Have been fighting noise in a Fender 5B3 clone. Cliff Schecht has been through this with the 5C3 circuit which is almost the same..I have three problems in this amp. First is a static type noise that sounds like a cold joint. This happens without anything plugged in to the amp. Hum and oscillation when both vol. are turned up. I reflowed all of the connections and no help. Changed all of the plate resistors from carbon comp to metal film. no help. pulled board and checked. nothing. disconnected the first input (mike). no change. Disconnected the second input. this is the one with the 50K resistor.(instrument). The static type noise went away. Removed the resistor and now no static type noise..I have read (Paul Ruby maybe) that this resistor will sometimes pick up noise. I am not picking up any radio stations yet. May just leave it that way or install a 10K metal film resistor.
I have changed all of the tubes and swapped all of them around. Have even changed the pre amp tubes from 6SC7 to 6SL7 and back again. Some difference but not much.
As for hum: My original grounding scheme was buss grounding for all and it connected on a lug right at the input jack. Used this type of grounding on another 5B3 I built and it has no hum. Started playing with a star ground scheme and got rid of some of the hum. Most improvement was when I separated the first filter cap from the others and ran it on a ground wire by it's self. Even with the star system I have excess hum. As recommended I am going to go to DC heater power. I intend to use the DC rectifier system that is used in the Marshall JCM DSL 50. Anyone got experience with the DSL 50 circuit?
Anyway, know this is long, thought some might enjoy it and have some sympathy for an old builder.
Fender 5B3 noise
Moderators: pompeiisneaks, Colossal
Re: Fender 5B3 noise
How are your center taps grounded?
Did you sand the surfaces before your tightened the star grounds together to remove the coatings and oxidation.
Have you tested your electrolytics for leakage?
Magnetic field between iron?
Dirty house power? an isolation transformer is helpful for troubleshooting grounding issues.
different tubes or known good tubes? - big difference.
the hum and oscillation can be voltage on the pots, or poor circuit design, you may need to change the values and or taper. Possibly tweak the circuit pre the pots to lower the gain.
hope it helps.
Did you sand the surfaces before your tightened the star grounds together to remove the coatings and oxidation.
Have you tested your electrolytics for leakage?
Magnetic field between iron?
Dirty house power? an isolation transformer is helpful for troubleshooting grounding issues.
different tubes or known good tubes? - big difference.
the hum and oscillation can be voltage on the pots, or poor circuit design, you may need to change the values and or taper. Possibly tweak the circuit pre the pots to lower the gain.
hope it helps.
Last edited by selloutrr on Tue Nov 08, 2011 3:08 pm, edited 1 time in total.
My Daughter Build Stone Henge
-
Cliff Schecht
- Posts: 2629
- Joined: Wed Dec 30, 2009 7:32 am
- Location: Austin
- Contact:
Re: Fender 5B3 noise
From experience I can say that the DSL50 rectifier circuit is prone to failure. The circuit itself isn't necessarily bad but the parts they use are underrated. The rectifier itself is known to fail often. Wouldn't go that route or if you do, beef up the components a bit.
Also my biggest noise change was isolating the input jacks and moving their grounds directly to that stages decoupling cap. I think I moved the whole V1 ground just to that decoupling cap actually, which seemed to help as well. I tried shielding this and that, changing parts, tubes, runs, etc and the V1 rewire (along with the DC heaters) did the trick.
I will pop inside my amp and see what I did for DC heaters. I used a big voltage regulator and put the whole amp on DC because it kept the wiring complexity down (i.e. I didn't have to rewire the hard to solder heater pins).
My amp was PTP instead of turret because of the chassis I used. This caused a lot of problems of its own..
Also my biggest noise change was isolating the input jacks and moving their grounds directly to that stages decoupling cap. I think I moved the whole V1 ground just to that decoupling cap actually, which seemed to help as well. I tried shielding this and that, changing parts, tubes, runs, etc and the V1 rewire (along with the DC heaters) did the trick.
I will pop inside my amp and see what I did for DC heaters. I used a big voltage regulator and put the whole amp on DC because it kept the wiring complexity down (i.e. I didn't have to rewire the hard to solder heater pins).
My amp was PTP instead of turret because of the chassis I used. This caused a lot of problems of its own..
Cliff Schecht - Circuit P.I.
Re: Fender 5B3 noise
Can't really help with the static, sorry. I think you are on the right track and need to keep looking. Set it aside for a week or so. That will help you get a fresh look. Untrain your eye!
For the hum, I agree that ground scheme is a likely contributor. I am assuming you know the difference between 60Hz and 120Hz and we are looking at 120Hz. I'm sure all of us have seen the debates and discussions about what ground methodology works best. I concluded long ago, there is no single correct answer, but, to a large extent, each of us goes by experience. I think I'm in the minority with the scheme I usually use, but it works for me. Since you have gone down some other paths without satisfactory results, I think you don't have much to give up to try yet another variation on how it works.
This is how I approach it.
There are sections to the circuit to be identified. In the preamp, each section ends with the cathode. I'm assuming a standard layout where V1 is the input, V2 is gain, V3 is the PI (or more tubes if you have them, but the PI is last). There is an exception for the typical 18 watt layout where V2 is the PI, but you work it as if it is last in line. So two tubes for gain and one PI make potentially 5 sections. The number of sections is then reduced when a common filter cap is used. For example, if both sections of V1 are fed B+ from a single filter cap, both triodes constitute a section ending at the second cathode. If both V1 and V2 share the same filter cap, then you end up with one preamp section and the PI.
Imagine the ground buss as a linear star. Attach ground leads from low potential to high potential. So, start with the input jack, V1 cathode, the filter cap ground, etc. Group each section together on the buss wire. The PI goes last, closest to the power section. The buss ends there with a lead going to a single chassis point ground bolt that has directly connected high potential grounds.
To the single chassis point ground bolt connect the PT CT, main and screen filters from the power section, OT ground, power cathode ground, and any other high potential power section grounds I failed to mention. Notice that the buss is not grounded by the input jack like many people do.
The only ground I can think of that doesn't get included is the faux CT from the filament string, which can go anywhere but on that single ground bolt. If there is a PT CT for the filaments, it does go on that single ground bolt.
I have found this simple scheme is easy to implement and works well. I have a number of hum free amps using it. I didn't invent it. Someone told me about it. What have you got to lose?
For the hum, I agree that ground scheme is a likely contributor. I am assuming you know the difference between 60Hz and 120Hz and we are looking at 120Hz. I'm sure all of us have seen the debates and discussions about what ground methodology works best. I concluded long ago, there is no single correct answer, but, to a large extent, each of us goes by experience. I think I'm in the minority with the scheme I usually use, but it works for me. Since you have gone down some other paths without satisfactory results, I think you don't have much to give up to try yet another variation on how it works.
This is how I approach it.
There are sections to the circuit to be identified. In the preamp, each section ends with the cathode. I'm assuming a standard layout where V1 is the input, V2 is gain, V3 is the PI (or more tubes if you have them, but the PI is last). There is an exception for the typical 18 watt layout where V2 is the PI, but you work it as if it is last in line. So two tubes for gain and one PI make potentially 5 sections. The number of sections is then reduced when a common filter cap is used. For example, if both sections of V1 are fed B+ from a single filter cap, both triodes constitute a section ending at the second cathode. If both V1 and V2 share the same filter cap, then you end up with one preamp section and the PI.
Imagine the ground buss as a linear star. Attach ground leads from low potential to high potential. So, start with the input jack, V1 cathode, the filter cap ground, etc. Group each section together on the buss wire. The PI goes last, closest to the power section. The buss ends there with a lead going to a single chassis point ground bolt that has directly connected high potential grounds.
To the single chassis point ground bolt connect the PT CT, main and screen filters from the power section, OT ground, power cathode ground, and any other high potential power section grounds I failed to mention. Notice that the buss is not grounded by the input jack like many people do.
The only ground I can think of that doesn't get included is the faux CT from the filament string, which can go anywhere but on that single ground bolt. If there is a PT CT for the filaments, it does go on that single ground bolt.
I have found this simple scheme is easy to implement and works well. I have a number of hum free amps using it. I didn't invent it. Someone told me about it. What have you got to lose?
Re: Fender 5B3 noise
Phil, that is pretty much to way I do a buss with a couple of exceptions. one is that I don't ground the filter caps with their section, I ground the buss on a lug right at the input jacks, I have a separate ground lug for the mains, I normally connect the filament ct or faux ct to the top of the power tubes cathode resistor for dc elevation. Will be pretty easy to convert to yours. I'll give it a try and let you know if it improves this amp.
One question, do you use isolated input jacks?
What you have described is pretty close to what Hoffman over on EL34 recommends.
You are correct that all have their favorite way of grounding and there are lots of ways to do it that work..When I look in some of the older amps I am surprised that they even work. Seems that there are ground points every and any where.
One question, do you use isolated input jacks?
What you have described is pretty close to what Hoffman over on EL34 recommends.
You are correct that all have their favorite way of grounding and there are lots of ways to do it that work..When I look in some of the older amps I am surprised that they even work. Seems that there are ground points every and any where.
-
Cliff Schecht
- Posts: 2629
- Joined: Wed Dec 30, 2009 7:32 am
- Location: Austin
- Contact:
Re: Fender 5B3 noise
Hrmm... Just fired on a 5B3 clone I finished last night and it's one of the lowest noise amps I've built. Can't even tell it's on with no input. I used 6SL7's in the preamp and didn't have enough capacity on the transformer to fit a tube rectifier (it's rated at 5.5V at 1.1A, how annoying) so I went with a SS bridge. This was kind of a goofball transformer that I've used in various builds in the past (250V@200mA no CT HV secondary, 6.3V at 1.8A CT and that pesky 5.5V winding) from an old Japanese reel-to-reel and I think it's finally found its resting place.
The amp sounds KILLER to boot, lots of distortion on tap. I like it more than the 5C3 I think because the lack of negative feedback really opens things up. I was thinking of adding a switch to swap in a feedback resistor ala 5C3 (weighting the value of the resistor appropriately for a 5B3 PI) but I might not bother, the amp is too much fun as is.
I did not use any DC heaters on this build yet it's much quieter than the previous amp that was in this chassis (the PTP 5C3 I built ages ago). It also doesn't have that "swirling" that was present in my last build (the fix is a Zener diode clamping the input grids of the power stage). I think the low noise is from careful lead dress in the preamp area and the lack of swirling is from getting the B+ voltages close to spec (actually they seem spot on with what I expect from an early 6V6 tweed, about 340-350V B+).
I think after building enough Fenders to make most normal humans happy I have figured out a "formula" for getting these things to come out consistently low noise. It's essentially a method of grounding and building that takes into consideration all of the high current return grounds and sensitive nodes. It's not rocket surgery and obviously it's not all completely original fodder but nonetheless I've taken a lot of time to sort through what many consider "proper" grounding (along with what I know through solid state PCB and IC design) to come up with something that follows my ideologies and works well in practice. I've never really used the Fender brass grounding plate though, I've always run buss bars for grounding and played around with where everything is grounded (IMO a better looking technique). I'm going to make a bigger post about this and general building safety and such once I organize my thoughts a bit more which will hopefully help others out in the future. Might be a good reference for people chasing down hum/buzz in their amps as well.
Also FWIW the 6SL7's I'm using are "special" ones, one is a factory selected low noise variant (has the white dot specifying a factory select) and the other is just a solid performer that I bought or pulled from something, somewhere. I've used these same tubes in other octal amps with great success. Specifically I point this out because others have had noise problems in the past with octals and don't know how inconsistent these tubes really are from lot to lot. I'm not sure how the new manufacture stuff works but even in the heyday of tube manufacturing they couldn't get these high gain octal triodes to consistently come out low noise and low microphonics. You gotta buy a bunch and sort through them, or pay more for the premium versions (last time I bought them I was paying $2-3 for good 6SL7's, there are still deals to be had!).
I'll also try to start a thread with pictures of all of my Fender builds, once I get headshells built for everything. There's a lot of great, unique sounds available from the various Fender amps available in the past and it's fun building them and playing them to see how Fender was evolving his designs. Besides, imitation is the sincerest form of flattery (*looks at Dumble CLONERS*).
The amp sounds KILLER to boot, lots of distortion on tap. I like it more than the 5C3 I think because the lack of negative feedback really opens things up. I was thinking of adding a switch to swap in a feedback resistor ala 5C3 (weighting the value of the resistor appropriately for a 5B3 PI) but I might not bother, the amp is too much fun as is.
I did not use any DC heaters on this build yet it's much quieter than the previous amp that was in this chassis (the PTP 5C3 I built ages ago). It also doesn't have that "swirling" that was present in my last build (the fix is a Zener diode clamping the input grids of the power stage). I think the low noise is from careful lead dress in the preamp area and the lack of swirling is from getting the B+ voltages close to spec (actually they seem spot on with what I expect from an early 6V6 tweed, about 340-350V B+).
I think after building enough Fenders to make most normal humans happy I have figured out a "formula" for getting these things to come out consistently low noise. It's essentially a method of grounding and building that takes into consideration all of the high current return grounds and sensitive nodes. It's not rocket surgery and obviously it's not all completely original fodder but nonetheless I've taken a lot of time to sort through what many consider "proper" grounding (along with what I know through solid state PCB and IC design) to come up with something that follows my ideologies and works well in practice. I've never really used the Fender brass grounding plate though, I've always run buss bars for grounding and played around with where everything is grounded (IMO a better looking technique). I'm going to make a bigger post about this and general building safety and such once I organize my thoughts a bit more which will hopefully help others out in the future. Might be a good reference for people chasing down hum/buzz in their amps as well.
Also FWIW the 6SL7's I'm using are "special" ones, one is a factory selected low noise variant (has the white dot specifying a factory select) and the other is just a solid performer that I bought or pulled from something, somewhere. I've used these same tubes in other octal amps with great success. Specifically I point this out because others have had noise problems in the past with octals and don't know how inconsistent these tubes really are from lot to lot. I'm not sure how the new manufacture stuff works but even in the heyday of tube manufacturing they couldn't get these high gain octal triodes to consistently come out low noise and low microphonics. You gotta buy a bunch and sort through them, or pay more for the premium versions (last time I bought them I was paying $2-3 for good 6SL7's, there are still deals to be had!).
I'll also try to start a thread with pictures of all of my Fender builds, once I get headshells built for everything. There's a lot of great, unique sounds available from the various Fender amps available in the past and it's fun building them and playing them to see how Fender was evolving his designs. Besides, imitation is the sincerest form of flattery (*looks at Dumble CLONERS*).
Cliff Schecht - Circuit P.I.
Re: Fender 5B3 noise
Hi Cliff,
your experiences with 6SL7 and lead dress would be very much appreciated. I have an empty 60's Ampeg Reverberocket here that I calls for them and I would like to get it right.
Last year I went a step too far and tried the 6SL7 for the 1st time in an idea of my own. That was one unknown too many. What ever I did I got a roughly 100Hz buzz noise. If I listen past the buzz the 6SL7 is a great tube. I tried over 40 of them so far and the buzz was always there, no matter how many re-builds I did - so I think I am the issue, not the tube type!
I should build the 5B3 to get to know the tube in a proven circuit. Is the 6SL7 somehow more sensitive with regard to proximity / distance to other tubes or components than the novals?
Best, tony
your experiences with 6SL7 and lead dress would be very much appreciated. I have an empty 60's Ampeg Reverberocket here that I calls for them and I would like to get it right.
Last year I went a step too far and tried the 6SL7 for the 1st time in an idea of my own. That was one unknown too many. What ever I did I got a roughly 100Hz buzz noise. If I listen past the buzz the 6SL7 is a great tube. I tried over 40 of them so far and the buzz was always there, no matter how many re-builds I did - so I think I am the issue, not the tube type!
I should build the 5B3 to get to know the tube in a proven circuit. Is the 6SL7 somehow more sensitive with regard to proximity / distance to other tubes or components than the novals?
Best, tony
-
Cliff Schecht
- Posts: 2629
- Joined: Wed Dec 30, 2009 7:32 am
- Location: Austin
- Contact:
Re: Fender 5B3 noise
The proximity of any of the wires in the preamp to the heater wires is one biggie. Unlike 9-pin sockets where I tend to let the wires be a bit long, I made all of the preamp connections as short and deliberate as possible. Heater wires are really short and tucked sort of in the corner of the chassis (especially around V1) which is about as far from the rest of the wires as I can get them.
I also made sure to put the V1 tube as far away from the PT as possible. For whatever reason this amp had the power transformer right in front of the V1 and V2 tubes instead of the 6V6's when I found it. I built my 5C3 this way but always hated how close these two were. When I rebuilt this amp I literally flipped the chassis so that the back-side of the 5C3 build is now the front side of the 5B3. Reusing the original faceplate made this impossible to detect
.
Another big deal is grounding the pot bodies to the same point as the preamp ground. If you build using a buss bar soldered to the back of the pots or use a Fender style brass plate then you already have this advantage but if your volume pot bodies have to go through a bunch of chassis before they see the circuit ground then you lose the shielding benefits that this can offer.
FWIW I didn't realize the importance of this until I shielded the body of the reverb pot on my Vibroverb build (as well as my 6G15 tube reverb head build). I literally added one of those little lugs that Trainwreck builders use to add a buss bar with stainless steel pots and soldered the floating end of the tab to the ground lug on the pot. I also ran a ground wire directly from the pot to the RCA jacks (this ties more into general grounding rules maybe). This trick has worked more than once to clean up that last little bit of noise in amps in the past, but only when I'm not relying on the chassis for my main grounding points (which I don't like to do).
Also make sure your hardware is cranked down tight! I keep mentioning this because I went through more than one build where something or another wasn't tight. Securely fastening everything killed the hum. It may not seem like a biggie but loose mechanical connections, especially anything made of steel, can couple noise into your circuit. Now I make sure everything is tight to the point of almost stripping. I've heard of others using nail polish or Lock-Tite to secure hardware down but I've never found this necessary. If your hand doesn't hurt after you secure everything down then you aren't doing it hard enough
.
I also made sure to put the V1 tube as far away from the PT as possible. For whatever reason this amp had the power transformer right in front of the V1 and V2 tubes instead of the 6V6's when I found it. I built my 5C3 this way but always hated how close these two were. When I rebuilt this amp I literally flipped the chassis so that the back-side of the 5C3 build is now the front side of the 5B3. Reusing the original faceplate made this impossible to detect
.Another big deal is grounding the pot bodies to the same point as the preamp ground. If you build using a buss bar soldered to the back of the pots or use a Fender style brass plate then you already have this advantage but if your volume pot bodies have to go through a bunch of chassis before they see the circuit ground then you lose the shielding benefits that this can offer.
FWIW I didn't realize the importance of this until I shielded the body of the reverb pot on my Vibroverb build (as well as my 6G15 tube reverb head build). I literally added one of those little lugs that Trainwreck builders use to add a buss bar with stainless steel pots and soldered the floating end of the tab to the ground lug on the pot. I also ran a ground wire directly from the pot to the RCA jacks (this ties more into general grounding rules maybe). This trick has worked more than once to clean up that last little bit of noise in amps in the past, but only when I'm not relying on the chassis for my main grounding points (which I don't like to do).
Also make sure your hardware is cranked down tight! I keep mentioning this because I went through more than one build where something or another wasn't tight. Securely fastening everything killed the hum. It may not seem like a biggie but loose mechanical connections, especially anything made of steel, can couple noise into your circuit. Now I make sure everything is tight to the point of almost stripping. I've heard of others using nail polish or Lock-Tite to secure hardware down but I've never found this necessary. If your hand doesn't hurt after you secure everything down then you aren't doing it hard enough
.Cliff Schecht - Circuit P.I.