Silvertone 1335

[mresistor]

Anyone know where I can find a schematic of the Silvertone 1335. It has two 6L6G, 5Y3, 12AX7, 6SL7. Two sets of triple inputs, mic and intrument.  And tremolo. 

[sluckey]

I used google. It seems to be available on a couple pay sites.

[mresistor]

Steve  I was able to get a copy from the Radio Museum.   Do you want a copy to post to the schematics library?

[mresistor]

I have a barn find Silvertone 1335 on the bench, I had to sterilize it amd clean it because of mouse droppings and spiders. First thing I found was a shorted RCA 5Y3GT. I stuck a sovtek 5Y3 (5C4M actually) and the amp works with some really really noisy volume pots. The trem isn't functioning but now that I have the schematic I think I'll stick a jumpered 1/4 plug in the jack and see if it works.

[sluckey]

It has two 6L6G, 5Y3, 12AX7, 6SL7

Your schematic shows one more tube than you originally listed?

[Paul1453]

This circuit also seems to have a different kind of fixed bias then I've seen before.

Instead of using - voltage on the grid, they seem to have elevated the cathode voltage to accomplish the same thing. 

But that cathode voltage also seems to be used as the 12AX7s heater voltage.   

This is unusual, and throws me for a loop on understanding how it works.   

Sluckey, or one of the other experienced guys, care to explain what is going on with that?   

[HotBluePlates]

This circuit also seems to have a different kind of fixed bias then I've seen before. Instead of using - voltage on the grid, they seem to have elevated the cathode voltage to accomplish the same thing.

But that cathode voltage also seems to be used as the 12AX7s heater voltage.

Cathode bias, not fixed bias.

Pins 4 &5 of a 12A_7 are used when you operate the 2 internal heater segments in series, rather than parallel. For this operation, the heater runs at 12.6v @ 150mA per heater.

There are 2 tubes with their heaters in series. It will still run at 150mA, but should be 25.2v total. But apparently, with the plate and screen voltages indicated, the 6L6 output tubes idle at 150mA per pair (75mA per tube) with 19v of bias. So each tube's heater sees 19v/2 = 9.5vdc rather than 12.6v. That's low, but apparently the free d.c. heating for these tubes was judged more important than nailing the correct heater voltage.

The key thing to remember is you can assume the output tubes are idling at/very near 150mA total. The idle current is d.c., and therefore the preamp tube heaters see d.c. for their heaters (reducing the chance of hum).

[Paul1453]

OK, so since this circuit doesn't have an adjustable bias we call that Cathode bias?
Or, is it cathode biased because the heater filaments are actually the cathode bias resistor?

With regular cathode bias, we normally have a large wattage resistor on the cathode of the output tube.

I also don't know why they call it fixed bias, when we use a pot to adjust the bias setting. 

To me, cathode biased seems to fit the "fixed bias" label better.
It's bias is fixed by the value of the resistor on it's cathode.

On a fixed bias circuit, we adjust the bias with the - voltage we supply it.
That term just doesn't seem to properly reflect the circuit's functionality, IMO.
Maybe it's a holdover term from earlier developments?

[HotBluePlates]

OK, so since this circuit doesn't have an adjustable bias we call that Cathode bias?
Or, is it cathode biased because the heater filaments are actually the cathode bias resistor?

With regular cathode bias, we normally have a large wattage resistor on the cathode of the output tube.

I also don't know why they call it fixed bias, when we use a pot to adjust the bias setting. ...

Cathode bias, because the grid is held at ground potential (through resistors to ground through which no appreciable current flows), and the cathode is elevated to some voltage above ground. The net result is the grid is negative of the cathode.

In fixed bias, we typically held the cathode at ground potential and apply some negative voltage to the grid. The net result is the grid is negative of the cathode (same as in cathode bias).

There are a lot of ways you can cathode bias. You might stack the cathode on a heater string to power the heaters from output tube cathode current (as here). For small tubes, you might use an LED between cathode & ground (the LED turns on at a particular forward voltage). Or you might simply insert a resistor between cathode & ground, using cathode current to define a voltage drop which biases the tube.

That last one is often called "self-bias" or maybe "automatic bias" because increased cathode current develops a bigger voltage drop, and the resulting bigger grid-to-cathode voltage difference tends to turn the tube more-off, lowering cathode current. So it's "self-adjusting". If a fault happened causing a big enough cathode current to burn the cathode resistor open, all cathode current would be stopped, so resistor-derived cathode bias is inherently safe and self-limiting.

By comparison, the typical fixed bias of an applied negative grid voltage is not self-adjusting based on tube current. If you set it wrong, or if there's a fault condition, it will cheerfully let your tubes melt (plus whatever other damage is caused). You might think of it as adjustable, if you have an adjustment pot, but many fixed-bias circuit don't have such a pot and at no time are you fiddling the pot continuously for every little current variation. So by comparison to resistor-derived cathode bias, it is "fixed".

[Paul1453]

Thank you HBP!   

A clear explanation of the concepts involved.

I knew cathode bias was inherently safer for malfunctions.

It seems that most High Watt designs tend to be fixed bias.

I'm guessing that was mostly a manufacturing cost decision to eliminate multiple high wattage resistors then.

[HotBluePlates]

... It seems that most High Watt designs tend to be fixed bias. I'm guessing that was mostly a manufacturing cost decision to eliminate multiple high wattage resistors then.

You can't bias an amp to class B (idle at 0mA with huge peak plate currents) with cathode bias; 0mA at idle is 0v of bias which is full-on.

You can bias a class A amp with cathode bias, and if the tubes and output stage are perfectly balanced (with zero distortion), the reduction of current in one side is exactly offset by the increase in current of the other side when signal is applied. Bias stays perfectly steady (because net current through cathode resistor is constant).

You can push into class AB with cathode bias. There will be some shifting of the bias point with large outputs, because the peak current of one side is more than twice the idle current of that side. Therefore it's impossible for the net current to stay completely steady under all signal conditions. Distortion will also make the net current change (usually an increase).

Going far into class AB (meaning closer to class B than class A) doesn't work well with cathode bias. The tubes run too hot over the signal cycle, which you'd typically resolve by increase the negative bias (reducing tube current), making the tubes cut off for a bigger part of the signal cycle (and allowing them to cool). But with cathode bias, your only means of increasing the bias voltage is to increase tube current. So the desired effect and the means to get there are at odds!

And when you consider peak plate current in one of those "lean class AB amps" will often be 3-4 times the idle current value, but off is still off, you'll see how the net current through the cathode resistor will rise sharply even if you could cathode bias into lean class AB. So all the big amps, using lean class AB to produce a lot of power for the given tubes, use fixed bias. There's really no other good choice.

[mresistor]

It has two 6L6G, 5Y3, 12AX7, 6SL7

Your schematic shows one more tube than you originally listed?

Yes in my haste to post, I got it wrong. The amp tube compliment on the main chassis is;


5Y3GT, 6L6G, 6L6G, 6SN7GT, 12AX7, 6SJ7


This being a split chassis amplifier, there is also a 12AX7 preamp tube on the input chassis above.


Pics coming.

[mresistor]

A true attic find...

[mresistor]

Notice someone has replaced one of the 6L6G tubes with a 6L6GC. In the past, the amp has been spray-painted black. And the carry handle is missing. One knob and on/off switch are missing.

[mresistor]

HPB - thanks for the discussion of the circuit. The engineers were really thinking when they implemented the cathode bias on this amp through the 12AX7 heaters.  Kind of cool. I'd never seen that before.

[Paul1453]

HPB - thanks for the discussion of the circuit. The engineers were really thinking when they implemented the cathode bias on this amp through the 12AX7 heaters.  Kind of cool. I'd never seen that before.

Seconded, HBP!   

Your clear explanations "in a language we can all understand" (Living Color) are invaluable!

Yes, this design seems like something I might like to copy.   

Your explanation of the principles involved also gives me an idea why the heaters seem underpowered at first.   

As you described, when the amp is played hard the current and voltage should rise on those heaters, No?

If they were set higher at idle, they might exceed recommended heater voltages when played.   

[HotBluePlates]

Your explanation of the principles involved also gives me an idea why the heaters seem underpowered at first.
As you described, when the amp is played hard the current and voltage should rise on those heaters, No?

Probably not, as the amp is almost certainly class A. One output tube's current increases towards 2*idle, while the other output tube's current drops towards zero. Net cathode current through the heaters should remain unchanged.   (Okay, it'll try to rise slightly because of screen current rise at full output, but that's small compared to total cathode current in this case.)

... If they were set higher at idle, they might exceed recommended heater voltages when played. ...

I could try to theorize, but it will be easier if Mresistor simply measures cathode voltage while playing. I'm gonna guess it doesn't change much, if at all.

... Yes, this design seems like something I might like to copy. ...

You have to really, really want d.c. heating of the preamp tubes by way of cathode current & bias of the output tubes to try to copy this or any similar plan. That's because everything else about the power supply & output stage largely gets dictated to you by making this choice. In other words, everything else has to bend to fit whatever this choice dictates, just to make it work.

There will be no such thing as "modding an amp to use this idea," or "using whatever parts I have on hand." Class of operation, OT load impedance, output tube bias voltage, output tube idle current, and supply voltages (because of the dictated bias & idle current) are all pretty well set in stone because of the choice to use this particular d.c. heating trick. As a result, the requirements for the phase inverter and preamp are largely driven by the resulting output stage.

[Paul1453]

Is Class A operation a bad thing for a guitar amp?

I know very few are designed that way.

If you had an aggressive preamp like the Plexi circuit,
you could still get sweet 12AX7 distortion,
and extremely clean headroom and output, if desired, from Class A operation, No?   

[HotBluePlates]

Is Class A operation a bad thing for a guitar amp?

I know very few are designed that way.

All single-ended amps, plus a portion of the push-pull amps (especially the older models of the early 50's) are class A designs. Nothing inherently bad or good about it; class AB just allows more power output from the same bottles.

Every guitar amp preamp I can think of operates in class A.

If you had an aggressive preamp like the Plexi circuit,
you could still get sweet 12AX7 distortion,
and extremely clean headroom and output, if desired, from Class A operation, No?   

Class of operation, by itself, doesn't dictate if the amp will be clean or distorted.

[Paul1453]

OK, my lack of depth of knowledge is showing again.   

I thought that Class A operation has the least distortion,
with the following Classes of operation allowing more and more distortion while still meeting their specs.   

I see Hi-Fi tube amps touting Class A operation.   

[HotBluePlates]

Notice someone has replaced one of the 6L6G tubes with a 6L6GC. ...

Note the the smaller "6L6GC" tube is really a Russian 6π3C (6P3S) tube, which only has a ~20-21w plate dissipation rating.

There are a lot of those floating around with that label, but not carrying the specs implied by the label.

[mresistor]

You have a good eye HPB. It sure enough is a rebranded 6n3c. I got one of my newer ones out to compare. I think I'll put another one in this amp. What's cool about this one is it says "made in USSR" on it.

[HotBluePlates]

... What's cool about this one is it says "made in USSR" on it.

What's really cool is National/Richardson Electronics actually put the correct country of origin on it.

Mostly National/Richardson rebranded & sold tubes from other manufacturers. It seems that when the Cold War was a little colder, Russian-made tube would get labels claiming British or German origin. I used to have (I think I got rid of them) such 6P3S tubes marked "Made in England," etc.

There is an amp for sale on another forum, in which the seller appears to tout the "NOS tubes" that will come with the amp, including 2 "6CA7" tubes marked "Made in Germany". Except "6CA7" is a U.S. type number, typically coming in big 6L6GC bottles. And these were skinny-bottle EL34's. And they were marked "National Electronics" which meant almost certainly they were Russian EL34's with false labeling.

Anyway, this does look like a cool amp, and the good news is the 6P3S tubes should be quite happy in that particular amp.

[mresistor]

Gut shots    as it's possible some of you would like to see what the innards look like. Interestingly the OT is mounted inside the main chassis.

[HotBluePlates]

Notice someone has replaced one of the 6L6G tubes with a 6L6GC. ...

Note the the smaller "6L6GC" tube is really a Russian 6π3C (6P3S) tube ...

You probably already know, but the "5Y3" is also a Russki tube, not an actual 5Y3.

Interestingly the OT is mounted inside the main chassis.

Yep, no room to put it on the outside! If it works, it works...

Anyway, everything on the inside looks 100% stock.

[mresistor]

HPB -  that rectfier is my 5C4M  that I was using to see if the amp would fire up originally. The original RCA 5Y3 is shorted and the 6L6G is also shorted, all the other tubes check good. I'll be using a real 5Y3 in it from here on out. Maybe even a 6087. Those 60 year old electrolytic caps have to be replaced. 2 - 10uf 450v and 2 - 16uf 450v . One cap is a twin cap. The owner wants to play this amp. He's a performing musician.


When I got the amp I noticed that the 12AX7 socket in the control panel was really loose, and when I got it opened up, the nuts holding it in place on the bolts were backed off and the whole things was wobbling around. I still wonder why or how nuts can come loose like that. The nuts had to loosen up going against gravity. My theory is someone, probably the same person who took the on/off switch out, had loosened that socket mount up and forgot to tighten it up. (for what reason I have yet to discover)  The main chassis also had two mount nuts loose.

[mresistor]

Well I got this amplifier working and the tremolo works well, but I am concerned about the voltages. Looks like I may have to pull the chassis again and check or change some resistances.
I am getting 330V on pin 8 of the 6SJ7 and that exceeds the max rating of 300v.


Another consideration, would it be wise to upgrade the power cord to a grounded cord in this amp? It's a split chassis and it appears like the upper chassis is only grounded through an rca jack for the tremolo signal.


I have replaced all 4 of the electrolytics and I am still getting some pretty good crackling and low level hiss noise. Do you think it might be related to the voltages being too high? The plate voltage on the input 12AX7 is a little low at 120V though, which is good thing that it is not too high.


I'm going to check R23, R24 and R25 .

[sluckey]

I am getting 330V on pin 8 of the 6SJ7 and that exceeds the max rating of 300v.

I would not worry about that.

would it be wise to upgrade the power cord to a grounded cord in this amp? It's a split chassis and it appears like the upper chassis is only grounded through an rca jack for the tremolo signal.

Grounded power cord for sure. Then run a dedicated heavy duty ground wire between the two chassis.

...crackling and low level hiss noise...

That could be lot's of things, tubes, resistors, caps, etc.

[mresistor]

Thanks Steve. 


I replaced the power cord and the strain relief grommet. This sucker is grounded now. Replaced the first PS node 2.2K dropping resistor because it was reading 750 ohms. Also replaced the second PS node dropping resistor which read 13K, but replaced them with 2 watt resistors. I cannot believe they were half watt.  Upper chassis is grounded good.
Fired it up and still have a lot of raspy noise. Probably some caps or resistors need replacing, but as mentioned, it can be anything. Going to have to go through this amp with a finer comb.

But it sounds really really nice, apart from the noise.

[mresistor]

old National 12AX7 on main chassis was noisy,  replaced and amp alot quieter on the left inputs, but the right side inputs on the control panel are injecting background static noise. At least I have in narrowed down to a lot less components and physical contact points.

[HotBluePlates]

old National 12AX7 on main chassis was noisy ...

Got a pic of that tube?

I've come to have a low opinion of National Electronics as tube-rebranders (in mostly the least-nice way) but I haven't seen a 12AX7 with their label. Just curious as to what it really is.