Realistic Carnival 6V6GT Tube Amp Help

[misterc57]

My newly acquired little carnival tube amp sounds distorted and I need help trying to figure out why.

This is a 3 tube audio mono amp; 12AX7 6V6GT 6X5GT with original Toshiba tubes. 3 pots; Volume 1, Volume 2, on/off Tone.

I do not really notice the distortion when running a single tone signal, only when playing music full of different audio signals. Rock song intros do not sound too bad, but once the whole band kicks in, the distortion is there. I have tried this with different speakers and a different set of pre-amp and output tubes, still have distortion.

I have checked every resistor in the amp and they are all very close to spec.

I do not have the schematic and I think the voltages on the 12AX7 plates seem low.

All VDC readings to ground, no signal, all pots set to zero. All heaters were between 11.3 and 11.4 VDC. PT taps were 260 VAC. Below are my voltage readings

12AX7 pins

1: 94.7 V
2: -6 mV
3: .95 V
4: 11.39 V
5: 11.39 V
6: 79.4 V
7: .17 mV to 0 (keeps moving around)
8: -.23 mV
9: 11.4 V

6V6GT pins

1:
2: 11.3 V
3: 273 V
4: 245 V
5: 19 mV
6:
7: 11.3 V
8: 11.7 V

6X5GT pins

1:
2: 11.3 V
3:
4:
5:
6:
7: 11.3 V
8: 290 V


Attached is a picture of the chassis.



Thank you

[Enzo]

So how much ripple is on the main B+ at the first filter?

You did not report the cathode voltage on the 6V6, pin 8.

You did not report the grid voltage at pin 7 of the preamp tube, and the lack of voltage at the associated cathode on pin 8 is troubling.

[misterc57]

I have updated the missing readings in the original post.

Can I measure ripple (never done this before) using a DVM? If so how?

Thank you!

[Jazz P Bass]

To measure 'ripple' you set your DVM to read Volts AC.

Black lead on ground.

Red lead is the probe.

[misterc57]

There is a 20-10-10 capacitor can in the unit. There are color markings on the bottom (chassis side), 2 yellow, 1 red, 1 black. Here is what I measured for AC.

yellow: 69.1 mVAC
yellow: 64 mVAC
red: 2.8 VAC (this appears to be the input to the can from the rectifier pin 8)
black: .27 mVAC going into a 25 MFD cap, 78.5 mVAC leaving the 25 MFD cap to pin 8 of the 6V6GT.

[misterc57]

Any more thoughts on this? Are those ripple readings acceptable? Should I be checking the ripple at the plates of the 12ax7?

I wonder if the input signal level is too high at the input to the amp. My first test was with a CD player. Lots of distortion. I do not know what level the CD players puts out.

My recent tests are with an iPod player, if I make the volume level of the iPod very low (I think I was averaging around 300 mv of AC input signal), the distortion is barely noticed.

I still wonder about the plate voltages on the 12ax7, are they too low to generate a healthy signal?

Thank you!

[Jazz P Bass]

Those 'ripple' voltages look o/k to me.

What are the dc voltages on those same capacitors?

Did you ever get the plate voltages higher (pin 1 & 6) on the 12AX7?

As Enzo pointed out, the negative voltage on the cathode (pin 8) is not good.

The CD player at full volume into the amp is a 'line level' signal.
Probably 1 volt or more.

300mvs from the ipod is still pretty high.

I usually use a 100mv signal at test.

[misterc57]

I will do more testing with input signal levels.

This is a 1950s audio amp. It has two inputs, one for mic, one for phono/tuner. Anyone know what the expected signal levels would be for each input?

Thank you.

[DRH1958]

One thing I noticed is that you have DC voltage readings for the heaters. By the looks of the pic you posted, you should be getting 6.3VAC on the heater connections as they come directly off the PT. Measure those again, this time with the meter set to AC volts and measure directly on the pins of the socket just to be sure all tubes are getting proper heater voltage.

Also remember, this is a single 6V6 amp, so it's probably putting out only about 4 watts at those somewhat low voltages. It's not going to give you much clean sound before it starts to distort.

As Jazz P Bass says, 100mV is good as this is the average for a guitar/mic. Turn down the vol out of the ipod so it's at 100mV and hook up to the mic jack.

One last thing, I've worked on many older amps to convert to guitar use and many of these older amps used these low plate voltages.

[stokes]

I noticed the same thing about the heaters.It looks like the heater CT is connected to the cathode of the 6V6.That would explain the 11+ dcv's.The ripple volts are fine,but I question the 12AX7 readings,in particular V1/B or pins 6,7 and 8.Take plate pin measurements with no tube and see if they come up some.Then try a different 12AX7 if you have one.

[Jazz P Bass]

Here is the schematic.

Realistic Carnival.pdf

[misterc57]

All of this info and schematic has been great help!

I will get back to the amp tonight or tomorrow night and report back.

Thank you!

[DRH1958]

Great work Jazz for finding that obscure schemo. We can now see that the input stage(pins 6,7,8) is biased by the old grid leak method. I've worked on some old PA amps with this type of bias and usually they distort because of that method. You can tell it's this bias method because of the 3 meg value of R4 and the cathode tied to ground. R4 sets the bias. They did away with this in the '50s because it's an unstable method. Changing tubes usually results in different bias voltages. I think that 0v reading on pin 7 might be a typo. The cathode is at ground(0 volts), so the grid can't be at 0v too. There needs to be some bias voltage on the tube. Usually, this method puts the grid somewhere around -.5 to -.7 volts. Since this is off center bias for a 12AX7, it usually ends up distorting the input signal right off the bat. This could be much of your problem. If the input stage distorts, there's no way to clean it up. It just gets amplified further.

I also noticed the amp has the old 2-wire power cord from your photo. It would be a good idea to change it out to a new 3-wire cord and hook the green ground to the chassis. This will establish safety for you so the chassis can't end up being "hot" and establishes a good ground reference for voltage measurements. Once you have replaced the cord, see what the voltage is on pin 7. If it's too low, you could increase it by adding another megohm R in series with R4 to see if you can get closer to center bias. Changing tubes might change bias as well. You need to leave C4 in there to isolate this bias voltage from the input.

The other option might be to re-do the input network to a cathode bias setup and get rid of the grid leak bias method. But try the cord replacement and then possibly adding a 1 meg R in series with R4 first.

One last option would be to decrease R13 in the power supply to increase the preamp B+ to higher than 100v. But try one of these at a time starting with the power cord and see what happens before making wholesale changes.

[misterc57]

Very interesting. I do not understand why the 3 wire AC cord could change the readings? Because of a better ground?

I also noticed from the schematic that the input I was using (mic input) has two gain stages and the other input (phono-tuner) has one gain stage. By using the two gain path with a strong input signal I wonder if that tied into the distortion also.

I have a lot to try and do when I get back to the amp. I also wonder why there is no fuse in this amp (but I want to solve one issue at a time, as you said no wholesale changes).

Thanks again!

[DRH1958]

The main reason for the 3-wire power cord is for your safety. These old amps can have the chassis become energized because there is no grounded third wire. It's always recommended to either replace with a 3-wire cord(least expensive) or use an isolation tranny. Right now, your chassis has no way to be at the same potential as the ground in your 120v outlet because there is no third ground wire to accomplish this. Adding this 3-wire cord also stabilizes the chassis ground since it's solidly connected to your house ground once you plug it in. This might make a difference in your readings and might not, I don't know for sure. Your chassis is acting as a ground for the amp but is not properly grounded to the house ground. Also, you might have an energized chassis right now. I suppose I should ask, do you have a house that has the 3-wire outlets? If so, it's also good to know if the outlets you use are wired correctly. One of those little plug-in testers is cheap insurance to know this. I should also say that if you do replace the power cord, disconnect C9. You won't need it.

Many of these amps didn't come with fuses. This one was most likely built in the late 50's, so the requirements back then were quite different. You're right, it would be good to add a fuse too. Where the power cord enters the chassis, it should say what power the amp draws. Use that to determine your fuse size

Yes, you could use the phono-tuner input if you have a device with sufficient signal to drive the amp with only that one gain stage. I've never used this input because I've always rebuilt these old amps for guitar, so it never provided enough gain. I've always converted the mic input to cathode bias and used that input because with that old grid leak bias method, that stage almost always distorted. So yes, that could very well be why you are getting distortion. I'm not sure what signal level you will need to use the phono-tuner input. Experiment and see.