Tube Amps / Music Electronics
|For current discussions, please visit Music Electronics Forum.||New: visit Schematic Hell!
The sunn still shines online!
|12/30/2003 4:16 AM|
||Hum in Silvertone 1482|
My tech chops aren't what they used to be...any help would be appreciated.
This amp is humming loudly when the volumes (mic and instrument share the first 12AX7) are turned up. It's quiet when the volumes are down, which leads me to suspect the circuit before the volume pot. I've tried a new tube and there is no improvement.
Here's the odd part: When I touch the chassis, the hum drops and when I move my other hand near the first preamp socket, the hum just about disappears.
I've tightend all the grounds and the pots.
I've added a three prong power cord and that hasn't helped.
If any of you have any ideas, I'd like to hear them.
|12/31/2003 8:41 PM|
I have a 1472 which is the same amp produced in earlier years but has a lower voltage PT, otherwise supposed to be the same as the 1482.
this amp does not have shorting input jacks so a much greater amount of hum is normal when nothing is plugged into the jacks, compared to a Fender Deluxe which is a similar circuit built from better materials. Plus the mic input on the 1482 does not have any series resistance, which would help to quench some forms of interference, I added a used 33K with short leads soldered to the socket pin, then ran a couple inches of shielded wire to the mic jack.
Then when a guitar is plugged in, the unused jack(s) and wire leading to them still act as better antennas to pick up interference compared to the way a Fender has its input jacks. then there is the unused channel, normally I would plug into an INST jack, then run a short jumper cord from the other INST jack over to the mic jack, that way the guitar signal is present at the grid of each channel's input triode and you can adjust the tone & volume of each channel separately to contribute to a final blend. (try it, you may be instantly satisfied with both tone & noise reduction) But if you are only plugging into a single channel, it would be better if the unused channel had a shorting jack on its grid so at least any noise from upstream of there would be silenced. The two volume controls will still be interactive like on a tweed Deluxe but at least no amplified noise will be available on the unused channel in case you like its volume knob maximized (I do). Plus there is no complete metal enclosure to shield the circuit even when the chassis is installed in the cabinet. you may also have no ground connection to the input jacks other than their mechanical connection to the aluminum chassis, I ran a decent gage copper ground wire betwen the jacks' ground terminals, and then hooked an alligator clip lead to that so I could test various ground points in the amp for lowest noise before selecting the ground point to solder to eventually.
There is also higher gain from the 12AX7's having their 330K & 220K plate resistors (compared to 100K on a Fender) so that will amplify hum & noise more so than necessary, plus on mine all the 330K's (not just the plate resistors) had drifted to much higher ohms over the decades, one was even over 1meg.
To make a valid comparison to a Fender for idle hum & noise, set both amps facing you about 3 feet apart, then using the same guitar to test each amp, adjust the controls for as similar a tone from each amp as you can get, and as high a clean volume as you can get without any significant distortion, while trying to have each amp output be the same loudness to your ears (or a SPL meter). then with a single non-shorting dummy phone plug in the Fender No. 1 input you can compare to the stock silvertone. or alternatively with one shorting dummy phone plug in an INST jack of the Silvertone and another in the MIC jack, you can compare the noise to the stock Fender when nothing is plugged into the fender. To compare each amp with a guitar plugged in, keep the amps' settings the same as above and just remove one shorting jack from the silvertone when you plug in the guitar for testing. Leave all the guitar knobs maximized and compare the background noise from each amp separately then.
Just wanted to add, if the 3 prong cord was a direct replacement for the original, the neutral (white, larger blade opening at the wall socket) AC line would probably be best if it was connected to the terminal having the original cap to ground (so-called death cap) if you did not remove the cap already. On a non-polarized two prong system when the amp was made, people simply reversed the power cord in the wall socket to get the position with lowest hum, namely when the neutral was cap-coupled to the chassis. If you cap-couple the hot (black) AC line to the chassis by mistake, there is supposed to be full-amplitude 60Hz on that line and a good amount of that will come through the cap to the chassis, making chassis ground points hum to a different degree depending on their distance from both the cap as well as the point where your new green earth wire connects to the chassis. this is also not safe, in case the cap shorts you will get full AC line voltage on the chassis, that's probably why it's called a death cap, although this was more of a problem when there is not a safety ground wire.
|1/2/2004 5:26 AM|
Thanks for the reply and the suggestions.
I replaced the three input jacks with shorting jacks and the amp is now plenty quiet with nothing plugged in and is reasonable with the amp wide open.
It has the original tubes and seems to work fine. I don't have much to compare it to, but it doesn't seem particularly loud. This is the only 6V6 amp I have, but it's much quieter than my EL-84 Dr. Z Maz Jr.
I'll try to link the channels to see if that changes anything.
|1/2/2004 1:25 PM|
I have aqquired a 182 recently and appreciate the boatload of knowledge you just imparted.
The biggest thing I see about this amp is that it is very dark. I haven't had a chance to fool with it, but I intend to try some different speaker setups to see if that will help. Would changing out some of those old caps help, do you think? I would really like to brighten it up some.
I have also heard of guys changing the circuit around to resemble the 5E3, what do you think of that?
|1/5/2004 9:22 PM|
remember that the tone controls arranged like this are just treble-cut knobs so you mainly need to just keep them both maximized at all times. My tone pots were only about 750K so I also added a 330K in series with each one to reduce the loss of highs when they are maximized.
Then in case you are not leaving the volume controls both maximized as well, you might benefit from a bright cap of about 100 to 150pf silver-mica across each volume pot, so that the highest highs (those which can make it through the cap) do not get attenuated as you reduce the volume knobs.
Weak output can be categorized as either truly low power, or as lack of gain.
Bad filter caps, especially the 20uf section of the can marked C17C, can result in low power even if the hum is not loud enough to be annoying. These narrow-can electrolytics have not been made in years, but amazingly mine was not worn out in a way that yielded excessive electrical leakage, no it just had lost most of its capacitance over the decades. So I left the cap there and added a 47uf in parallel to C17C, and a 22uf in parallel to C17B, and another 22uf in parallel with C17A. Plus 4.7uf in parallel with C1 so the preamp had a little more filtering. Plus once I got really going, I noticed the way that the plate supply of the 6AU6 is connected directly to the screens of the 6V6's, and there were not any screen stopper resistors either. So I put a 390ohm resistor on the screen pin of each 6V6, then after they join C17B, added a 1.5K resistor between there and R30, plus a new 4.7uf electrolytic to ground where the new 1.5K meets R30. This gives the 6AU6 tremolo oscillator its own little filtered power supply node. Also paralleled a 2.7K with R35 after adding the extra power supply filtration, plus replaced R15 with a 3K and that brings up the preamp & PI voltages to respectable levels.
Another place where a weak old cap results in low power is C9, the 10uf/25V cathode bypass cap for the 6V6's. Plus it seems like a better idea to use a cap rated for more than 25V on the cathode of a 6V6, I replaced the original with 15uf & 50V rating.
One more thing about the power, you do not want a rectifier with a weak section in it. This circuit is really asking a bit much from such a small rectifier as a 6X4 to begin with, it does not need to be brand new but can not be weak. I had about a dozen used ones and a couple NOS which I tested plus listened to the difference. There were only 2 that were unsuitable and it was obvious on the tube tester that one plate was weaker than the other. Other than that the used 6X4's sounded as good as the NOS.
When it comes to gain, the 330K plate resistors on the first 12AX7 do give more gain than if they were 100K, but the 2.2K cathode resistors on these two triodes do not have bypass caps, so that limits the overall gain of both parallel input stages. Here is where I made the first major modifications to get a little more conventional Fender sound, I changed one channel to have a 100K plate and a 1.5K cathode just like a Fender, but I put 4.7uf bypass cap on the 1.5K instead of the fender 25uf value. Then on the other channel I paralleled another 330K with the one that was there to result in about 160 to 170K at the plate of that channel, and left the original 2.2K cathode resistor there, also bypassing that one with a 4.7uf film cap.
So each channel now has a little bit different response, and that gives greater variation when jumping them and blending them, plus when you only play through one at a time at least you have two different channels to choose from.
Then the second and final gain stage where R16, the 2.2K cathode resistor is also not bypassed, I put a 4.7uf there as well. This is the first one that I made switchable as a booster function for either or both channels, regardless of where you plug into.
Also, where the schematic shows R23 & R25 as 330K grid stoppers, mine did not have any resistors here originally. Experience has shown that values as high as 330K can really tame the brightness on some amps, I put in 100K for each grid which is a value that works well for me with the level of preamp distortion I am generating upstream of there.
Anyway, this circuit is already close enough to a 5E3 to where the Fender could be copied exactly and also have the trem feature, but it would still not have the same transformers as a Fender nor would it be built of as good materials and will never sound exactly the same. IMHO this is an excellent opportunity not to be restricted by the legendariness of the 5E3 circuit, especially since the chassis & cabinet are so much different. An amp which looks like or has the same chassis & cab dimensions as a 5E3 would not be generally as desirable without a close copy of the 5E3 circuit inside. The same circuit changes that would make a real 5E3 more suitable for yours or my playing would still not be a very good idea to do to a vintage 5E3 even in poor condition, they should probably just be restored to the original circuit & values. With the Silvertone though, just about anything goes so once you get it to where its equally suitable for you as much as a 5E3, you can keep on tweaking until its truly better. At least regarding the tone, the chassis is never going to be robust and the cabinet is so unroadworthy that you really have to be careful not to let anyone sit on it, not even once
|1/8/2004 7:04 PM|
My thanks, too, for that great primer/faq on ways to optimize the 1482. UPS just brought me mine yesterday, and the tone is amazing, but, like everyone else, there are hum and noise issues, and the volume is too weak. Even dimed, and using both inputs, it is not as loud as my DR on 3, and I was hoping to be able to A/B them; plus, there is no headroom at all. All that turning down the guitar volume does is, well, turn down the volume.
The three-prong cord (already got my first "tickle" :/ ), shorting jacks, better grounding, grid R on the mic channel, plate Rs, cathode bypass, and filter caps are all good places to start. I think this would also benefit from floating the heater voltage over DC, and from metal or carbon films in the preamp. There is also a thread over in General Discussion, I think, about reducing crossover distortion in cathode-biased amps by putting a zener in parallel with the cathode resistor.
I wonder if there would be any benefit to changing the PI tube to a 12dw7. Since the line splitter has no gain anyway, would the higher current capacity of the lo-mu half do a better job of driving the power tubes (esp. with tremelo)? It might even be a drop-in replacement.
|1/12/2004 10:25 PM|
Well, beefing up the filter caps took away most of the hum for me. The final touch was adding a 200ohm 2watt screwdriver pot for hum balance where the stock 68ohm resistors reference the heater leads to ground. I just unsoldered the resistors from their ground terminal and connected one to each outer leg of the pot, then the wiper runs to ground. I think 68ohms is too low anyway, even the 100ohm matched resistors many people commonly use I would rather have matched 330ohmers instead. Less heater noise than most amps now and the wires are not even twisted.
When it comes to crossover distortion, it hasn't cropped up since mine is a 1472 my B+ is 250V and it is Class A push-pull.
Definitely lower the value of all the dropping resistors if you have more capacitance in parallel or replacing the original can cap, a good rule of thumb for me with low-powered amps is to not go below 2K between power supply nodes. You also need to be aware of the voltage drop across each resistor compared to its wattage rating, resistors close to the rectifier may need to be 2watts or higher rated. This will give the 9-pin tubes more voltage to deliver both power & (a little more) headroom, while reducing to an extent the B+ seen by the 6V6's. You will have to reduce the 330K plate resistors on V1 to actually allow much more current to get through these triodes than there is presently, before you will get much lower B+ on the 6V6's. I did use carbon film in parallel with one of the 330K's to bring it down to 100K, as well as a few others. That will bring you a bit closer to class A, I would not expect to change the stock value of the cathode resistor on the 6V6's unless yours has drifted. It's obvious that the composition resistors in my Silvertone are specially designed to be low in quality, not much like the ones Fender was using at the time
A lower impedance driver like the low-mu side of a 12DW7 would probably not make much difference unless your amp is like mine was and the 330K grid stoppers are not there on the 6V6's. Well I put 100K's in that position which is still fairly high impedance and it sure seems good with a 12AX7 as phase splitter. Without the grid stoppers and with a low-mu triode to drive the 6V6's, you may still not get much difference unless you reduced the 330K grid loads at R21 & R22, but then you might be able to go as low as 47K or less for R21 & R22 where a 12AX7 would perform poorly by comparison. As you reduce toward 47K though, the tremolo may become more prominent and you may need to reduce the ohms of the Intensity pot, plus make other tremolo changes as well.
Either the series resistance like the 330K or 100K for R23, or the lower impedance driver arrangement, could help reduce the occurrence of *blocking* distortion which is often mistaken for crossover distortion many times. I did some measurement of the PI output (after removing the 6V6's and variacing down to the same PI voltage) and as long as R19 & R20 are the same nominal ohms, the output of both phases will match and the same level signal will be provided to both 6V6's when values of R19 & R20 are chosen anywhere between 49K & 91K. Keep in mind my 1472 has lower voltages overall so you may have a different range, but these resistances are not the ones which make a big effect from a small change. Also it makes no difference whether the 1Meg R18 is connected from the grid to the cathode directly, or if it has a 1.5K resistor between it and the cathode like seen on a Fender tweed Deluxe.
I have not tried the zener diode(s) to keep the cathode voltage from rising much more than it does at idle. I'll have to listen to it to be sure but it does seem like it might reduce some of the compression effect of the power stage, I wonder if you can hear a worthwhile difference when just approaching distortion or maybe it is more prominent when overdrive is already high.
|Page 1 of 2||Next>||Last Page>>|