5B3 deluxe refurb- advice needed

[Aaron Smith]

Greetings all... I've got a 5B3 that I'm refurbishing, and I could use a little assistance. Background on me: I've been messing with tube amps on and off for 7-8 years. I've built a few simple ones from scratch (AX84, 5F1, etc.) and I've refurbished a few broken old ones (Silvertone 1482, etc.) I'm a biomedical engineer and I studied quite a bit of electrical engineering, but I've since forgotten most of the hard stuff.

The amp is a bit of a mutt. Original transformers are gone, but they were replaced many moons ago with a 125P1A PT (Harvard/Brown Princeton), and a 125A10B OT (blackface Princeton Reverb). The electronic components on the board were mostly original, except for the filter caps... they were replaced with some Sprague's sometime during the disco era.

It was playing when I got it, but just barely. Had some bad filter hum, so I replaced the filter caps and B+ dropping resistors. Definitely cleaned up the hum, but it's still not a good sounding amp. For starters- it is very quiet. About 1/2 the volume of my SF Princeton Reverb, for comparison. It also has zero clean headroom; it goes from off to distorted.

My main question today is about the power supply. To reiterate, the filter caps and dropping resistors were just replaced. Here are the voltages I measured, and what they're supposed to be per the 5B3 schematic/layout:
B1+ = 361V (350V)
B2+ = 302V (280V)
B3+ = 283V (250V)

There's only an 11V difference between measured and desired for B1, but the difference gets bigger with B2 (22V) and B3 (33V). Any idea what's going on here?

All the downstream voltages are wacky too, but I'm sure I need to replace a lot of bad resistors and caps- almost all of which are original.

Finally- how do you post pics on this forum? Thanks in advance

[Cliff Schecht]

Before you go swapping too many parts blindly (essentially wrecking any value that's left in the amp), I'd figure out why the amp isn't drawing enough current. What shape are the tubes in. IME the power supply and plate resistors are what go bad, sometimes other parts go randomly but it's usually only a few, not the whole lot of them. It sounds like you have a leaky signal capacitor, this will cause the signal to have an unwanted DC component that gets amplified through the next stage and causes all sorts of weirdo clipping. It saps volume and makes things sound farty and distorted, i.e. no headroom like you are dealing with. The easiest way to test these parts in circuit is to lift one side of the capacitor (ideally the side that should have no DC, like the side of the cap that feeds the grids of the next stage), solder a 1 meg resistor from that floating end to ground and measure for DC. There should be none or the cap is considered leaky.

You can also pull tubes and measure the value for most of the resistors (unless they have something connected to both ends). Did you change the cathode bypass capacitors on all of the stages? These guys surely go bad too after 60-70 years (especially considering they have a shelf life of 10-20 years!!).

[Phil_S]

...measured, and what they're supposed to be per the 5B3 schematic/layout:
B1+ = 361V (350V)
B2+ = 302V (280V)
B3+ = 283V (250V)

There's only an 11V difference between measured and desired for B1, but the difference gets bigger with B2 (22V) and B3 (33V). Any idea what's going on here?

Food for thought. When that amp was built, line was probably 110 or 115. Today you probably get 120 or more. I've read reports of people getting 130. This means B+1 at 3% over and B+2 at 8% over are within expectations. Building on what Cliff says, B+3 at 13% over is where I think your problem with current draw exists. You may also have a problem on the B+2 node (power tube screen supply), but I'd concentrate on the preamp and PI sections first.

[Aaron Smith]

Thanks for the feedback. Turns out I made the typical newb mistake... just now I tried a new set of power tubes, and the voltages are spot-on perfect. Plugged in, volume is much higher. So I guess my big problem was bad power tubes (duh...)

There is still some other stuff going on... a static-y whoosh sound whenever a loud bass note is played, for instance. I also have a fair amount of noise. Will check the signal caps and resistors around the input to see if I can find anything amiss.

[Aaron Smith]

Tonight's work... replaced both 25uF cathode bypass caps, replaced the .05uF coupling cap and 5M grid leak resistor on the instrument input. Results- strange ghost distortion is gone, and she's playing at full volume. Couldn't really wind it up as the kids are in bed, but I like what I'm hearing so far. Still has a fair amount of noise that goes up when the volume knob goes up; tomorrow I might try swapping the 50k input resistor.

Making progress... thanks for the help!

[Prairie Dawg]

On a lot of Fender schematics it tells you that voltage can be as much as +/- 20 per cent from the schematic. Nothing to worry about.

[Cliff Schecht]

Tonight's work... replaced both 25uF cathode bypass caps, replaced the .05uF coupling cap and 5M grid leak resistor on the instrument input. Results- strange ghost distortion is gone, and she's playing at full volume. Couldn't really wind it up as the kids are in bed, but I like what I'm hearing so far. Still has a fair amount of noise that goes up when the volume knob goes up; tomorrow I might try swapping the 50k input resistor.

Making progress... thanks for the help!

A couple of tricks to reduce the noise.. First and foremost you can lift the heater centertap (or the artificial centertap from two 100 Ohm resistors) and connect it to the cathode of the power tube. This elevates the heaters with a DC component that tends to reduce capacitive coupling of the heater noise into the cathodes of all of your tubes (preamp and poweramp).

Also keep in mind that single-ended topologies have no inherent noise cancelling like push pull amps do. More filtering never hurts and I usually up the first filter cap to at least 47uF as a starting point. If the rectifier tube can handle it you could increase this value, or if you are using a SS rectifier then you can go crazy here (although past a certain point there are diminishing returns).

If your amp uses a 6SC7 then you might consider finding some spares. Most all octal preamp tubes are notoriously noisy little fuckers. Problem is, 6SC7's are used in a LOT of equipment and good ones command a premium. I personally never use these, I instead use the 6SL7 which is the same tube (electronically) with a different pinout. You can even build an adapter to switch between these two or even build an 8 to 9 pin adapter and use 12AX7's instead. I spend a lot of time sorting through my octals to find the low noise ones and it sucks that so many (maybe 1 out of 5 to 7) are actually usable for, say, V1 of a higher gain amp.

[Aaron Smith]

Tonight's work... replaced both 25uF cathode bypass caps, replaced the .05uF coupling cap and 5M grid leak resistor on the instrument input. Results- strange ghost distortion is gone, and she's playing at full volume. Couldn't really wind it up as the kids are in bed, but I like what I'm hearing so far. Still has a fair amount of noise that goes up when the volume knob goes up; tomorrow I might try swapping the 50k input resistor.

Making progress... thanks for the help!

A couple of tricks to reduce the noise.. First and foremost you can lift the heater centertap (or the artificial centertap from two 100 Ohm resistors) and connect it to the cathode of the power tube. This elevates the heaters with a DC component that tends to reduce capacitive coupling of the heater noise into the cathodes of all of your tubes (preamp and poweramp).

Also keep in mind that single-ended topologies have no inherent noise cancelling like push pull amps do. More filtering never hurts and I usually up the first filter cap to at least 47uF as a starting point. If the rectifier tube can handle it you could increase this value, or if you are using a SS rectifier then you can go crazy here (although past a certain point there are diminishing returns).

If your amp uses a 6SC7 then you might consider finding some spares. Most all octal preamp tubes are notoriously noisy little fuckers. Problem is, 6SC7's are used in a LOT of equipment and good ones command a premium. I personally never use these, I instead use the 6SL7 which is the same tube (electronically) with a different pinout. You can even build an adapter to switch between these two or even build an 8 to 9 pin adapter and use 12AX7's instead. I spend a lot of time sorting through my octals to find the low noise ones and it sucks that so many (maybe 1 out of 5 to 7) are actually usable for, say, V1 of a higher gain amp.

Cliff, thanks for your advice. I have the amp sounding pretty good, but I still have a fair amount of 60 cycle hum when I wind it up; I'm assuming that is heater hum. My power transformer doesn't have a heater center tap- at least one that I can find. Can you talk me through installing the artificial one with the 100 ohm resistors? Do I just parallel one resistor off of each of the heater wires? Is it better to take them to a chassis ground, or to the power tube cathode? If cathode, do I take it right to the power tube pin, or to the ground side of the cathode resisitor? If it's not already obvious, I'm not entirely sure what I'm trying to accomplish by putting DC on the heater circuit. Thanks for your patience.

[martin manning]

The two 100R resistors create an artificial center tap which centers the heater AC voltage around whatever the junction of the resistors is tied to. The simplest thing to do would be to put one 100R from pin 7 and one from pin 2 to pin 8 (cathode) on one of the output tube sockets. You could try it this way, and also taking the junction of the 100R to chassis ground (the other side of the cathode resistor or a ground lug placed under one of the socket attachment screws) and see which is better.

[Aaron Smith]

The two 100R resistors create an artificial center tap which centers the heater AC voltage around whatever the junction of the resistors is tied to. The simplest thing to do would be to put one 100R from pin 7 and one from pin 2 to pin 8 (cathode) on one of the output tube sockets. You could try it this way, and also taking the junction of the 100R to chassis ground (the other side of the cathode resistor or a ground lug placed under one of the socket attachment screws) and see which is better.

Thanks for your help Martin. Starting to make sense to me now...

[Firestorm]

The 5B3 originally grounded one side of the heater supply to the chassis (which is the cheap and dirty way to give heaters a ground reference). Before you do any of the things suggested, make sure this is no longer the case and it has individual heater wires for each side. If you have to make this change yourself, note that one terminal on the lamp is also grounded and must be unsoldered from its bracket.

[martin manning]

Good point. Check continuity to ground from any heater pin to see if the string is grounded somewhere already.

[pdf64]

It may be obvious if there's only one heater wire going from tube to tube, rather than a twisted pair.
But before messing, bare in mind that improvements to the heater arrangement will only reduce noise from the heaters, ie 60Hz hum and possibly buzz.
It can't help with 120Hz hum from the B+, or hiss.
So it may be best to positively identify heaters as being the source prior to proceeding.
Pete

[Aaron Smith]

It may be obvious if there's only one heater wire going from tube to tube, rather than a twisted pair.
But before messing, bare in mind that improvements to the heater arrangement will only reduce noise from the heaters, ie 60Hz hum and possibly buzz.
It can't help with 120Hz hum from the B+, or hiss.
So it may be best to positively identify heaters as being the source prior to proceeding.
Pete

Good suggestion... how would I isolate that the hum is indeed coming from the heaters? I only have the noise with V1 in place, and it goes up proportionally with the volume knob. I assumed it was heater hum from V1 that is being amplified through the subsequent gain stages. I have tried 4 different tubes in V1, they all had the same noise.

There is also some hiss, but I'm going to address that separately. Think the old carbon comp input resistors might be the culprit there.

[Aaron Smith]

Here is a gut shot, FWIW. You can't see it in the pic, but there is only one heater wire going from tube to tube, not a twisted pair.

[img764]http://i305.photobucket.com/albums/nn20 ... 69ccd7.jpg[/img]