voltage on 6v6 GT

[rafaelctt]

No, there are only two valves (V3 and V4)

It’s not clear from the photo what is happening at the input. Where is the input?

There is no input in this partial circuit. Refer to the schematic in Reply #79. The problem is a 100KHz oscillation on the speaker. Pulling V3 will kill the oscillation. Put a ground on V3 pin 7 will kill the oscillation. Put a ground anywhere else in the V3A circuit will NOT kill the oscillation. What is the solution?

There must be some positive feedback going on somewhere, or the 1st stage is acting like a radio antenna. What about grounding Pin 2 of v3? (A normal guitar amp has a ground tip switch on the input Jack, or an instrument plugged in.) Or running a meter over the PS rail to check the filter caps are properly decoupled?

That is what I find strange, that putting pin2 of V3 to ground, there is a signal in the speaker. Even putting to ground the union of R19 with C12, there is still a signal.

[shooter]

signal implies you are providing it, is that the case, you're injecting a signal?

[tubeswell]

No, there are only two valves (V3 and V4)

It’s not clear from the photo what is happening at the input. Where is the input?

There is no input in this partial circuit. Refer to the schematic in Reply #79. The problem is a 100KHz oscillation on the speaker. Pulling V3 will kill the oscillation. Put a ground on V3 pin 7 will kill the oscillation. Put a ground anywhere else in the V3A circuit will NOT kill the oscillation. What is the solution?

There must be some positive feedback going on somewhere, or the 1st stage is acting like a radio antenna. What about grounding Pin 2 of v3? (A normal guitar amp has a ground tip switch on the input Jack, or an instrument plugged in.) Or running a meter over the PS rail to check the filter caps are properly decoupled?

That is what I find strange, that putting pin2 of V3 to ground, there is a signal in the speaker. Even putting to ground the union of R19 with C12, there is still a signal.

Have you measured the signal frequency? The universe is abuzz with electromagnetic radiation at all sorts of frequencies (some of it from outer space, some of it man-made) and an unshielded amplifier will pick up whatever is out there. (The more gain stages there are, the shorter the aerial needs to be to amplify it). If grounding the grid of V3b kills it, its coming from before the grid of V3b. Is there a radio transmitter/repeater station in your neighbourhood? You might be picking up 'raw' AM or FM carrier waves (that haven't been demodulated through a superhet) or something. 

[rafaelctt]

signal implies you are providing it, is that the case, you're injecting a signal?

no, I mean the 50Hz that appears in the speaker with levels between approximately 30 and 400 mV. Tomorrow I will try to inject a signal

[sluckey]

signal implies you are providing it, is that the case, you're injecting a signal?

no, I mean the 50Hz that appears in the speaker with levels between approximately 30 and 400 mV. Tomorrow I will try to inject a signal

That is just power supply hum.

[rafaelctt]

Do you think this simplification of the circuit could work?

[PRR]

"Work" like play? Probably. Work well? I bet with a complete AA Champ, no NFB or tone, volume, and then a THIRD triode gain stage, it will be a bee's-nest at any knob setting, and overload or scream past "3".

Have you tried a faithful copy of a Fender Champ? Especially the layout! When I tried that, I realized there are hard-learned unwritten lesions in how Leo placed those parts and wires. I was able to get it clean pushing wires around with a 'lead' pencil, but I had 90% stolen Leo's layout so I was closer than I knew.

[sluckey]

Do you think this simplification of the circuit could work?

NO!

[bmccowan]

I have watched this thread with interest. I applaud the patience shown.
Having swum in Lake Stubborn more than a few times myself, I think this is going to go one of two ways.

• Chasing rabbits down multiple holes until exhausted; disassembling the amp and rebuilding with a proven schematic and layout.
• or; proceed right to - disassembling the amp and rebuilding with a proven schematic and layout.

[rafaelctt]

I have watched this thread with interest. I applaud the patience shown.
Having swum in Lake Stubborn more than a few times myself, I think this is going to go one of two ways.

• Chasing rabbits down multiple holes until exhausted; disassembling the amp and rebuilding with a proven schematic and layout.
• or; proceed right to - disassembling the amp and rebuilding with a proven schematic and layout.

Not the idea. The circuit I'm trying to make is proven. The problem is that I'm trying to simplify parts to make sure it works and build it little by little to learn how it works. Thanks to this and to you I am learning how these amplifiers work. My idea is to learn what amplifiers I have to spare. Thanks for your help
NOTE: Sorry for the Google translation

[pdf64]

Do you think this simplification of the circuit could work?

Regarding V3b -
A DC path between a grid and its cathode is essential. It’s normally achieved by a grid leak resistor.

[rafaelctt]

Do you think this simplification of the circuit could work?

Regarding V3b -
A DC path between a grid and its cathode is essential. It’s normally achieved by a grid leak resistor.

Yes, I totally agree. It's a mistake. In fact I was going to do it with a 220k-470k resistor. Thanks

[rafaelctt]

Hello!
I have redone the circuit by shortening cables and following the layout of a Fender as precisely as possible.
It is made as in the circuit of the image.
The measurements I get now seem correct to me except that:
1- They seem a bit high in value to me
2-In pin 7 of V3B I have a voltage of -4.1 V
3- In pin 4 of V3 I still have higher voltage than in pin 3

Are they normal values?



tank you

[sluckey]

The measurements I get now seem correct to me except that:

1- They seem a bit high in value to me   (Quite normal for that Hammond 372BX)
2- In pin 7 of V3B I have a voltage of -4.1 V  (should be zero. Maybe it's really -4.1mV but you misread your meter?)
3- In pin 4 of V3 I still have higher voltage than in pin 3  (pin 4 is filament. What's your point?)

[rafaelctt]

The measurements I get now seem correct to me except that:

1- They seem a bit high in value to me   (Quite normal for that Hammond 372BX)
2- In pin 7 of V3B I have a voltage of -4.1 V  (should be zero. Maybe it's really -4.1mV but you misread your meter?)
3- In pin 4 of V3 I still have higher voltage than in pin 3  (pin 4 is filament. What's your point?)

I meant V4

in the 6v6 gt pin 4 is screen

[sluckey]

I meant V4

Sorry. I'm not a mind reader. 

It's very possible that your OT has been damaged by that BIG high frequency signal you had in the beginning.

[rafaelctt]

Is there a way to check it without replacing it?

[sluckey]

Is there a way to check it without replacing it?

Use your ears. How does it sound? If not sound good, replace the OT. No guarantees.

I have redone the circuit by shortening cables and following the layout of a Fender as precisely as possible.

Show us a hi-rez pic of the redone amp.

[HotBluePlates]

3- In pin 4 of V3 I still have higher voltage than in pin 3

Are they normal values?

I meant V4

Overall, having 6V6 screen voltage a bit higher than the plate voltage doesn't matter.  And it happens due to voltage drop across the output transformer primary.

Your diagram shows 406v in the power supply, and 386v at the 6V6 plate: 20v dropped across the transformer primary.
Hammond says the primary has a resistance of about 360Ω.  20v / 360Ω = about 55.6mA of plate current.

Your diagram shows 24v at the cathode, across a 460Ω resistor:  24v / 470Ω = 51mA
We can see that either the cathode resistor is lower than 470Ω, or your 1760C is higher than 360Ω.

6V6 Dissipation:  (386v - 24v) x 56mA = 20.3 watts ---> way too hot.

Reduce the 6V6 dissipation:
   Assume the plate voltage will land at 393v
   14w / 393v = 35.5mA
   Add 3mA of screen current:  35.5mA + 3mA = 38.5mA
   Calculate cathode voltage:  470Ω x 38.5mA = about 18v

   Using the triode curves in a 6V6 data sheet (page 5), I estimate the 6V6 needs a screen voltage of about 275v to pass 38.5mA with a bias of 18v (I can show the method some other time if you like).

   Your diagram shows a voltage drop from "B+ B" to "B+ C" that indicates preamp current is about 5mA.
   We'll estimate our 3mA 6V6 screen current plus 5mA preamp current flows through R20: 8mA
   We have 406v at "B+ A" and need 275v at "B+ B":  (406v - 275v) / 8mA = about 16kΩ ---> use an 18kΩ resistor for R20

[rafaelctt]

3- In pin 4 of V3 I still have higher voltage than in pin 3

Are they normal values?

I meant V4

Overall, having 6V6 screen voltage a bit higher than the plate voltage doesn't matter.  And it happens due to voltage drop across the output transformer primary.

Your diagram shows 406v in the power supply, and 386v at the 6V6 plate: 20v dropped across the transformer primary.
Hammond says the primary has a resistance of about 360Ω.  20v / 360Ω = about 55.6mA of plate current.

Your diagram shows 24v at the cathode, across a 460Ω resistor:  24v / 470Ω = 51mA
We can see that either the cathode resistor is lower than 470Ω, or your 1760C is higher than 360Ω.

6V6 Dissipation:  (386v - 24v) x 56mA = 20.3 watts ---> way too hot.

Reduce the 6V6 dissipation:
   Assume the plate voltage will land at 393v
   14w / 393v = 35.5mA
   Add 3mA of screen current:  35.5mA + 3mA = 38.5mA
   Calculate cathode voltage:  470Ω x 38.5mA = about 18v

   Using the triode curves in a 6V6 data sheet (page 5), I estimate the 6V6 needs a screen voltage of about 275v to pass 38.5mA with a bias of 18v (I can show the method some other time if you like).

   Your diagram shows a voltage drop from "B+ B" to "B+ C" that indicates preamp current is about 5mA.
   We'll estimate our 3mA 6V6 screen current plus 5mA preamp current flows through R20: 8mA
   We have 406v at "B+ A" and need 275v at "B+ B":  (406v - 275v) / 8mA = about 16kΩ ---> use an 18kΩ resistor for R20

Thank you very much for your explanation. Understanding that the power consumed would be 1, 152w, with which a resistance of 18k and 3w would be enough

[tubeswell]

3- In pin 4 of V3 I still have higher voltage than in pin 3

Are they normal values?

I meant V4

Overall, having 6V6 screen voltage a bit higher than the plate voltage doesn't matter.  And it happens due to voltage drop across the output transformer primary.

Your diagram shows 406v in the power supply, and 386v at the 6V6 plate: 20v dropped across the transformer primary.
Hammond says the primary has a resistance of about 360Ω.  20v / 360Ω = about 55.6mA of plate current.

Your diagram shows 24v at the cathode, across a 460Ω resistor:  24v / 470Ω = 51mA
We can see that either the cathode resistor is lower than 470Ω, or your 1760C is higher than 360Ω.

6V6 Dissipation:  (386v - 24v) x 56mA = 20.3 watts ---> way too hot.

Reduce the 6V6 dissipation:
   Assume the plate voltage will land at 393v
   14w / 393v = 35.5mA
   Add 3mA of screen current:  35.5mA + 3mA = 38.5mA
   Calculate cathode voltage:  470Ω x 38.5mA = about 18v

   Using the triode curves in a 6V6 data sheet (page 5), I estimate the 6V6 needs a screen voltage of about 275v to pass 38.5mA with a bias of 18v (I can show the method some other time if you like).

   Your diagram shows a voltage drop from "B+ B" to "B+ C" that indicates preamp current is about 5mA.
   We'll estimate our 3mA 6V6 screen current plus 5mA preamp current flows through R20: 8mA
   We have 406v at "B+ A" and need 275v at "B+ B":  (406v - 275v) / 8mA = about 16kΩ ---> use an 18kΩ resistor for R20

Thank you very much for your explanation. Understanding that the power consumed would be 1, 152w, with which a resistance of 18k and 3w would be enough

Are you using language translation software?

[tubeswell]

Hello!
I have redone the circuit by shortening cables and following the layout of a Fender as precisely as possible.
It is made as in the circuit of the image.
The measurements I get now seem correct to me except that:
1- They seem a bit high in value to me
2-In pin 7 of V3B I have a voltage of -4.1 V
3- In pin 4 of V3 I still have higher voltage than in pin 3

Are they normal values?



tank you

Based on this schematic and these reported voltages, the 6V6 is dissipating 16W

(I get 51mA cathode current, i.e., 24V/470R, including 6.9mA* screen current)


So, 44.2mA plate current x 362V plate-to-cathode = 16W plate dissipation


*based on the voltage drop across the supply resistors. i.e. 12V across 1k (R20) minus 92V across 18k (R21). This is way too much idle screen current for a little ole 6V6. In order to lower the screen current, you need to lower the screen voltage by about 50V. So try 10k (instead of 1k) for R20. This should bring the 6V6 back into more normal operating conditions. (Note also that when you increase R20 to 10k, you may want to decrease R21 to something between 4k7 to 10k if you want to maintain the preamp plate voltages around 250V)

[rafaelctt]

3- In pin 4 of V3 I still have higher voltage than in pin 3

Are they normal values?

I meant V4

Overall, having 6V6 screen voltage a bit higher than the plate voltage doesn't matter.  And it happens due to voltage drop across the output transformer primary.

Your diagram shows 406v in the power supply, and 386v at the 6V6 plate: 20v dropped across the transformer primary.
Hammond says the primary has a resistance of about 360Ω.  20v / 360Ω = about 55.6mA of plate current.

Your diagram shows 24v at the cathode, across a 460Ω resistor:  24v / 470Ω = 51mA
We can see that either the cathode resistor is lower than 470Ω, or your 1760C is higher than 360Ω.

6V6 Dissipation:  (386v - 24v) x 56mA = 20.3 watts ---> way too hot.

Reduce the 6V6 dissipation:
   Assume the plate voltage will land at 393v
   14w / 393v = 35.5mA
   Add 3mA of screen current:  35.5mA + 3mA = 38.5mA
   Calculate cathode voltage:  470Ω x 38.5mA = about 18v

   Using the triode curves in a 6V6 data sheet (page 5), I estimate the 6V6 needs a screen voltage of about 275v to pass 38.5mA with a bias of 18v (I can show the method some other time if you like).

   Your diagram shows a voltage drop from "B+ B" to "B+ C" that indicates preamp current is about 5mA.
   We'll estimate our 3mA 6V6 screen current plus 5mA preamp current flows through R20: 8mA
   We have 406v at "B+ A" and need 275v at "B+ B":  (406v - 275v) / 8mA = about 16kΩ ---> use an 18kΩ resistor for R20

Thank you very much for your explanation. Understanding that the power consumed would be 1, 152w, with which a resistance of 18k and 3w would be enough

Are you using language translation software?

Yes, I'm sorry, I'm writing from Spain

[HotBluePlates]

Based on this schematic and these reported voltages, the 6V6 is dissipating 16W

I agree.

But didn't I mention I don't exactly trust the numbers provided?  Because I don't exactly trust the numbers provided.   

[rafaelctt]

I have made a series of measurements of voltages and currents. I have done it with resistance R20=1K and with R20=10K. In both images you can see the measurements obtained.

with R20=1k





with R20=10k

[rafaelctt]

I think that with the 10k resistor the measurements are more or less good. The problem is that I have a lot of distortion and too much gain with the volume at minimum. The reverb also gets really loud with the volume all the way down. How would I measure the gain of each triode? I have a signal generator but it barely injects a level above 100mV when I plug it in. (no load if it goes up to a few volts). It must be the low impedance it has (600 Ohm)

[sluckey]

I think that with the 10k resistor the measurements are more or less good.

I agree

How would I measure the gain of each triode?

Inject a 1KHz 100mV zero volt offset signal at the input jack (If you don't know what offset is then use a .1µF cap between the FG and the amp). Set scope to AC coupling on input. Use scope to measure signal amplitudes at V1A grid and plate. Divide plate signal voltage by grid signal voltage. This resulting number is the gain for V1A. Repeat for other triodes. For example, if you measure 100mV_pp at the grid and 2V_pp at plate, the gain would be 20.

I have a signal generator but it barely injects a level above 100mV when I plug it in. (no load if it goes up to a few volts). It must be the low impedance it has (600 Ohm)

The input impedance of your amp should be 1MΩ. That's almost like "no load" to your FG. There's either a problem with the input wiring of your amp ***OR*** a problem with your FG.

To confirm your FG is good... Set the FG for 1KHz sine with zero offset. Connect a 1MΩ resistor across the 600Ω output of your FG. Now measure the signal voltage at the output of the FG. If you can still only get 100mV signal on the scope then the FG is bad. If you can get several volts on the scope then you have a wiring error between the amp input jack and V1A grid.

You can confirm the input is wired correctly by measuring the resistance between V1A grid and chassis. You should measure 33K with nothing plugged into the input jack. Now connect a guitar cable to the input but DO NOT connect the other end of the cable to anything. The resistance between V1A grid and chassis should now read 1MΩ.

[PRR]

...I have a signal generator but it barely injects a level above 100mV when I plug it in. (no load if it goes up to a few volts). It must be the low impedance it has (600 Ohm)...

FWIW: Looks like a 3311 series.
"Ten V p-p into 600 ohms (20 V p-p O.C.). This output may be attenuated by >30 dB by a variable attenuator and offset by ± 5 V."
https://calright.com/product/agilent-hp-3311a-function-generator-1-mhz/
https://engineering.case.edu/lab/circuitslab/sites/engineering.case.edu.lab.circuitslab/files/docs/Hewlett_Packard_3311A_Function_Generator_Operating_and_Service_Manual.pdf

[rafaelctt]

These are some captures with the oscilloscope with high volume.
The current through R20 is increased to 16 mA. Therefore the resistor has to dissipate almost 3w.

[sluckey]

The current through R20 is increased to 16 mA. Therefore the resistor has to dissipate almost 3w.

I suggest replacing R20 with a 10W resistor. I'd probably use a 5KΩ 10W resistor.

Have you plugged a guitar into the amp? How does it sound?

What did you do to resolve the low output from your FG?

[rafaelctt]

In order to work optimally with the FG, I put a 0.1 capacitor as you indicated. From then on I was able to do the measurements without problems. He was injecting me with a voltage of 7 volts DC.
Tomorrow I'll try the guitar but I think I'll have distortion.
Something strange that happens is that a sound is generated in the output transformer with the same frequency as the input. I am doing the measurements without a speaker and with a load of 8 Ohms 15W. Even so, the frequency of the input of the FG is heard in the OT.

[sluckey]

Something strange that happens is that a sound is generated in the output transformer with the same frequency as the input. I am doing the measurements without a speaker and with a load of 8 Ohms 15W. Even so, the frequency of the input of the FG is heard in the OT.

That's very normal. The OT laminations are vibrating. Doesn't mean anything is wrong. You need to turn the FG output down to 100mV_pp.

He was injecting me with a voltage of 7 volts DC.

The DC offset knob controls that dc voltage. Read the manual to learn how to use it.

[rafaelctt]

As expected, the sound with a minimum volume already sounds distorted and the reverb sounds too much with the P5 potentiometer turned down. How can I lower the gain of the reverb?
Thank you

[ghostinthescope]

I definitely don't know as much as the rest of the folks on this forum and have learned A LOT lurking here and reading.

In my humble opinion I suggest you connect a 50k resistor from pin 7 of V3 to ground.

I built an amp like this a while back and although I can't find my notes I remember looking at schematics for other Fender Reverb amps that have higher gain phase inverters, like the long tail pair phase inverters, had some gain reduction through the 50k tremolo pot. The princeton reverb has a cathodyne phase inverter which has a gain of less than 1, so it doesn't have the 50k resistor. I think your last gain stage has a lot more gain than a cathodyne phase inverter. I believe/hope this will also help with the reverb.

I'm sure other folks here can better speak to all of these points or help to correct what I have said here.

[ghostinthescope]

After second thought I think most of what I said is flawed, but I still maintain that some sort of grid leak resistor at v3 pin 7 will help you tame the gain.

The 50k in the other fender circuits comes after that gain stage, but before the phase inverter.

[sluckey]

R19 and the Reverb pot are the grid leak for V3 pin 7. If you put a 50K between V3-7 and ground you will create a voltage divider (66:1) with that 3.3M. That will almost destroy the signal arriving at V3-7.

[ghostinthescope]

Thanks for the correction, Sluckey. That makes a lot of sense.

What's up with the 220K resistor in the other AB763 circuits? Would that be that more appropriate?

[rafaelctt]

I inject a 90 mV rms signal into pin 2 of V1. With the volume at maximum (P1) I have 30 volts rms on pin 5 of V4, isn't it too much? The wave is very distorted on its positive side.

[rafaelctt]

These are the rms measurements injecting 97mV 1Khz

[sluckey]

What's up with the 220K resistor in the other AB763 circuits? Would that be that more appropriate?

Only Leo knows.   

I can speculate... Maybe it's to provide a dedicated ground reference for the tube grid. Maybe the signal was too hot so the 220K was added to drop the level a bit.

[sluckey]

These are the rms measurements injecting 97mV 1Khz

I don't see any alarms.

You have analyzed this amp to death with your FG. IT'S TIME TO PLUG IN A GUITAR!!!

[rafaelctt]

These are the rms measurements injecting 97mV 1Khz

I don't see any alarms.

You have analyzed this amp to death with your FG. IT'S TIME TO PLUG IN A GUITAR!!!

I connected the guitar and that's why I commented that it distorted a lot and that the reverb immediately had too much level
Reply #132

[ghostinthescope]

     Hello rafaelctt, did you make any more progress on your project? I have been really appreciating your perseverance and I think you're probably very close. I appreciate long threads with people who are dedicated to finding solutions. I know I've burned a lot of candles chasing down silly mistakes or even sometimes faulty components.

did you double and triple check that the reverb pot ground lug is definitely connected to ground?

what kind of settings did you have your eq on when you plugged your guitar in? were they all the way up?

are you using humbuckers or single coil pickups?

can you share any high resolution photos of your latest work?

[HotBluePlates]

I inject a 90 mV rms signal into pin 2 of V1. With the volume at maximum (P1) I have 30 volts rms on pin 5 of V4, isn't it too much? The wave is very distorted on its positive side.

These are the rms measurements injecting 97mV 1Khz

Your diagram shows 33.8v RMS on the 6V6 grid.

The diagram in this post shows your cathode (bias) voltage is 19.8 volts.  That means the 6V6 cannot accept more than about 18.5v peak before it draws grid current & distorts heavily.

"The wave is very distorted on its positive side" because 6V6 grid current is lopping off the top of the wave.  The 6V6 is receiving 33.8v RMS x 1.414 = 47.8v peak, or about 2.6x more than it can handle cleanly.  PRR told you this would happen back in this post.



You might be able to knock down some of the signal level by making R19 smaller (reduces the Dry signal strength, but makes the Reverb stronger), splitting R14 into 2 resistors to reduce the drive into the Reverb Driver 12AT7, and use a 100kΩ Audio taper for the Reverb pot so the reverb is not so strong at low settings.

[sluckey]

I wonder what it sounds like if you just turn the volume control down?

[ghostinthescope]

It looks like he had the volume control down when he tried it with the guitar... from reply #132: "As, expected, the sound with a minimum volume already sounds distorted and the reverb sounds too much..."

Maybe the circuit is  "proven" as mentioned earlier by rafaelctt in this thread, but this isn't a commercial circuit... or is it??

maybe the extra gain stage is just overdrive right out of the gate?

maybe its worth incorporating some gain reduction techniques?

an amp I built a few years back for my GF is also single ended with reverb. almost exactly the same as this except I used the fender tweed style tone control. The tweed tone control doesn't bog down the signal much and... holy moly it had way too much gain at the start. I definitely used some resistance to bring the signal down to a reasonable level and may have removed a bypass cap somewhere too. I'd have to pull the chassis, because I can't quite remember and have no idea where that notebook is.

[PRR]

> The 6V6 is receiving 33.8v RMS x 1.414 = 47.8v peak, or about 2.6x more than it can handle cleanly.

It would be reasonable, for most genres and styles, for the amp to be on the edge of overload when dimed and 20mV input. Since he has 97mV input when he has gross distortion, he's not far off; unless this is already turned-down some.

[bmccowan]

Maybe the circuit is  "proven" as mentioned earlier by rafaelctt in this thread, but this isn't a commercial circuit... or is it??

I and others questioned the circuit several posts back. It seems that the basic circuit is proven, but there are multiple added gain stages. In addition, the subject of the non-standard practice layout is still out there. Requests for pics of the revised layout have not been answered making this a frustrating thread.
Were I to work on this amp, I would temporarily disconnect the reverb and the extra gain stages and get the amp working producing clean sound. Then I would start adding things back in, testing and adjusting along the way.