Princeton reverb problem

[SleepLess]

Hi everyone.
I finished building a PR replica last week and have played it since then. It sounds great from 1 to 7 on the volume pot. The problem is that starting from 7 it sounds harsh and metallic. On 10 the sound seems to cut at times and it has ugly, harsh distortion.

I need you guys to give me leads on where to look to find a solution. Can it be a bad component that behaves ok as long as moderate volume is applied? What else could it be?
If you need voltage readings, don't hesitate and say so.

Thanks in advance!

[shooter]

seems to cut at times and it has ugly, harsh distortion.

do you have a way to measure the AC signal driving you PA tubes?  Almost sounds like you're driving them to cut-off, or a really distorted drive signal, fwiw

[SleepLess]

seems to cut at times and it has ugly, harsh distortion.

do you have a way to measure the AC signal driving you PA tubes?  Almost sounds like you're driving them to cut-off, or a really distorted drive signal, fwiw

I have a MM yes. Tell me how to do that and I'll do it!
Thanks for your help!

[eleventeen]

(drive-by idea) Measure (DC) plate and cathode voltages on your preamp tubes. Should be ~~170-180 volts on the plates, about 1.5-2 volts 
on the cathodes.


The reverb driver tube has HIGH HIGH volts on its plate. Be careful. (meaning, it is different from the preamp tubes)
The 12AT7 phase splitter has rather different volts than the preamp tubes. (meaning, it is different from the preamp tubes)


Yet, per: http://el34world.com/Forum/index.php?topic=18468.0   PRR claims preamp plate volts do not much affect gains.


I defer to his deeper knowledge but I would draw up a little spreadsheet and just measure those 13 or so points real quick like, just to make sure those tubes are operating "normally". You actually need to make those measurements twice, once with volume at zero and another at full blast. I would do that before I started looking around for a harder-to-find cap or R that has either drifted off value (very strange in a brand new build but not unheard of)


If you make the SS in advance, it take very little time to make & record those measurements.


It is also possible that you have a HF oscillation (that you can't hear) but which is driving the amp nuts at the extreme rotation of the volume control.

[SleepLess]

Hi! V7 is a GZ34S by JJ. All other tubes are brand new JJ right out of the box as well. B+ is 396V. V5 is set at 21mA, V6 is at 24mA. Spreadsheet

Tube     Pin 1        Pin 2       Pin 3       Pin 4          Pin 5         Pin 6        Pin 7        Pin 8        Pin 9
V1         166            0           1.3         3.1AC         3.1AC        166           0             1.3          3.1AC
V2         380             0           7           3.1AC         3.1AC        380           0              7            3.1AC
V3         166            0           1.3         3.1AC         3.1AC        168          0(noisy)    1.3         3.1AC
V4       196/263       0         2.1/2.6    3.1AC         3.1AC         238        23.9           69          3.1AC
V5         0            3.1AC       392         386            -33.5           0          3.1AC          0
V6          0           3.1AC       392         386            -33.5           0          3.1AC          0
V7          0           396             0          313AC       0.9AC         315AC    0.8AC        396

Thanks for your help!

[Fresh_Start]

Do you have a listening amp?  Or second choice an oscilloscope?  Either would be very helpful in finding out where the nasty distortion is introduced in the signal path. There's a distinct possibility that the ugly distortion is showing up at the phase inverter. If so, there's a mod you could try to imcrease the headroom of the cathodyne.

How about some alternative tubes?  I'm not partial to JJ 6V6s because they can sound harsh.  Tung Sol  6V6s for new production would be my first alternative to try but tastes vary.

IME there seems to be a lot of variance in how 12AT7s perform. Does the nasty sound appear when you have reverb turned up?  How about when it's off?

A vintage 12AX7 in the V1 spot might sweeten up the sound.

Last but far from least, what is your power tube dissipation?  Getting the bias dialed in right is important, especially with bias vary tremolo.

Hope that helps,

Chip

[eleventeen]

Nicely done!

V4       196/263       0         2.1/2.6   <<----I take it this means that these are two measurements [plate / cathode]  taken with volume = "low" and "high", "high" being where you have a problem. I can't see those numbers being right.


I'm stumped as to how the setting on your volume control (which sits in between halves of V1 - first preamp) could affect voltages on V4 . But I smell HF oscillation.

If this was built from a kit, probably you wired the output transformer blue to one plate and brown to the other and those colors were diagrammed out for you and you wired them properly.


But at the risk of sounding silly, it is an absolute requirement that the OT is properly phased. No matter what any diagram says. It *is* possible that a given manufacturer could supply one flavor of OT one month and find a better deal or run out of one type and substitute another type the next month. And oops, still send the same diagram. So, I believe it is a small possibility that your OT is not properly phased---which means you have to flip and test whether (either, not both) the primary or secondary are flopped, phase wise.


Most amps will motorboat or oscillate at *any* volume level with the OP phase reversed. Some don't. So...if you are following what I'm saying, you have to test this on at least an experimental basis. Because NOTHING ELSE will fix it if this is the problem, you sort of have to prove it not to be the problem, at least in my peanut brain. I'd assume you've already cut the blue & brown plate wires to length so those are not easy to flip---though you could, by tacking in some short wires for test purposes. (High voltages there, right?) But you could try flipping the green and black on the output leads. If you flip the wires and you get immediate oscillation or the problem is much worse, then it was right the way it was. False alarm, flip them back. Just make the flip on a temp basis---tack solder the wires. EITHER pri or secondary: Not both. (The preference is to flip the blue & brown)

Yeah, it sounds stupid, but if it's wrong, it will never ever work right misphased. 

[tubeswell]

Hi everyone.
I finished building a PR replica last week and have played it since then. It sounds great from 1 to 7 on the volume pot. The problem is that starting from 7 it sounds harsh and metallic. On 10 the sound seems to cut at times and it has ugly, harsh distortion.

I need you guys to give me leads on where to look to find a solution. Can it be a bad component that behaves ok as long as moderate volume is applied? What else could it be?
If you need voltage readings, don't hesitate and say so.

Thanks in advance!

Going by the symptoms you describe (occurs above 7) sounds like you could have gain spike or frequency-doubling from overdriving the cathodyne stage. Solution is to put a 470k to 1M grid stopper right at the grid pin of the cathodyne inverter stage.

[SleepLess]

Nicely done!

V4       196/263       0         2.1/2.6   <<----I take it this means that these are two measurements [plate / cathode]  taken with volume = "low" and "high", "high" being where you have a problem. I can't see those numbers being right.


I'm stumped as to how the setting on your volume control (which sits in between halves of V1 - first preamp) could affect voltages on V4 . But I smell HF oscillation.

No there's only one measurement. Being the tremolo tube, voltage varies. All my measurements were done with intensity, speed, reverb on 1 (which is OFF), bass and treble on 5 and volume on 1 (OFF).

If this was built from a kit, probably you wired the output transformer blue to one plate and brown to the other and those colors were diagrammed out for you and you wired them properly.

I tried that already. Didn't change a thing.

But at the risk of sounding silly, it is an absolute requirement that the OT is properly phased. No matter what any diagram says. It *is* possible that a given manufacturer could supply one flavor of OT one month and find a better deal or run out of one type and substitute another type the next month. And oops, still send the same diagram. So, I believe it is a small possibility that your OT is not properly phased---which means you have to flip and test whether (either, not both) the primary or secondary are flopped, phase wise.


Most amps will motorboat or oscillate at *any* volume level with the OP phase reversed. Some don't. So...if you are following what I'm saying, you have to test this on at least an experimental basis. Because NOTHING ELSE will fix it if this is the problem, you sort of have to prove it not to be the problem, at least in my peanut brain. I'd assume you've already cut the blue & brown plate wires to length so those are not easy to flip---though you could, by tacking in some short wires for test purposes. (High voltages there, right?) But you could try flipping the green and black on the output leads. If you flip the wires and you get immediate oscillation or the problem is much worse, then it was right the way it was. False alarm, flip them back. Just make the flip on a temp basis---tack solder the wires. EITHER pri or secondary: Not both. (The preference is to flip the blue & brown)

Yeah, it sounds stupid, but if it's wrong, it will never ever work right misphased.

[SleepLess]

Do you have a listening amp?  Or second choice an oscilloscope?  Either would be very helpful in finding out where the nasty distortion is introduced in the signal path. There's a distinct possibility that the ugly distortion is showing up at the phase inverter. If so, there's a mod you could try to imcrease the headroom of the cathodyne.

I have other amps at home yes. A Tremolux, a Trainwreck Express.

How about some alternative tubes?  I'm not partial to JJ 6V6s because they can sound harsh.  Tung Sol  6V6s for new production would be my first alternative to try but tastes vary.

IME there seems to be a lot of variance in how 12AT7s perform. Does the nasty sound appear when you have reverb turned up?  How about when it's off?

I will try other tubes. Which ones do you suggest first, the 12AT7 and V4?

A vintage 12AX7 in the V1 spot might sweeten up the sound.

Last but far from least, what is your power tube dissipation?  Getting the bias dialed in right is important, especially with bias vary tremolo.

Hope that helps,

Chip

[Willabe]

Link for Doug's listing amp, scroll down to find it;

http://el34world.com/Hoffman/tools.htm


            Brad   

[SleepLess]

I think I've solved it. I had a wire going to the cap for 40uF instead of 20uF and vice versa. I reversed those. The amp sounds better. I can play it on 10 now without reverb, with reverb, the effect drowns the signal and it doesn't sound good. I suppose that is normal. Thanks for your help guys.

I still think that my voltages for pins 6 and 8 of V4 are a bit high. I have changed tubes to no avail. All my readings all read a bit below the Fender specs but that V4 reads higher...
Thanks!

[SleepLess]

Crap... I haven't solved it at all actually. Same booms and bad sound when pushed above 8... Damn, I've neevr experienced that thing. What can it be?

[SleepLess]

Guys I've just shot a video of the amp. The beginning shows it on 6 1/2 and it sounds great, and then I push the volume up to 8 and it's chaos... I swapped V4 with two other tubes to no avail, V1 as well.

Let me know what you think! Thanks!

http://youtu.be/pRIIwlxsvc4

[tubeswell]

Crap... I haven't solved it at all actually. Same booms and bad sound when pushed above 8... Damn, I've neevr experienced that thing. What can it be?

See my earlier suggestion

[dude]

Grid stopper control

[SleepLess]

Thanks a lot for your suggestions guys. It is the fourth Princeton Reverb I'm building, its' for a friend of mine and it's the first time I've had this issue. I could be adding your "mods" to control that issue but I'd rather try and find the cause of it at the source. Something must be wrong since I never installed any of those mods before and they were perfectly well.

Not to neglect your approaches, but I'd rather find why I would actually need these mods to find the root of the problem and then solve it.
Thanks!

[Willabe]

It is the fourth Princeton Reverb I'm building, its' for a friend of mine and it's the first time I've had this issue.

Ok, did you change anything in this build? (Chassis, layout/grounding scheme, iron set, cabinet if combo, ie, how close the speaker magnet is to the tubes?)


 
                  Brad   

[Willabe]

If this was built from a kit, probably you wired the output transformer blue to one plate and brown to the other and those colors were diagrammed out for you and you wired them properly.

I tried that already. Didn't change a thing.

Just to be clear, you did switch the 2 OT primay plate wires with each other?


                  Brad     

[tubeswell]

... I'd rather find why I would actually need these mods to find the root of the problem and then solve it.
Thanks!

For a useful discussion of gain spike and frequency doubling distortion in cathodyne PIs, see either one of Merlin Blencowe's preamp books (pp169-172 of the 1st edition, or pp152-153 of the 2nd edition)

[dude]

From that sound clip seems like a wiring mistake, especially if it's a proven circuit.

I'm with Willabe, check all the wiring maybe use a magic marker to put a dot next to all connections re-checked. Perhaps a wire you forgot to solder, it's usually some simple mistake but can be very time consuming to find.

I think this is your first step before any changes to the circuit.

al

[Fresh_Start]

Try alternate 12AX7 in V4 spot.
Try grid stopper on cathodyne.
Try different reverb tank - I have one Accutronics that sounds like crap in any amp. No explanation.

Most of all try listening amp to locate problem.

Respectfully,
Chip

[HotBluePlates]

Thanks a lot for your suggestions guys. It is the fourth Princeton Reverb I'm building, its' for a friend of mine and it's the first time I've had this issue. I could be adding your "mods" to control that issue but I'd rather try and find the cause of it at the source.

I understand your reasoning and agree with the sentiment. But note also Tubeswell's suggestion is a quick, 1-resistor change. I'd consider it a diagnostic test in your case, to see if it even solves the problem (which then confirms/denies overdriving the inverter as the problem).

If it does fix the problem, then you have a choice: leave the resistor in, or do a side-by-side comparison of tubes, parts, voltage conditions and drive signals between this amp and one of your previous Princeton's to see what's different to make this one misbehave.

For all troubleshooting, the most valuable tests to perform are ones which exclude potential problem-sources as they reduce the list of could-be's. If Tubeswell's grid-stopper suggestion does not solve the issue, you have at least ruled out phase inverter overdriving and grid current as a "could be."

[SleepLess]

Got it HotBlue! Will do it! Thanks!

[dude]

What's the difference between a grid stopper and a grid leak resistor?

[Willabe]

A grid leak goes from grid to ground.

A grid stop is in series, signal comes in 1 end and goes out to the grid.



                          Brad   

[dude]

I knew that, but some amps use both some only the grid leak.

Can you use a grid stopper without a grid leak.

[sluckey]

Can you use a grid stopper without a grid leak.

No. There must be a dc resistance path from the grid back to the cathode for proper biasing.

[dude]

Thanks, I assume the grid stopper goes directly to the grid and the grid leak just before to ground.

I see usually the grid leak is 1M

Stopper anywhere from 1K to 470K

In a Fender circuit I would think a lower valve (1K5 to 4K2) as not to take any highs away and lose touch sensitivity.

al

[Willabe]

I assume the grid stopper goes directly to the grid and the grid leak just before to ground.

Yes.

I see usually the grid leak is 1M

On 12 _ _ 7 preamp tubes but not on power tubes.



                        Brad   

[tubeswell]

The size of the grid stopper that can be used on a cathodyne stage's grid without affecting any audible HF is quite large (470k-1M) because the input capacitance of a cathodyne stage is very low (even if a 12AX7 is used).

[Fresh_Start]

Note that the effective grid stopper on a typical Fender hi/lo input jack circuit is 34K for the hi I out and 68K for the lo input.

Randall Aiken has a nice piece on grid stoppers in his tech notes.

Cheers,

Chip