5F6A Bassman total custom build - super low output

[tomgilmartin]

There are inevitably bugs to work out on a build, so don't get discouraged. Your experience is not singular. Actually, given the unconventional layout of that amp I'm quite surprised there aren't more bugs.

I'm not in the least discouraged! This is fun for me. I love wiring stuff up.

Twisted pair noise rejection only works if the current in the two wires flows in opposite directions. It's strange to me how tube amp designers of the Fender/Marshall era only worry about the positive side of the signals. I thinks it's because they were trying to make the amps as cheap as possible to manufacture. In industrial instrument wiring you NEVER use the chassis to carry a signal ground. Only Shields go to chassis.

[J M Fahey]

Thanks Chuck. It has a funny sort of sizzle at high treble. Still have the bias super cool. I'm not really blown away with the "chime" I was hoping for.

No chime to be expected from Hi Fi speakers with incredibly thick and heavy cones, incredibly heavy voice coils, and 10dB less sensitivity than whatīs expected in a Guitar speaker.

The original Bassman was designed, developed and tuned-by-ear using 4 Jensen P10R , as different from your as can possibly be.



Parameters ..... Jensen ..... Dipole-10
Resonance ..... 99 Hz ....... 26 Hz
Moving mass .. 13.7g ....... 54.6g
X max ........... 0.8mm ..... 12mm
Sensitivity ..... 95.1 dB .... 86.5 dB

Pity Dipole10 frequency resīponse is not available, but being an expensive Hi Fi woofer I bet itīs ruler flat within the low band, has little mids and absolutely no treble, while Guitar oriented Jensen P-10 is "pre equalized" with full bass and treble boost built in.
Just look at its frequency response curve:

[tomgilmartin]

No chime to be expected from Hi Fi speakers with incredibly thick and heavy cones, incredibly heavy voice coils, and 10dB less sensitivity than whatīs expected in a Guitar speaker.

The original Bassman was designed, developed and tuned-by-ear using 4 Jensen P10R , as different from your as can possibly be.

Parameters ..... Jensen ..... Dipole-10
Resonance ..... 99 Hz ....... 26 Hz
Moving mass .. 13.7g ....... 54.6g
X max ........... 0.8mm ..... 12mm
Sensitivity ..... 95.1 dB .... 86.5 dB

Pity Dipole10 frequency resīponse is not available, but being an expensive Hi Fi woofer I bet itīs ruler flat within the low band, has little mids and absolutely no treble, while Guitar oriented Jensen P-10 is "pre equalized" with full bass and treble boost built in.
Just look at its frequency response curve:
[ATTACH=CONFIG]46767[/ATTACH]

Great stuff, I agree the speakers are totally different. At this point, it sounds pretty good!

[tomgilmartin]

Can someone explain to me why guitars cords are made with coaxial cable? It would seem any noise picked up on the shield would appear on the hot leg anyway through the coil of the guitar. How does this give any semblance of noise attenuation? <NO SARCASM, I DON'T UNDERSTAND THIS.>

I suppose because it's the only alternative for the unbalanced output of a guitar. Only other choice when you have only two connection points at each end is twisted pair. One could use shielded twisted pair, but for the shielding to work properly the shield would have to be connected at one end only, and that end would have to be plugged into the amp end. If the shield ground end were plugged into the guitar end, then shield noise would couple through the negative line and thus into the hot leg of the signal.

[eschertron]

There are certainly better ways to shield the signal in a guitar cable, but decades of tradition... sigh.

There's a relatively high impedance through the pickups vs. to the chassis via the grounded side of the unbalanced signal. Think of the cable shield as an extension of the chassis itself, there will always be noise 'cast at the amp, but path of least resistance sends that noise signal to ground (well, most of it, most of the time!).

[tomgilmartin]

Attached file is an mp3 of a frequency sweep of the speakers. Driven from a phone with a function generator app through my studio amp. Recorded on another phone. So there may be some phone artifacts. File is mono. If you want to get the frequency response, the sweep is 20sec for 20hz-20khz.

To analyze for frequency then:
0.1sec = 100Hz
1sec = 1kHz
5sec = 5kHz
ETC.

Checked this with my old software CoolEditPro to see the waveform and do Frequency Analysis.
Seems like rolloff doesn't start until about 4kHz, then drops off, then peaks again at 8.2kHz.
Not an anechoic chamber test, nor high-tech equipment, but it gives some idea of the response.



Question: How do I get the site not to log me out all the time?

[tomgilmartin]

New pics with all wire shielded twisted pair. It took me about 12 hours actual time to wire all this. Pretty much one hour per shielded wire.
Everything is shielded except these signals which are all unshielded twisted pair:

• Inverter output to power tubes
• DC power from power section to preamp section
• Tube heater filaments
• Choke wires to DC bus
• Power Amp to OT
• Speaker output from OT to phono connector.

[Leo_Gnardo]

One could use shielded twisted pair, but for the shielding to work properly the shield would have to be connected at one end only, and that end would have to be plugged into the amp end. If the shield ground end were plugged into the guitar end, then shield noise would couple through the negative line and thus into the hot leg of the signal.

I've tried building guitar cable like that, the "proper" way with a separate interior line for signal ground, shield NOT served at the guitar end of the cable. Tried several cable types too, Belden 8412 mic cable, Belden "Brilliance" quad conductor cable, Canare quad, Mogami & some other brands too. It only seems to make a marginal audible improvement in the most noisy environments, even then it's barely detectable over ordinary competent guitar coax, Conquest USA-1 for instance. There's an analogous method for mic cables, called "Westrex wiring." For a balanced cable you need 3 interior conductors plus shield. Here again, not much difference can be heard unless the stage/studio is in a noise polluted zone such as southern Manhattan. By "noise" I mean electrical fields present from AC wiring 50-60 Hz, radio transmissions at all imaginable frequencies, then up to and beyond microwave. I've done work in some NYC studio locations where entire rooms are lined with grounded copper screen to turn them into Faraday cages. In environments like that, copper screening the whole room is the only thing that seems to help.

[tomgilmartin]

Smoke test passed, but no change in sound. If anything it's worse. More hum than before. Maybe I have a mistake or a shorted shield somewhere. Will check tomorrow.

[Chuck H]

One big leap in a new direction. Two steps forward. One step back. There will be sideways steps too. As long as the forward steps continue to outnumber backward or sideways. Realize that you're brand new to this, doing something "out of the box" (my theme for this thread ) and killing it

[g1]

Question: How do I get the site not to log me out all the time?

There should be a box to check off that says 'remember me' near the login button.

[tomgilmartin]

I'm wondering about this sizzle noise I hear. I have made a lot of mistakes building this amp.

When I received the speakers I decided to do some slap bass with them just sitting on the railing. No, I didn't drop one, but I did blow one up. Huge loud pop! Then I could hear it rubbing. Sanded down the voice coil a bit to fix the rub, but I can still hear it. I wonder if this is causing a feedback, because it effects the movement of the speaker, which will be present on the leads for sure, then coupled back through the neg feedback. I can see the second and third harmonic in the measured output of the speakers, check out the file I sent.

Also, while testing DC during early startup, I managed to cause an arc. Twice. I may have damaged a power tube.

[tomgilmartin]

Frequency response to pink noise. Through the studio amp, close mic. Recorded on phone.

[tomgilmartin]

Still puttering with it. Found a crack in a signal ground that had touched the shield. Fixed that, now it sounds like before. Disconnected my shield ground point and shields show infinite resistance to ground, so no touch points between shield and signal ground except the shield ground point.

Next up, the output from the PI to the power tubes. I was poking at the wire with the chopstick method (a Sharpie marker) and seemed to help. Looking again at it, the input and output of the power tubes are running parallel for quite a while. Perfect place to pick up a feedback. I might go all the way and shield the output too. I'll try the input first, but I used up the entire sleeve of solder I bought! Gotta get some more.

Even if nobody answers, writing it out here is helpful to force me to think clearly about it. Thanks again to all who have helped.

[tomgilmartin]

THAT WAS THE ONE!!! Sizzle GONE! Clean tones abound! Crunch tomorrow when people are not sleeping here. Audio attached.
Shielded PI Output.mp3