Silvertone 1482 - Why so distorted?

[dbishopbliss]

I bought an old Silvertone a while back and started making some changes... then life got in the way. I finally got around to finishing the wiring but now the amp sounds really bad.


I replaced the can filter cap with 3 discrete caps of similar values. I routed the last cap as suggested by sluckey. I also took the filament center-tap and raised its potential by grounding at the cathode of the power tubes (above the 270R cathode resistor). Other than that, I have not made any changes to the circuit (I talked about making some changes, but I haven't done any of those).


Now that amp is VERY distorted. The HT voltages are a little high compared to the schematic but not horrible considering the line voltage is probably higher - 368 vs 350, 350 vs 330 and 280 vs 260. All other voltages are similarly high except one is way out of whack... measuring the cathode pin 8 on one 6V6 is 57 volts and the other is 18 volts. They use a common cathode resistor so I was surprised to see two different voltages. I'm guess this is the problem.  Bad tube or something else? I don't have any other 6V6s around to test at the moment.


Here is what it sounds like.


https://youtu.be/TLAgo65w0Mc


Thanks for your help.

[sluckey]

measuring the cathode pin 8 on one 6V6 is 57 volts and the other is 18 volts. They use a common cathode resistor so I was surprised to see two different voltages. I'm guess this is the problem.

I guess so too. How can they be different voltages? They are supposed to be tied together. Apparently they are not. Look for a bad solder joint or broken wire.

[dbishopbliss]

Look for a bad solder joint or broken wire.

That was one problem. I couldn't see that the lead had broken off and the wire was being held on by melted insulation.


However, it still sounded terrible once I rewired the cathode. Then I swapped speakers... aha! Its the speaker. Which is strange because I just took it out of the box tonight for the first time. I will be contacting the company. Hopefully they won't care that I bought it back in January.


Another issue I just noticed... measuring across the heater wires (one yellow lead to the other) I am only getting like 5.1 Volts AC. Could this be causing the flubby bass issue or do you think that has to do with the Phase Inverter Grid Stopper (or lack thereof).

[sluckey]

I would expect 6.3v on the filament string. Unplug all tubes. What do you measure between the yellow wires now?

[dbishopbliss]

I would expect 6.3v on the filament string. Unplug all tubes. What do you measure between the yellow wires now?

Voltage is 5.6v from yellow to yellow. I have grounded the heater center tap (two enameled wires) at the cathode of the 6V6 tubes to raise the potential.  Would that cause the lower voltage readings, or do you think I may need to order a new transformer for the heaters? If I recall someone on this board posted they had to get a separate one for their 1482.

[shooter]

I have grounded the heater center tap

just tape off the CT and use 2 100ohm to create an artificial CT

but while it's down n discharged ohm between each yellow n ct for grins

[jjasilli]

That was me. 
5VAC loaded should work for 6V heaters.  But maybe that's not the point.  5V loaded raises the question that something is wrong.   I'd try disconnecting the heater secondary, and hooking up a good outside supply to test.

[dbishopbliss]

but while it's down n discharged ohm between each yellow n ct for grins

0 ohms between each yellow lead and center tap. 263 ohms between yellow lead and chassis ground, which makes sense because I have the CT going through the 270R cathode resistor... it's probably 263 ohms.

I'd try disconnecting the heater secondary, and hooking up a good outside supply to test.

Unfortunately, I don't have any spare power transformers any more. I moved a couple of months ago and stuff had to go. Might be just as cost effective to buy a filament transformer. I think AES had them for $12. I need to get another speaker jack plate which is $0.85 but I was holding off because shipping was $5.00.

[jjasilli]

Shipping charges are a major bugaboo!  If you have another amp, maybe that'l do to borrow a filament supply? Otherwise, hard to test w/o test equipment, bench supplies, etc.  A variac would work for this and other purposes.

[dbishopbliss]

The filament transformer arrived today. I got a 4A transformer. 2*6V6 (0.9) + 2*12AX7 (0.6) + 1*6AU6 (0.3) + 1*6X4 (0.6) = 2.4A so I figured better to go bigger instead of getting a 2A transformer.


Now I'm trying to figure out where/how to mount this thing. Its a little bigger than I expected. I'm thinking either mounting on top of the chassis next to the power transformer, directly above the filter caps. Or I could try to mount it below(ish) the power transformer by using an extra long screw through one corner of the PT, then drilling a new hole and using a standoff.


Check out the pictures to get an idea of what I mean. I'm open to other suggestions.

[sluckey]

Before you mount it anywhere, hook it up and see if that solves any problems.

Then look hard at the left side of the chassis (as seen in your pics).

[dbishopbliss]

I just installed the Phase Inverter Grid Stopper and much of the flubby-ness has been tightened up. The amp doesn't sound half bad now, but it is not very bright at all, even with the tone knob turned all the way up. EDIT: I just discovered something... the speaker I am using is VERY directional. I had the speaker so that the side was roughly even with my seating position and it was pointing behind me. As soon as I rotated the cabinet 90 degrees so that it pointed AT me the highs came back. Wow! I don't think I have ever experienced THAT directional a guitar speaker.

I'm still getting roughly 5 VAC on the heaters. After listening to the speaker head-on (see edit above), maybe I don't need to worry about the heaters. The challenge with hooking up the new transformer is the leads for the heaters are tightly wound and pretty short. Hopefully I can disconnect them without too much trouble.

Then look hard at the left side of the chassis (as seen in your pics).

Are you suggesting you see something? All other voltages are pretty close to those listed on the schematic.


Here's a question... if only one half of the CT was grounded, would that impact the VAC? I'm pretty sure I was able to get both leads cleaned of enamel and soldered properly, but you never know.

[sluckey]

I'm suggesting possibly mounting the filament PT on the left side of the chassis.

[DummyLoad]

wind another secondary on the existing PT?

if the bobbin window has enough room to push some 20AWG magnet wire through it. does require removal and disassembly of the PT. probably more work than you'd want to take on.

--pete

[dbishopbliss]

Here's a thought... could I just run 5V6 tubes instead of 6V6? Not sure where I can get a matched pair of 5V6 tubes though. Not finding them on Ebay.

[DummyLoad]

Here's a thought... could I just run 5V6 tubes instead of 6V6? Not sure where I can get a matched pair of 5V6 tubes though. Not finding them on Ebay.

https://www.tubesandmore.com/products/5v6gt-tetrode-beam-power  - 33 in stock. they may match for a fee - ask.

what of the preamp tubes? i'd run with the aux. fil. transformer or wind a new 6.3V secondary on the existing PT if there's room.

--pete

[sluckey]

Here's a thought... could I just run 5V6 tubes instead of 6V6?

I would not go down that rabbit hole!

[dbishopbliss]

Message sent to AES about matching.

I did some searching and a number of people have said the low heater voltage isn't as big a problem for the preamp tubes but that the low voltage could cause "cathode poisoning" in the power tubes. I suppose I may be worrying about nothing. Would a transformer change from 6.3 VAC to 5.1 VAC over time? I believe the tubes are all original in this amp (Silvertone labeled). Maybe it has been this way for 55 years and I really shouldn't worry about it.

[dbishopbliss]

I would not go down that rabbit hole!

If I want to mount the filament transformer there aren't a lot of options. It doesn't fit very well on the left side of the chassis as Sluckey suggests... there isn't enough room between the edge of the chassis and the power transformer. One thing I thought of is mounting it to the front panel of the chassis below the Tremolo and Strength pots. The problem from an aesthetic perspective is that I would have to drill a hole into the Sears | Silvertone logo, which would be a shame.


Here's an idea... could I attach it using something like liquid nails? As long as I ran a ground wire from the transformer frame to ground would there be any downside other than if I wanted to remove it someday?

[dbishopbliss]

Here's an idea... could I attach it using something like liquid nails? As long as I ran a ground wire from the transformer frame to ground would there be any downside other than if I wanted to remove it someday?

I'm thinking this may be the way to go... but instead of permanently mounting the transformer to the chassis, I will attach two bolts. Then I can mount the transformer with kep nuts as if they were coming through.  J-B Weld says they have a strength of 2424 PSI. The surface area of a screw head pretty small but as long as I don't torque the screw too much I'm thinking it would have enough strength.


What do you think?

[1blueheron]

JB weld might work if you never plan to move the amp around and think it will never get banged or bumped.  If it does get banged or bumped, that tranny is likely to come loose and go thrashing about until you arrive at your destination.  That thought makes me cringe.

Will it fit within the square indicated below if you relocate that bundle of green caps or replace them with modern caps of smaller size?

[sluckey]

You have the transformer. Just temporarily connect it to the filament circuit. Of course you must disconnect the yellow leads of the original PT. Does this solve your distorted issue? If not, you may not even need the filament transformer. Very easy to try. Worry about physically mounting later.

I kinda like the JB Weld idea. I've had some good results with that stuff. I'd want a quarter size glob to hold a 8-32 nut. I'd want a nut that's about 1/4" thick too. Maybe get a 1/2" long 8-32 standoff and cut in half.

Cathode poisoning is a myth when dealing with receiving tubes. That applies to really high power tubes that operate with very high voltages, such as 50KW radio transmitters.

[dbishopbliss]

Does this solve your distorted issue?

Actually, the amp isn't distorted any more. The Phase Inverter Grid Stopper seemed to have fixed that. Should I just leave well enough alone? I was worrying about the 5VAC on the heaters, but maybe its not really an issue.

[sluckey]

I'd just button it up. It's a budget amp.

BTW, what meter did you measure that 5VAC with? Do you have another meter you could use for a comparison?

[dbishopbliss]

Radio Shack meter from the early 90's. That's my only meter. One of these days I will get a Fluke.

[DummyLoad]

Radio Shack meter from the early 90's. That's my only meter. One of these days I will get a Fluke.

measure that new hammond fil. transformer - what do you get? it may be the meter reading low on that particular range. 


--pete

[dbishopbliss]

New transformer measures 6.6VAC unloaded. It don't think its my meter.

[1blueheron]

Had this idea last night while working on mine.  Why not drill the rivets out of the Silvertone badge and move it to somewhere else on the chassis.  Then you can mount your trans where it was without messing up the plate.  Pretty simple to remove with small drill bit and a punch.

[jjasilli]

New transformer measures 6.6VAC unloaded. It don't think its my meter.

This is not science.  You cannot reason your way to accuracy for test equipment.  An unloaded filament tranny could be expected to read 7.1VAC, which is 0.5V more than your meter reads for the new tranny.


Also, your prior reading of 5.6V + 0.5V = 6.1V which is within spec for a loaded tranny.  We have determined nothing.  The accuracy of your meter is still in doubt. 

[dbishopbliss]

Had this idea last night while working on mine.  Why not drill the rivets out of the Silvertone badge and move it to somewhere else on the chassis.  Then you can mount your trans where it was without messing up the plate.  Pretty simple to remove with small drill bit and a punch.

No rivets on my Silvertone badge... its silkscreened on the front panel.


I took it out to play with a band last night and it was actually pretty quiet (no buzz, hum) but it is VERY dark. Even with the tone control all the way up. Is this typical for this amp? Or could this be the speaker (WGS ET65) or could this be a symptom of the low heater voltage?

[dbishopbliss]

This is not science.  You cannot reason your way to accuracy for test equipment.  An unloaded filament tranny could be expected to read 7.1VAC, which is 0.5V more than your meter reads for the new tranny.

Also, your prior reading of 5.6V + 0.5V = 6.1V which is within spec for a loaded tranny.  We have determined nothing.  The accuracy of your meter is still in doubt.

Someone needs to send me a calibrated Fluke so I can determine if my RS meter works. :-)

[sluckey]

You've already proven that your meter works.

[shooter]

VERY dark

start with a speaker, to me the schematic seems to be "bright"
tack a 470pF from top of vol pot to wiper n see if it moves you in the right direction

[sluckey]

Some good info about dark in this thread...

     http://el34world.com/Forum/index.php?topic=24009.msg258025#msg258025

In fact, there's a lot of info about this amp. Just search this forum for "Silvertone 1482".

[jjasilli]

Upon request, attached is a PDF version of my modded Silvertone 1482 schematic (which appears in ExpressSCH in the thread posted by sluckey).  The purpose of the mods was to brighten the input stages, and "improve" PS filtering for an overall less grungy tone.  It still retains its bargain basement flavor; I consider this project a success. 

[1blueheron]

Had this idea last night while working on mine.  Why not drill the rivets out of the Silvertone badge and move it to somewhere else on the chassis.  Then you can mount your trans where it was without messing up the plate.  Pretty simple to remove with small drill bit and a punch.

No rivets on my Silvertone badge... its silkscreened on the front panel.


I took it out to play with a band last night and it was actually pretty quiet (no buzz, hum) but it is VERY dark. Even with the tone control all the way up. Is this typical for this amp? Or could this be the speaker (WGS ET65) or could this be a symptom of the low heater voltage?

OK got ya,  I thought you were referring to the other badge that you showed in your picture under the trans. 

[sluckey]

Seems to me you could just rotate the filament transformer 90° as seen in the pic 1blueheron just posted. I still suggest you temporarily hook it up to see if you gain anything before drilling any holes.

[dbishopbliss]

I still suggest you temporarily hook it up to see if you gain anything before drilling any holes.

I'm going to do that tonight (hopefully).

Seems to me you could just rotate the filament transformer 90° as seen in the pic 1blueheron just posted.

You can't tell from the picture, but under the transformer 90° are the screws used for mounting the tag strips that hold the filter caps. If I need to mount the transformer, would there be any issue with mounting it as shown in the picture above or is that only a concern with output transformers?

[sluckey]

The transformer will be fine mounted as shown in the pic.

[dbishopbliss]

So I hooked up the the new filament transformer tonight... it measures basically 6.1 VAC loaded. It seems to me that the amp has much more life to it. There are more highs... still fairly dark, but definitely more. Also, the microphone channel has much more gain. So, I think there was something to the low filament voltages and the sound.


In addition, I swapped in new tubes. I think that also helped. I have a new speaker coming so I will try that before I start making more circuit changes.


Now I just gotta figure out how to mount this transformer without accidentally drilling through other wires. Maybe now IS the time to rewire all the heaters.

[jjasilli]

Upon request note the chnages I made to the stock 1482:

1. 56K grid stoppers instead of 68K: Re Ch2 these are technically not grid stoppers; they form a voltage divider with the 1M grid leak resistor.  This provides slightly more signal voltage than the usual 68K.
2. 1.5K cathode resistor on the instrument channel instead of 2.2K.  I wanted to keep Ch1 more true to stock; mod Ch2 to be more Fender-like
3. 104K plate resistor on the instrument channel instead of 330K.  same as 2.
4. 280R cathode resistor on the power tubes instead of 270R.  Don't recall why but nor a significant difference.
5. 72uF cathode bypass cap instead of 25uF.  Experimented with different values while playing guitar.
6.  Bypass caps on cathodes preamp and stage of of PI.  See below

Here are my notes from 2007:  This amp was pretty good stock.  As modded it has great tone and sustains forever.  Tone is more like a Marshal, definitely not Fender.  The sustain however can make the bass muddy when notes overlap, so it’s necessary to pay attention to damping and muting strings.  It needs its tone control turned way up.  It’s very sensitive to the guitar’s volume & tone controls.  Not every setting sounds good.   But when you get it balanced, which is easy enough, it’s wonderful.  Produces vintage rock tone by the gallon. 

Doesn’t like my Strat too much. Les Paul sounds good.  The sustain really shines for slide.  Ibanez Talman with 3 Kent Armstrong lipstick pickups, tuned to open G, sounds dazzling for slide.  (That guitar is gone replaced with a Gibson L6S.)

It was necessary to remove the rear speaker cover.  Otherwise tone was muffled, with too much baffle board distortion.  Before the last set of mods, it sounded fine with the rear cover on!  (The rear baffle board is now back on)

The stock 2-piece Masonite baffle board was replaced with 1/4" plywood and a full 12" cutout.    A thicker baffle would not have allowed clearance for the speaker against the rear cover, which is permanently removed now anyway.  Joints in the paper board cabinet were re-glued.  Rubber or foam a/c insulation was added around the cab near the amp chassis to keep it from rattling against the cab.  Most of the cabinet rattling is now gone.  This is a cheap amp and I’m gonna keep it that way.  Someday I’d like to clone a scratch build with a proper layout and grounding scheme.

Comments on the mods:

Ch 1 grid stop resistor 10K: decouples guitar chord from tube’s grid.  Nicer highs.

Ch 2 input resistors reduced to 56K (carbon comp) from 68K: a bit more gain. Lower values resulted in too much gain.

Cathode Bypass resistors on all preamp stages.  Cleaner highs.  Amp less muddy.  Values from 1uf to 5uf seemed to make little difference from one another.

Ch 2 Vol pot bypass cap.  Doesn’t seem to make much difference.

No treble bypass caps on attenuating resistors which follow the volume pots.  A range of values on a decade box  from 500pf down to 47pf were auditioned.  There was no detectable change in tone!

Power tube cathode resistor bypass cap increased to 50uf from 24uf.  Beefier tone.

Filter choke replaces 2.2K dropping resistor.  Relaxed tone.

27K dropping resistor in the B+ line reduced to 22K.  The actual stock resistor read 34K.  Preamp plate voltages are now closer to spec.  Spec calls for 3 watt.  Stock resistor was ½ watt.  That’s what I used, because it was on hand. 

Diode to last filter stage.  Increases transient response.  Makes bass notes seem more powerful. 

.22uf bypass cap across last filter cap.  Reduces hiss.  Generally improves “tonality”.  A range of values on a decade box were auditioned from  .047 up to .22uf. .22uf sounded best to me. 

[dbishopbliss]

I have a new speaker coming so I will try that before I start making more circuit changes.

I hooked up the WGS G12C today... so much better. Now there is quite a bit more high end than with the ET65.


I decided to measure each resistor and compare to the value identified in the schematic... maybe that was a bad idea. They are all over the place. I'm trying to decide if I should keep the mojo of an amp with drifted 60 year old resistors, or should I try to bring it closer to spec.


There is at least one resistor in the tremolo section that I can get no reading when I measure across it. I don't have a footswitch handy so I haven't even tried it to see if it works.


Any thoughts about replacing the resistors?

[sluckey]

There is at least one resistor in the tremolo section that I can get no reading when I measure across it. I don't have a footswitch handy so I haven't even tried it to see if it works.

Well, I'd replace that resistor. You don't need a footswitch for the trem to work.

[dbishopbliss]

There is at least one resistor in the tremolo section that I can get no reading when I measure across it. I don't have a footswitch handy so I haven't even tried it to see if it works.

Well, I'd replace that resistor. You don't need a footswitch for the trem to work.

I just checked and the resistor I cannot read is a 2.2M going to ground. Could it be that the value is to high for my meter?


And I just remeasured, maybe the resistors aren't as bad as I thought at first. The voltage divider/grid stoppers on the instrument inputs are measuring 87K and 80K (supposed to be 68K) which may be contributing to the low-gain sort of sound in that channel.


The resistors going from the volume pots to the grid of the phase inverter are 289K from mic, 329K from instrument. Not sure if it matters they are 15% different. These seem to change depending upon where the volume pot is set so they were wildly different when I first measured.


There is one resistor that is supposed to measure 330K in the tremolo going from G2 to ground (pin 6) that measures 198K... that is about 30% off.


Other than those, I think they are all close to spec. I was thinking in terms of numbers, not percentages. 

[sluckey]

Your meter should be able to measure a 2.2M resistor. Grab a new resistor and see.

You really can't accurately measure resistors in circuit. You need to disconnect one lead to measure accurately.

That 330K that only reads 198K. It's rare for a carbon comp resistor to decrease in value. They almost always increase in value. That .5µF cap in parallel could be leaky, causing a lower reading. Once again, disconnect one end of the resistor and measure to be sure. My money is on a leaky cap.

[dbishopbliss]

Diode to last filter stage.  Increases transient response.  Makes bass notes seem more powerful.

Can I use any diode here? I think I have some UF4007.

[dbishopbliss]

You need to disconnect one lead to measure accurately.

Whoever wired up this amp was pretty tight with the wire wraps and liberal with the solder. Makes disconnecting one end pretty difficult. Any harm in leaving it.

My money is on a leaky cap.

Stupid question... what exactly is a leaky cap. I guess I always thought it was actually leaking oil or whatever is in them. I'm guessing I'm wrong. And what effect does having a leaky cap have on a circuit.

[sluckey]

Whoever wired up this amp was pretty tight with the wire wraps and liberal with the solder. Makes disconnecting one end pretty difficult. Any harm in leaving it.

Yes, even Sears made good mechanical connections before adding solder back in the day. You need more tools to properly remove a component than was used to install it. I have a drawer full of picks, needles, dedicated desoldering tools, etc. Plus I have an expensive vacuum/desoldering station. It's not easy. Many times its just easier and definitely more cost effective to snip the leads close to the component body then proceed to carefully remove the individual leads to make way to install a new component. Many jack-leg techs would skip this last part and just solder the new component to the old leads. More pride in the $$$ than in their workmanship.

what exactly is a leaky cap.

Not talking about literally spilling it's guts. Most often when someone talks about a leaky cap, they are talking about leaking dc voltage. A perfect cap is just two electrically conductive plates separated by a thin insulator. Heck, I made a cap with a 24x24 piece of window glass, two pieces of aluminum foil, and a bit of duct tape for a science fair project. (see attached pic). Anyhow, that perfect cap with perfect insulation between the two plates is said to have zero leakage. IOW, you could charge the cap to 100V, lay it on your workbench, and the cap would hold that 100V charge forever. But in the real world, the insulation between plates is not a perfect insulator. The insulation will begin to slowly break down with age. (This is very common with old paper/foil caps from the '60s.)The insulation rarely becomes a perfect conductor. It usually just begins to have some high value resistance. This resistance will now allow dc current to pass through the cap. We call this "leaking". You can put a 100V charge on this leaky cap, put it on your workbench, and the 100V charge will slowly 'leak' down to zero. The less resistance the insulator has, the faster the cap will leak. The resistance of the insulation looks just like a resistor in parallel with the cap. It will discharge the cap just as if you placed a resistor across the cap.

The effects of a leaky cap depend on the circuit the cap is used in. In the case of that .5 cap across that 330K resistor, you may notice that that 330K and another 330K resistor right above it form a voltage divider to put a certain voltage on G2. The .5 cap is just used as a filter cap, just like the electrolytics in the power supply. But let's suppose that .5 cap is leaky and it actually has a resistance of 500KΩ. This 500K is parallel with a 330K resistor. Do the math and you see that the total resistance is now 198K. The most noticable effect this will have in this circuit is that the voltage at G2 will be lower than you expect.

And a leaky coupling cap between the plate of one tube and the grid of the next tube will have different symptoms. The plate voltage that leaks thru the cap will appear on the grid of the next tube and upset the biasing for that next tube. If it happens to be a power tube, you may even see red plating. And if there is a volume/tone pot involved the leaky cap will put dc voltage on them and likely make them sound scratchy as they are turned.

If you're still with me at this point, here's the bottom line... Your 330K carbon comp resistor only measured 198K in circuit. Why? A CC resistor usually increases resistance with age, so it's not likely that is the problem, so look to see what is in parallel with the resistor. Ah ha! It's an old .5µF cap, probably paper/foil. That cap is much more likely to be leaky than the CC resistor value decreasing. Hence, my money is on the leaky cap.

I would disconnect the cap and check it for a resistance reading. A good cap will initially 'kick' the meter, then settle down to infinity ohms. A leaky cap will measure some (usually high) resistance. Replace if so. BTW, while the cap is disconnected measure the 330K resistor again. I bet it now reads 330K or higher. Replace if out of tolerance.

And of course, this cap may be fine and the resistor really has decreased in value. But you gotta unsolder some leads to ever know. My money is still on the leaky cap.   

[dbishopbliss]

As always, thanks for the clear explanation. It makes sense now. I will have to play with it this weekend.

[PRR]

> Could it be that the value is to high for my meter?

I can't see your meter from here. Does it have any specs?

A lot of popular-price DMMs only go to 1.99Meg. NOBODY (except tube-amp nuts) uses resistors that large. (Also in implies a test-current of 1uA, which is comparable to leakage in a low-price meter.)