedbackloop. Just like anyhwere else, if you change the resitor size you also need to change the capacitor size to keep the same frequency response.

The polarity of the cap shouldn't do anything to the resposne. It will just help ensure that the cap doesn't get fried because the DC potential is going through it the wrong way. the + side should always be toward the highest DC potential. For AC, the cap doesn't care.

Let us know how it goes.

Chris

[loudthud]

One thing I noticed is that you have the resonance control connected backwards. In the fully clockwise position, there is no boost of low frequencies. Move the center terminal connection to the other side.

An easy way to measure the feedback factor is as follows. Adjust the resonance control so it shorts out the capacitor and the presence control so it shorts it's capacitor to ground. Connect a signal generator to the input of the phase inverter and a dummy load resistor to the output. Monitor the output with a voltmeter. Now adjust the generator for a small output (1V on the voltmeter) at some mid frequency of 200Hz to 1KHz. Now, disconnect the feedback and the reading on the voltmeter should (ideally) pop up to something like 10V. If not, adjust the 68K resistor until you get a 10 to 1 differential. If you don't have a signal generator, you could use a noise source such as the output of an old FM radio where you can hear the hiss between stations.

[defaced]

Alright, here's what I've got. It doesn't make since, but it's what I've got.

With a 500Hz sine wave, at 1v on the input of the phase inverter, with the feedback loop disconnected, I get 2.8v at the 8 ohm output tap of the OT. With the feedback loop engaged, as I lower the feedback resistor, I can get the output up to 20.1v, but I can hear the tubes oscillate quite loudly (this is with the feedback resistor as a dead short). With a feedback resistor size of 6.8k the output voltage is 18v and the oscillation at the tubes is barely audible. If I maintain the same feedback/tail resistor ratio I had before of 250k/130k, I get 16v at the output. If I shoot to get the output voltage in the 6-14v range, it is not stable.

Now if I go the other way with the feedback resistor, up to 250k, there is no change in the output voltage whether the NFB is engaged and not.

This sounds counter to what LT is saying, but with these conditions, the resonance works, the presence does not. However, when the resonance control is maxed, the amp honks, which is not a good thing.

I've tried several different pots (10k to 1M) and many different caps (.15pf to 470uf). I was hoping it was a component issue, but I've not been so lucky. I did find that one of my 1M audio taper pots was bad though.

[loudthud]

Something is going wrong here. I suspect that the amplifier is oscillating, probably at an ultrasonic frequency that you cannot hear from a speaker. The usual cause of this on a new build is that the phase of the feedback is wrong. Unfortunately, the Deuce schematic doesn't give us the color of the wires on the transformer primary. The usual fix for this is to reverse the plate wires from the transformer to the 6L6's but you may also reverse the two wires going to the inputs of the 6L6's at the circuit board.

Since (I'm assuming) there is no oscilloscope available, we will have to use the voltmeter to detect the oscillation. With no signal input to the amp, and the meter connected to the output of the amp, the meter should have a very low reading, something below 0.1V AC. Even if the OT is connected properly, there is probably some setting of the "68K" pot (the feedback resistor) that will cause the amp to oscillate. The amp may also oscillate when the presence control is turned up. So disconnect the feedback and try this test.

Hopefully the amplifier does not oscillate when the feedback is disconnected. That is a lot harder to fix. It's usually caused by wires being too long or a problem in the power supply or a bad ground. We'll have to fix this before we can go any further.

Now connect the generator. Adjust it so there is 10V on the output of the amp. Now when the feedback is connected, (Check with an ohm meter that there is a short across the presence and resonance caps) the output should go down to 1V at some setting of the "68K" pot and it should be close to 68K or maybe 47K. If you can't get there, we'll have to look at the resistor values in the phase inverter.

[defaced]

I was reading the Aiken's amp page (link) about global negative feedback, and came to the same conclusion as you: the feedback is in phase, not out of phase. I reversed the feedback wires that I had connected to the jack, and now as I lower the feedback resistor, I can see a reduction in voltage. With no feedback applied, it's 3.3v, with maximum feedback (Rf = 0), the output voltage is 0.9v.

Depending on how Rf is adjusted, I can get both controls to do something, I wouldn't necessarily call them presence and resonance yet, but they do something.

[loudthud]

Ahhh, PROGRESS! Now, you should re-examine the values in the phase inverter. I would lean towards the more standard values like the ones in Aiken's schematic and a little more B+ would probably help. This will give the controls a little more range and then you can tweek the cap values to suit your taste.

[defaced]

I'm still tweaking.

I've tried to raise the B+ and am not getting much. Between pins 1 & 6 and ground, I see between 200v and 220v depending on source voltage and resistor. It also doesn't seem like current draw is a constant, but also a dynamic element. Is there an equation or curve that dictates this?

I also can't get less than 0.9v at the output, and at that point the amp oscillates, but not the same way it did when I added positive feedback. This time, something inside the tube starts to pulse and the feedback loop breaks down and the voltage across the speaker jack starts to raise. Does this mean I've hit the limit on the amount of feedback possible? Would the screen voltage affect this, as the Deuce used 100 ohm screen resistors on two of the tubes, the other two are with out resistors?

I've also been playing with the cap across the resonance control, and it seems to make no difference in its value. I've yanked the it and I get the same response as if it's in place (this is with multiple cap values). If I short the resonance control, I have a presence control, if I turn up the resonance, I lose all of my presence control. Basically, messing with the resonance, messes with the effective value of the feedback resistor, so since I get adequate feedback with a feedback resistor in the 1-5k range, I'd need a resonance control that could stay within that range and still give me a usable control, yes?

[loudthud]

As you reduce the gain with feedback, there will be a point where the amp will oscillate. In fact you may not be able to get to the 10 to 1 ratio I talked about in previous posts. A partial remedy is a cap between the two phase inverter plates. Values of 47 to 220pf are typical. The high frequency response of the output transformer is the real issue and we don't have control of that.

A ratio of 10 is 20dB of feedback. That's on the high side of what is practical and may not be achievable in this case. A ratio of 3.2 would be 10dB and that should be practical. Now, the amount of feedback limits the amount of boost available with the presence and resonance controls. The feedback ratio is also approximently the ratio between the resonance pot (3.2) and the fixed feedback resistor (1). The resonance cap value is more dependent on the feedback resistor than the pot's value. Using the famous formula Freq = 1/(2*Pi*R*C) (resistor in ohms, cap if farads) try for a freq around 200Hz to make a first cut on the cap then go up or down depending on how it sounds. That should get the presence control to function across the range of the resonance control.

For screen resistors, I would lean towards the Fender value of 470 ohm on each tube.

B+ is not the same as plate voltage. B+ is meaured form the supply to ground where plate voltage is measured from plate to cathode. In the case of the phase inverter, there is significant voltage on the cathodes. As such, increasing the tail resistor (usually 10K to 22K) decreases the effective plate voltage. At some point you run out of headroom to drive the power tubes to full output. I was more concerned with the 300V B+ expecting 350V to 400V at that point.

[defaced]

Money! Both the presence and the resonance work. Having the formula for frequency is what did the trick, thank you. Right now it sounds pretty good with a resonance cap that give about 400Hz and a presence cap that gives about 2500Hz. I'm going to reverse engineer some other designs to see what they use.

Now the pot that is used for each control, what does that do to the sound. theoretically? I've played with it practically and the best I can come up with is that it's just different, but if there's a formula for it, then I'm all ears.

I can get ~350v (under load) on the B+ with little effort. The power supply uses two 22K resistors (IR6 and IR7) prior to the point where the 300v B+ is tapped. If I move the connection to between those two resistors, I get 350v B+. Having made that change, it sounds like I have more clarity overall. Will changing to 400V increase clarity?

[loudthud]

I play bass and the eq I like drastically cuts the midrange. I find the resonance control adds texture to the tone and loosens the bass but allows me to control it. Players that like tight bass would never like my tone. For me the presence control adds brightness but is not that much different from what a bright cap at the volume control does. For lead guitar I would expect some control on the raspyness of high notes in a solo or a little sparkle to big cords or rhythm parts. Think of presence and resonance as tone controls that add flexability to the usual tone controls.

The B+ to the phase inverter is something that has many subtle influences. Clarity is one but you can go too far and end up with a harsh overdrive tone. It's best to just experiment until you find the best overall compromise. Look at lots of other schematics to give you ideas and have fun.