2 amps, same circuit, same parts, one of them has muted top end?

[HBamps]

There is one circuit that I seem to build alot of for friends, a cathode biased JTM 45 type amp that is really a blast to play. The first few were really great then something started happening where the new builds have this muted top end, for lack of a better term.

The older ones breath and have a very 'open' quality without being bright.
the newer ones sound more constrained and less lively.

I realize part tolerances and understand not all amps sound / feel the same but when it's more than one, I start to wonder.

The circuits are the same, same values, mostly the same brand of components. The only differences I can think of are the following;

- Old ones use Hammond 270EX PT, new Weber W025130
- Old ones have a cap can for 1st 2 filters, new are individual.
- Old yellow Malloy, new white
- Old electrolytic bypass cap for V1, new film

I suspect it might be the PT but if all voltages are reading about the same, why would it make such a difference? The Weber PT is smaller, maybe less stout power and different response?

Any help would be appreciated. I am stumped.
Thanks,

[Enzo]

So take one of the new ones, and replace the Weber with a Hammond. Then you will know if that is the difference or not, and purchase future build parts accordingly.

Then, are all the tubes exactly the same? Got one of the old "good" amps around? Swap its tubes into a new dull amp. ANy change? It isn;t enough to say Sovtek 12AX7, the Sovtek 12AX7WA and 12AX7WC don;t act the same. And differnt brands have their characters as well.

Look at your parts bins carefully. A restock could have included some wrong values. Very easy to glance at a color code and not notice one wrong band or misthink a three digit numerical on a small cap. Not long ago I ordered from a trusted supplier a bag of 1 ohm 5 watt resistors. They sent me .1 ohm. I noticed that just as I was about to dump them in my 1 ohm drawer.

[R.G.]

... the new builds have this muted top end, for lack of a better term.

The older ones breath and have a very 'open' quality without being bright.
the newer ones sound more constrained and less lively.

"Muted top end" and "open quality" are usually indicators of absence or presence of changes in the 2kHz-4KHz range, usually thought of as "presence" in audio knob-bery.

I suspect the difference is less in the power supply and caps than in something like wire routing and capacitance.

It's *exactly* this issue that makes PCB construction valuable. The wires are in *exactly* the same relationship each time, so the parasitics are consistent, whatever they are. As you're discovering, consistency between builds is a valuable commodity in its own right.

There are other possibilities, some of which your'e already on, like capacitor tolerances, which are much worse than resistor tolerances in general, even for film types.

[greekie]

Great advice from the two tech senseis so far.

My advice is to the stiffness of things - chances are that your new filter caps might be more efficient than the old types. This can lead to a sort of stiffness, as if the capacitance was higher in the new one than in the old ones (which, as R.G. noted, might also be the case!). Try to go a step lower on the capacitor values. Of course, if it induces hum, it's a no go. But I find it valuable to mess around with that kinda stuff sometimes!

YM2C

Jake

[HBamps]

So take one of the new ones, and replace the Weber with a Hammond. Then you will know if that is the difference or not, and purchase future build parts accordingly.

Then, are all the tubes exactly the same? Got one of the old "good" amps around? Swap its tubes into a new dull amp. ANy change? It isn;t enough to say Sovtek 12AX7, the Sovtek 12AX7WA and 12AX7WC don;t act the same. And differnt brands have their characters as well.

Look at your parts bins carefully. A restock could have included some wrong values. Very easy to glance at a color code and not notice one wrong band or misthink a three digit numerical on a small cap. Not long ago I ordered from a trusted supplier a bag of 1 ohm 5 watt resistors. They sent me .1 ohm. I noticed that just as I was about to dump them in my 1 ohm drawer.

I tryed the tube swap, there's a little variation but it wasn't the real issue. I am going to buy another Hammond PT and try it, I resist because it's alot of work if it doesn't give me the results I'm after. Such as it goes I guess.

Two more observation regarding wire routing.
1) Old has power tube gris resistors on the board, new on the socket.
2) Old is a head, new is a combo where tubes are mounted at 90 degrees and slightly smaller chassis.

Thanks for the responses!

[loudthud]

How much test equipment is used to look at these amps after they are built? Do you have an oscilloscope? A Variac that measures line current? Have you plotted frequency response?

[HBamps]

How much test equipment is used to look at these amps after they are built? Do you have an oscilloscope? A Variac that measures line current? Have you plotted frequency response?

Little to none. I am a hobbyist. A multi-meter is about where it ends.
I have a O-scope but haven't learned to use it yet.....

[GregS]

I think it would be well worth your time and effort to learn how to use the oscilloscope. You can probe various points in your circuit to see where a high frequency roll off might be occurring. If you find one in one of the newer versions of the amp then test one of the older versions to see if there's a difference. This way you can possibly track down where the problem is occurring, eliminate guess work, and make only the change(s) necessary to fix the problem.

[loudthud]

Little to none. I am a hobbyist. A multi-meter is about where it ends.
I have a O-scope but haven't learned to use it yet.....

OK, any chance you crossed the wires from the Phase inverter to the power tubes in the new layout?

Set the meter to measure AC volts and connect it across the speaker. You should have a very low reading in the milli-volts. If there is an ultrasonic oscillation, you won't hear it but you should measure several volts at the speaker. The feedback resistor goes from the speaker jack back to the Presence control. Try the amp with that resistor disconnected. The gain of the amp will increase but see if the high end comes back.

If the above experiments don't reveal the problem I would set the two amps next to each other and cross connect them at the output from the Treble control. To clarify, disconnect the power amp at the Treble control of the good amp and connect it to the Treble control of the bad amp. Do the same with the bad amp. Did the problem stay with the bad amp, move to the good amp, or slightly affect both amps?

[HBamps]

OK, any chance you crossed the wires from the Phase inverter to the power tubes in the new layout?

Set the meter to measure AC volts and connect it across the speaker. You should have a very low reading in the milli-volts. If there is an ultrasonic oscillation, you won't hear it but you should measure several volts at the speaker. The feedback resistor goes from the speaker jack back to the Presence control. Try the amp with that resistor disconnected. The gain of the amp will increase but see if the high end comes back.

If the above experiments don't reveal the problem I would set the two amps next to each other and cross connect them at the output from the Treble control. To clarify, disconnect the power amp at the Treble control of the good amp and connect it to the Treble control of the bad amp. Do the same with the bad amp. Did the problem stay with the bad amp, move to the good amp, or slightly affect both amps?

Thanks so much for the suggestion above. The measurement is when the amp is on and at idle, correct?

I didn't cross the leads coming from the PI to the power tubes but they do cross OVER the leads that go to the PI plates. I always try to space (in 3D) wires to avoid interference. Looking at alot of old Fender layouts, the leads go from the PI, under the board and to the output tubes. Mine stays on top, parallel to the power tube coupling caps. I'll try to get a picture up of my layout.

Like I said earlier, the layout is slightly different between the two. I used to mount the grid stoppers (5k6) on the board, then read that they are best at the socket. been putting them there ever since.

Just thought of something and feel kind of stupid. I use solid core, PVC wire for my builds and this one I believe that is what I have that goes from the volume control, under the board to pin 2 of V2. It's probably that wire that is most susceptible to interference. I'll swap that with shielded tonight.

Thanks for all the tips. Hope I find a solution before tearing up this otherwise nice looking piece.

[Alex R]

On the scope issue - lots of scary knobs I know but the basic function is simple and it is ok to experiment. You need to identify:

Brightness and focus controls, function obvious
knob to move the trace up and down on the screen
frequency control, to display a readable number of waveforms
amplitude control, to get a readable height of waveform

That's about it really. Set the ground to AC to keep the waveform returning to the screen, twiddle the trigger till the wave steadies and there you go.

[loudthud]

I didn't cross the leads coming from the PI to the power tubes but they do cross OVER the leads that go to the PI plates. I always try to space (in 3D) wires to avoid interference. Looking at alot of old Fender layouts, the leads go from the PI, under the board and to the output tubes. Mine stays on top, parallel to the power tube coupling caps. I'll try to get a picture up of my layout.

If you've read many of the "I built an amp and it doesn't work right" threads, you'll frequently see the suggestion to reverse the (usually) brown and blue wires to the output transformer from the output tube plates. Assuming your first builds are correct, you would have the same problem with the new layout if you reversed the wires from the phase inverter plates or at the grids of the power tubes. This would turn the negative feedback into positive feedback creating an oscillator. The oscillation is usually at an ultrasonic frequency, easily seen with an oscilloscope or detected as AC voltage by a DVM.

The measurement is when the amp is on and at idle, correct?

Yes. You are looking for an ultrasonic oscillation. Disconnecting the feedback will stop the oscillation if positive feedback is the cause.

[HBamps]

If you've read many of the "I built an amp and it doesn't work right" threads, you'll frequently see the suggestion to reverse the (usually) brown and blue wires to the output transformer from the output tube plates. Assuming your first builds are correct, you would have the same problem with the new layout if you reversed the wires from the phase inverter plates or at the grids of the power tubes. This would turn the negative feedback into positive feedback creating an oscillator. The oscillation is usually at an ultrasonic frequency, easily seen with an oscilloscope or detected as AC voltage by a DVM.


Yes. You are looking for an ultrasonic oscillation. Disconnecting the feedback will stop the oscillation if positive feedback is the cause.

I forgot to mention, this amp has no negative feedback, and is cathode biased.

I added shielded wire from the vol to V2 and moved some flying leads around to make space between all. I couldn't test though, sleeping 1.5 year old daughter usually doesn't dig loud riffing will test soon.