Hoffman Amplifiers Tube Amplifier Forum
Amp Stuff => Tube Amp Building - Tweaks - Repairs => Topic started by: Opry Audio on April 27, 2019, 09:04:58 pm
-
Hello!
I would love to get some input on this schematic for a build I'm beginning to prototype. The power supply may change to get the desired voltages for each stage. I may end up simplifying the tonestack to just volume and tone since this amp basically has the same topology of a brownface deluxe. I may not keep the NFB either, just put it there to try for now. How much DC voltage should I expect after the rectifier? Definitely looking for suggestions regarding the power supply. Does anything look off to you?
-
tube types? or, do we get to guess? :icon_biggrin: 12AX7 + EL34? guessing EL34 based on the bias PS part selection...
--pete
-
You could inject the NFB loop to the grid of the 'slave' triode of the LTP, via a voltage divider...
You may also want to consider reducing the plate load on the LTP 'master' triode to something around 90% of the plate load of the slave triode to get a closer balance.
-
You could inject the NFB loop to the grid of the 'slave' triode of the LTP, via a voltage divider...
You may also want to consider reducing the plate load on the LTP 'master' triode to something around 90% of the plate load of the slave triode to get a closer balance.
agree - those plate loads work better with 10k or less for the tail R with that bias. for the proposed 22K/470R a 91K/100K pair would result in a much closer balance -or- maybe he wants wonky lopsided LTP: it does produce unique distortion.
resistors are cheaper than caps. 220R under the 22K and and 2K-3K NFB R would probably cost less and give similar tone and performance.
--pete
-
tube types? or, do we get to guess? :icon_biggrin: 12AX7 + EL34? guessing EL34 based on the bias PS part selection...
--pete
Oh wow! I totally left out the tube types! 2 12ax7 preamps and push pull 6v6's. Basically coming at it from a AB165 Deluxe without reverb and trem
-
You could inject the NFB loop to the grid of the 'slave' triode of the LTP, via a voltage divider...
You may also want to consider reducing the plate load on the LTP 'master' triode to something around 90% of the plate load of the slave triode to get a closer balance.
maybe 82k would be more closely balanced for the master plate load? I'm trying to keep the controls really simple with just volume, treble, and bass. I might even ditch the tonestack that I have altogether for just volume and tone, like a 6G3, although just from hearing them I would like to find a way to allow a healthy amount of balanced low end to come through, and if I end up keeping the tone stack, use the bass control to attenuate a really gentle slope between 300Hz and 700Hz to bring hopefully sweet and smooth high and mid frequencies to the front of the mix. This amp is being built point to point on terminal strips, so I do have room to do different things, but for now I'm trying to construct a really simple, but great sounding amp. I guess the LTP and Bias supply does complicate things, but the transformer had the windings so I figured I might as well take advantage. I already put a strike on the 5v rectifier tube windings since I was going for a tighter power supply on this one. It's hopefully going into a local studio when it's done so I would love to build something with extremely low noise, which I think my layout on its own will accomodate. Anyway, I would love the end result to have great cleans and a gradual and usable sweep into overdrive with favor going to output tube overdrive, thus my decision to use the LTP, and only 2 preamp gain stages. Going back to LTP, I may even experiment with wiring in a potentiometer to find the best value for balancing on my scope, and then see how much clipping I want coming from the PI. 350v on the PI and 100v on the cathode might be a good place to start, right?
-
TL;DR I'm trying to get the good cronch
-
Basically coming at it from a AB165 Deluxe without reverb and trem
Guessing you mean AB763 Deluxe Reverb? I would just copy the AB763 NFB loop. Disconnect later if you prefer, or maybe even make it switchable.
That 340-0-340 PT will put out 480VDC at the first filter cap when the Standby switch is open, ie no load. The B+ will drop depending on how hot you bias the 6V6s (hot bias equals heavier load and lower B+). The amount of B+ drop is also dependant on the current capacity of your PT, ie, a 100mA PT will have more internal resistance than a 150mA PT so the loaded B+ would be lower. I would expect about 450V with a 150mA PT and 70% bias.
-
Basically coming at it from a AB165 Deluxe without reverb and trem
Guessing you mean AB763 Deluxe Reverb? I would just copy the AB763 NFB loop. Disconnect later if you prefer, or maybe even make it switchable.
That 340-0-340 PT will put out 480VDC at the first filter cap when the Standby switch is open, ie no load. The B+ will drop depending on how hot you bias the 6V6s (hot bias equals heavier load and lower B+). The amount of B+ drop is also dependant on the current capacity of your PT, ie, a 100mA PT will have more internal resistance than a 150mA PT so the loaded B+ would be lower. I would expect about 450V with a 150mA PT and 70% bias.
yes! I did mean AB763, sorry really out of it today. The datasheet for the PT says 660v CT @ 120mA. So the initial secondary voltage on the schematic is probably wrong, I thought it was 680v-- I guess 20v doesn't make a huge difference but still. I'm also thinking about dropping the voltage with a 10W resistor instead of a choke, what are my options for dropping the voltage after the standby switch? obviously a resistor is cheaper and makes layout significantly easier although I can't complain when I'm doing it p2p
-
I already put a strike on the 5v rectifier tube windings since I was going for a tighter power supply on this one.
Based on that statement you really should stick with a choke.
Since you are building this with terminal strips then connect your choke to two lugs that are about 2" apart. If you decide later to replace the choke with a dropping resistor you'll have the necessary space to mount a 10 watt resistor on that strip.
Is this process still in the thinking stage? Any actual work done?
-
I already put a strike on the 5v rectifier tube windings since I was going for a tighter power supply on this one.
Based on that statement you really should stick with a choke.
Since you are building this with terminal strips then connect your choke to two lugs that are about 2" apart. If you decide later to replace the choke with a dropping resistor you'll have the necessary space to mount a 10 watt resistor on that strip.
Is this process still in the thinking stage? Any actual work done?
Well, yes and no. I have my layout mounted into a scratch 17 x 8 x 2 scratch chassis that I have built some clones into before deciding what size chassis, I need for the best layout, and I have more than enough of the parts I need at my repair shop. I'm using a JJ 50/50 can cap and 2 F&T axials for the PI and preamp supply, so most likely I will wire the choke right to the can cap since i'm trying to use as little wiring as possible and only use it in the power supply. Goal is 0 ac wiring except for the heaters and transformers, so with everything mounted in I guess I feel like it's halfway done even though it's really not!
-
I know it's a lot more work to do it that way, but I would definitely rather give someone an amp I got almost right on the first try than go back in and have to change a bunch of stuff I'm not happy with
-
For a proven 1-channel AB763 design with 12AX7s and 6V6s, that delivers heaps of output power with a healthy amount of Fender mid scoop with ample bass and highs and a great tone stack, try a 6V6 dumblesqued blackface design
(I reverse-engineered the attached schematic from one I had, so you'd just ditch the reverb circuit)
-
For a proven 1-channel AB763 design with 12AX7s and 6V6s, that delivers heaps of output power with a healthy amount of Fender mid scoop with ample bass and highs and a great tone stack, try a 6V6 dumblesqued blackface design
(I reverse-engineered the attached schematic from one I had, so you'd just ditch the reverb circuit)
Is the switch after the tonestack an EQ bypass?
-
Is the switch after the tonestack an EQ bypass?
It looks like a Dumble Rock/Jazz switch. In the Rock position the tone stack is similar to a typical Fender with a big mid-scoop. In Jazz mode it converts the tone stack to a James style circuit with an added mid control. The Jazz position has a flatter frequency response and a generally cleaner sound. Not everyone likes the Jazz position, but I've heard some very nice sounds from it.
If you want to know more, Merlin unrolled the two circuits separately in this thread: https://ampgarage.com/forum/viewtopic.php?f=4&t=12175 (https://ampgarage.com/forum/viewtopic.php?f=4&t=12175)
-
...I might even ditch the tonestack that I have altogether for just volume and tone, like a 6G3, although just from hearing them I would like to find a way to allow a healthy amount of balanced low end to come through, and if I end up keeping the tone stack, use the bass control to attenuate a really gentle slope between 300Hz and 700Hz to bring hopefully sweet and smooth high and mid frequencies to the front of the mix...
TL;DR I'm trying to get the good cronch
The thing about the classic AB763 Fender Tone stack is that it causes a LOT of signal loss. That tone stack and the NFB line are 2 of the defining differences between the Blackface era fenders and the earlier tweeds, where the blackface models have the mid scooped tone and much more clean headroom while the tweeds have lots of midrage push, little headroom, and less refined wilder distortion when pushed to overdrive. There are other contributing factors for these qualities as well, but these two things are pretty easy places to make changes to an AB763 circuit that can yield huge results in making it sound different and cronchier :headbang:
Eliminating the NFB line entirely or increasing the resistance there will cause the amp to breakup sooner and reduce the amount of clean headroom. Most old AB763 models used 820 Ohm resister in the NFB line, you could try different values here but I'd suggest starting with a 1.5K NFB resistor (almost 2X's the old fender standard) and increase from there or cut the line out completely if things still aren't wild and woolly enough for you. Lot's of people online like Rob Robinette (https://robrobinette.com/AB763_Modifications.htm#3-Way_Negative_Feedback ) have designed NFB switches to give the option of allowing more NFB through to clean things up and refine the sound, or adding NFB resistance or cutting NFB out entirely to allow the amp to break up sooner and make the distortion wilder. Also the .01uf capacitor you have in the NFB line isn't necessary, I'm thinking that might have been an error when you drew the schematic up?
The AB763 tone stack is trickier. You can play with it all you want but in the end it's still a low impedance passive tone stack that can only cut and subtract parts of the signal to shape tone, it can't boost any frequencies. As a result it's a heavy load down on the audio signal so you loose a good bit of volume and gain going through it no matter what. The benefit of that tone stack is that it's easy to get familiar sounding classic American mid scooped tone with though.
Going with a single tone control will allow for much more mid range to come through and you will also have much less signal loss, but you loose the ability to shape the tone as much, and there will be no way to make a mid scoop if you want it.
You could try a James/Baxandall tone stack instead. Still just 2 controls, treble & bass, but you can work them to make a mid scoop or a mid boost and there's not near as much loss as a fender AB763 tone stack. There are other options too, if you haven't downloaded the Duncan's Tone Stack Calculator then I highly recommend you do ( http://www.duncanamps.com/tsc/ ) . You'll be able to see everything I'm talking about in tone stacks and also find out how other tone stacks will effect and shape your signal.
I've attached a quick hand drawn edit on your schematic with a NFB switch, an old Ampeg style James/Baxadall tone stack so you can see what those might look like. Also a few suggested edits in red as well. Please excuse the chicken scratch, hope this helps you consider some of the options :smiley:
-
Your input is actually super helpful! I guess I didn’t consider a baxandall tonestack, but I think you’re right, it would do well in this circuit and the signal between stages is more streamlined. I have the tonestack calculator! I need to look into baxandall more closely. I see some schematics recommending 500k pots instead of 1M. Just wondering what the benefits would be between the two! Anyway, I think this tonestack will make a huge difference, I didn’t realize it was used in all the old Reverberocket designs. I did a repair on someone’s Reverberocket 12-R-T recently and it sounds really unique, also fantastic. And my ampeg Reverberocket 2 is one of my favorite amps, which has a weirdly magical Pyle speaker of all things.
I will experiment with the NFB. I would like the layout of the control panel to be super simple, so it’s something I may not worry about if I find something that sounds right to me in a general sense. I would love to build and amp that’s versatile in its own way, but has good tone literally everywhere, so we will see how that develops!
I will post an updated schematic and a layout for your roasting pleasure here soon! I still have some layout decisions to make about ac wiring layout since I plan on having an on-standby-off switch on the front of the amp which i want to wire up very cleanly to the IEC and fuse in the back. Maybe I will start a separate thread for that, or if there already is one please link!
And yes the cap in the NFB was a typo, I’m not really sure why I put it there in the first place lol
-
FWIW Since you're considering James/Baxandall TS, here's one of my single channel AB763 builds with a James Tone Stack. Uses 1M pots. The pot resistor and cap values give a classic Fender mid scoop. The Treble and Bass controls give a wide range adjustment. I didn't bother with attenuating the gain after the tone stack recovery stage, and the amp is super loud and very clean into 2 x 12" 100W speakers.
-
does not a GCA need to be CC? that pot with the GCA will upset grid-leak bias. works but sounds like $#~^...
see attached.
--pete
-
does not a GCA need to be CC? that pot with the GCA will upset grid-leak bias. works but sounds like $#~^...
see attached.
--pete
Can you clarify what GCA and CC is? Not familiar with that abbreviation
-
Grounded Cathode Amplifier. Capacitor Coupled.
--Pete
-
Grounded Cathode Amplifier. Capacitor Coupled.
--Pete
Got it, thank you!
-
...The thing about the classic AB763 Fender Tone stack is that it causes a LOT of signal loss...You could try a James/Baxandall tone stack instead. Still just 2 controls, treble & bass, but you can work them to make a mid scoop or a mid boost and there's not near as much loss as a fender AB763 tone stack...
The difference in signal loss between the Fender and the James looks pretty minimal to me?
Using TSC, with the mid at 6k8, and bass (log B) and treb (log A) at 50%, the Fender mid scoop dips to about -23dB, with the average 100Hz-10kHz looks about -18dB.
Whereas the James with bass and treb set to 50% has the mids and 100Hz-10kHz average at about -22dB.
-
Here is a layout with your suggestions in mind! I'm leaving out the negative feedback and impedance switch for now. Please roast away.... so? will it fail?
-
Maybe it will look better with 3 dimensions :think1:
-
does not a GCA need to be CC? that pot with the GCA will upset grid-leak bias. works but sounds like $#~^...
see attached.
--pete
Pete is absolutely right, I missed that in my quick edit. Thank you Pete!
-
...The thing about the classic AB763 Fender Tone stack is that it causes a LOT of signal loss...You could try a James/Baxandall tone stack instead. Still just 2 controls, treble & bass, but you can work them to make a mid scoop or a mid boost and there's not near as much loss as a fender AB763 tone stack...
The difference in signal loss between the Fender and the James looks pretty minimal to me?
Using TSC, with the mid at 6k8, and bass (log B) and treb (log A) at 50%, the Fender mid scoop dips to about -23dB, with the average 100Hz-10kHz looks about -18dB.
Whereas the James with bass and treb set to 50% has the mids and 100Hz-10kHz average at about -22dB.
Right, because its a passive Baxandall Tone Stack as well there is still signal loss. But boost the treble and bass to say 60%-%80 and see the difference between how the two tone stacks react. Not as deep of a mid cut in the Baxandall TS which makes for a lot more guitar signal, even with just a few extra db. Remember bulk of the guitar lives in frequencies from 80hz to maybe 4k or so in upper harmonics, but a lot of the meat in the signal lies in that 200hz - 1k range where the baxandall will be a bit less lossy across the board than the fender TMB. For a practical real world proof I just finished an amp conversion where I went from a baxandall TS to a Fender style with fixed 6.8k mid resistor and can definitely tell you I lost some signal strength in the process. I had plenty of output to give and wanted the more extreme mid scoop for that amp so it worked out fine, but if I had wanted to keep things simpler and also as high gain as possible the Baxandall would have been better. It all depends on the application and what you’re going for. :-)
-
I see some schematics recommending 500k pots instead of 1M. Just wondering what the benefits would be between the two!
Change the treble and volume pots in the tone stack calculator to 500k and see what it does. The ability to edit all the tone stack components in that program and see the response differences is one of its best features. :icon_biggrin:
Maybe it will look better with 3 dimensions :think1:
can you post a schematic again with this wiring diagram as well? It’s helpful to see both for those following along, thanks!
-
... I see some schematics recommending 500k pots instead of 1M. Just wondering what the benefits would be between the two!
1) Using bigger resistances (in tone stacks) reduces signal insertion loss slightly, and vice versa for smaller resistances.
2) But, power rating being equal, using smaller resistances lowers resistor hiss (and vice versa for bigger resistances). As I said, that's 'power rating being equal'. If you use higher power rated resistors (all else being equal), resistor hiss is typically lowered. So if you can find higher resistance pots in higher power ratings (i.e. 2x or 3x higher than the ubiquitous 1/2W), they will hiss less.
3) When designing filters, if you change resistance values, you also need to change cap values by an inversely proportionate amount, in order to preserve the same frequency rolloff/half-boost point. I.e. if you double resistance in the filter, you need to 1/2 the capacitance in the filter (or vice versa) to preserve the same frequency shelving.
-
can you post a schematic again with this wiring diagram as well? It’s helpful to see both for those following along, thanks!
[/quote]
It’s the exact schematic that DummyLoad posted but without NFB!
-
Circuit decisions aside, I would love some input on my grounding scheme. The power supply lead dress is hard to make look neat on layout creator, but I plan on zip thing all leads together on the far end of the chassis and bringing them to their destinations from there. Possibly wiring under where the rectifier is.. the terminal strips I’m using are vertical, so I have a lot of vertical space to work with. Maybe my layout could be nearer by running terminal strips horizontally instead of vertically given the position of the tube sockets
-
The schematic posted by DummyLoad posted has an error.
-
The schematic posted by DummyLoad posted has an error.
please point it out.
--pete
-
As Sluckey said, the error is in the schematic.
-
As Sluckey said, the error is in the schematic.
not any more. :icon_biggrin:
--pete
-
Awe, you're no fun.
-
nope.
--pete
-
Opry Audio :hello:
Any updates on this? How far along are you on the prototype you started?
-
I’ve been really busy with repairs but I have everything mounted in, my layout is finalized and most of what I have left is the preamp and phase inverter! I am however having some trouble soldering a ground bus to the back of these alpha pots I got. Any tips for getting solder to stick to them? I haven’t had a problem with others
-
Grinder and 150w soldering gun.
-
I was using a 200w iron but maybe I need to get out the dremel to sand the pot down,
-
Yeah, it doesn't like to stick to the shiny stuff.
-
Flux is your friend.
-
Any tips for getting solder to stick to them? I haven’t had a problem with others
Sorry for the late reply. In addition to what the others said about flux and or using a bit of sandpaper to rough up the backs of the pots, I've also had luck using some isopropyl alcohol on a bit of rag or sponge to clean it off again before applying solder. Also make sure your hands are clean too so you don't transfer any other dirt or grease while cleaning.
-
I just wanted to give a quick update about where this amp is at! I'm using a 40-18066 Classictone PT, which I intended as sort of a Deluxe style power supply for 6V6's. I had some layout and design regrets and turning this build more towards a single channel bassman amp with a 6V6 output stage and a baxandall tonestack. This PT was definitely intended to be used with a rectifier tube. I'm getting about 470vDC after the full wave rectifier and voltage is not dropping across the resistors on the B+ line as expected. I haven't put in tubes yet just because they would burn up instantly at these voltages. What are my options for dropping? What values and wattages should I check out for a build like this? Has anyone done a bassman build with 6V6's before?
-
You don't have any voltage drop across the dropping resistors because you don't have any tubes plugged in. The tubes draw current and that current flow will cause a voltage drop across the resistors.
No tubes = no current flow = no voltage drop.
-
You don't have any voltage drop across the dropping resistors because you don't have any tubes plugged in. The tubes draw current and that current flow will cause a voltage drop across the resistors.
No tubes = no current flow = no voltage drop.
Makes sense. What kind of voltages could I expect to see after the rectifier with tubes plugged in? any idea? At this point I am very much trying to avoid using a rectifier tube if possible. After doing some research I would love to figure out a way to use the unused 5v winding to supply a 6vDC heater supply to the preamp stages. Worth the trouble?
-
See reply #7
-
FWIW (and I know you're probably way past the stage of considering this Tone Stack now but), here are the plots for the tone stack in the amp in Reply #17.
(There's this cool little app called Tonestack Kit available on the App store for IOS devices, where you can export the charts as png images)
-
> cool little app called Tonestack Kit available on the App store for IOS devices
Apparently also for Android:
https://www.diyaudio.com/forums/software-tools/335891-tonestack-calculator-ios-android.html
However it is now 2 months later and I can't find the app in Android appstore.
-
Thanks for that handy link PRR - useful to know the various origins behind the tonestack names in the program.
-
Hey! Just wanted to update y’all on this amp, it’s built and it works— kind of! Upon start up I heard some crackling coming from the PT, and a bit of smoke. It seems like there might be too big of a load on the PT, it’s getting pretty hot. Here’s a pic of the build. Just want to focus on a good power supply for now. Suggestions? I’m using a 10K 5W to drop voltage for the PI but it might be too much— on second thought it could be that the bias supply isn’t wired correctly too
https://ibb.co/jW5d4mF
-
Hey! Just wanted to update y’all on this amp, it’s built and it works— kind of! Upon start up I heard some crackling coming from the PT, and a bit of smoke. It seems like there might be too big of a load on the PT, it’s getting pretty hot. ...Suggestions? ...
You have a short somewhere on the secondary side of the PT which is pulling too much current through the PT.
This can happen for a number of possible reasons:
- some part of the B+/power rail/filter caps positive nodes/supply resistors is connecting to the chassis somehow, (possibly a long/unclipped lead or a solder run is touching the chassis behind an eyelet board)
- the reservoir cap/one of the filter caps is shorted and the HT is dumping current straight to ground,
- one end of the filament winding is connected to the chassis somehow (e.g. at the lamp holder?)
- Your rectifier is shorted and it has destroyed the reservoir cap causing a B+ short to ground
- etc
Look for the short.
If you are lucky, the PT might have survived already, but not all PTs can handle being run into a direct short.
So before proceeding further, you should disconnect all the PT secondaries and (carefully) AC-test them with insulated test clip leads on your VAC meter (and don't let the loose winding ends come into contact with anything that would electrocute anyone or short the PT)
-
Yeah the transformer is no good. I just checked the secondary and there’s a short on the B+ line. That’s a huge bummer! But lesson learned I guess, dang! My can cap has a short in it