Can someone check my U45B schematic?

[ron vogel]

Hi...long time lurker here!

I did up a little schematic for some mods I'd like to do to my U45b. Just want someone with more knowledge than me to check my work. I want to do a standby switch, and I'm 90% I did that wrong, but I think the input side is OK. I wanted a little more versitality, so I did a gain switch, and a DPDT to swap front-ends.

[imaradiostar]

Hi Ron! I' not super knowledgeable but I have poked around inside a u45b before and I'm opinionated so I guess it'll count for something.

The grid leak bias on the original circuit's first stage is really not great. I would get rid of it altogether, meaning I wouldn't bother with the switching arrangement you've drawn. A useful alternative that would give you more tonal change would be a switch to select different cathode resistor and capacitor values- description to follow.

I'd connect the input directly to ground through a 1 meg grid leak resistor (to ground) and to the first grid. Get rid of the cap...you want the 1 meg resistor to establish the negative voltage on the grid. I'd use a 2.2k to 3.3k cathode resistor with a .47 to 1 uf cap across it. I'd also connect a second similar value resistor and a larger (10-25uf) cap in parallel and wire them to your switch, the cathode and ground. This will allow you to switch between a nice bass roll off like a marshall and a rounder bassier tone like a fender. The u45 generally sounds good cranked up so the marshall style might tighten or improve the sound at high volumes. The fender values should sound great at lower volumes.

What are you trying to achieve with the wire from the output? Negative feedback? It looks like you're using the circuit from an old thread about a U45b but I think you'll still need a 1K5 resistor from the cathode to ground where the 3k3 resistor was previously. You might wanna ask Dave what he did there- I don't quite get it myself. Like I said, not an expert.

Personally I'd wire it differently- I'd go with a 100-150k resistor (for starters) from the output transformer to the cathode of the second triode and a 5k pot wired in parallel with 4k7 resistor from cathode to ground. Then I'd put a cap (between .1 and 1 uf) from the pot's wiper to ground- it'll be a your new presence control. This will make the amp brighter where needed and tighten up the bass response.

There are a few other things I'd do in addition to the above and changing the input stage to cathode bias. I'd add a big grid stopper resistor on the phase inverter to help the distorted tone. This would connect at the tube socket and be around 100k. Merlin (valve wizard) recommends this and I've had great results with it. I'd also rewire the treble rolloff tone control so that it connected after the phase inverter. This way the second triode and the phase inverter can start to add some harmonic content and you can roll off a little high end as needed to smooth things out.

Of course...I haven't yet modded the u45b I've played with. These are only my thoughts.

Don't forget to recap it! I've never seen a univox with all the caps in perfect shape- lots of leaky ones in the univox that's on my bench right now.

jamie

[imaradiostar]

Thought I'd mention- I read further into the thread I linked and it does mention the negative feedback arrangement some more. I still would do it differently but it's all good- go for what sounds good to you!

jamie

[ron vogel]

Bear with me, I'm an overenthusiastic novice!

From what I understand, the neg feedback will slightly compress the signal, and increase the clean headroom a little. I won't always want that, so it might be better if I can switch it in and out.

I changed the front end to your sugesstions, take a peek and see if that is right. I am more of a visual person, so I couldn't really understand the rest of your sugesstions. What would help me immensely is if someone could draw in the changes in red and post it. I have ordered some nice caps for it, and will have most of the ones for the amp sans the ones for mods.

I have done some modding before, but nothing this extensive...so pardon my neediness...just a little nervous!

[imaradiostar]

Wire the 1uf 3.3k combo directly to the cathode (so it's connected all the time) and the second 3.3k/10uf combo to a switch so it connects them in parallel- it's a minor change but it'll have an effect on the bias level of the first triode.

I'm still quite unsure of how the negative feedback is connected in your drawing. It's a good thing to have negative feedback and to have it switchable but once again, I'd start with it wired like a pre-existing amp. We're lucky to exist in a digital world where you can analyze all kinds of circuits without leaving the comfort of your snowed-in house. Take advantage of all the different ways things have been built over the years.

take a look at this schematic:

http://schematicheaven.com/fenderamp...5e4a_schem.pdf

The 56k feedback resistor and the 1.5k cathode resistor form a voltage divider that establishes how much negative feedback there is. The 5k pot and the .1uf cap divert some of the higher frequency feedback to ground so that there is less negative feedback at higher frequencies. By altering the resistor or capacitor values you can change the response of the circuit. Making the cap smaller would increase the frequency that the presence control boosts at. Increasing the value of the 56k resistor decreases the overall amount of negative feedback.

There are formulas for all of the above relationships but for now I'd recommend just subbing in different values and listening to the changes.

Similarly- you could open the feedback loop (disconnect one end of the 56k resistor) and remove the negative feedback. The 5k pot would still effect the gain of the stage before the phase inverter though- you could use this to your advantage depending on what you want. If might be fun to wire a dpdt switch so that it would open the feedback loop and connect a larger value cap in parallel with the .1uf feedback cap- then you'd retain an adjustable treble boost but with a lower frequency.

See this graph courtesy of Merlin the Valve Wizard:



It represents the gain change of a 12ax7 gain stage with a 1.5k cathode resistor and a 100k plate resistor based on the value of the cathode bypass cap- clearly the smaller the cap, the higher the boost frequency. The difference between the boosted and non-boosted signal also changes with the values of the plate and cathode resistors. Once again, formulas exist but I'd start by using my ears and listening to the changes brought about.

I'm really serious when I say this- put a large (100k-220k) half or quarter watt grid stopper resistor on the tube socket pin for the phase inverter grid. It will improve the "cranked" tone of this little amp- you won't regret it.

jamie

[Enzo]

And I don;t see the need for the standby switch at all, but the way you show it ought to work as well as any of the 15 other ways.

[ron vogel]

Yeah, I am going to just do the cap mod on the front end with the switch for now, then re-cap it. I really need to keep it simple, I may mod it more later though. I got some very nice NOS caps for it, and am replacing just about everything except the trannies.

I got the amp working very poorly, and looks like it had been sitting for years unused. The astetic condition is very good for 42 years old, and am doing a full resto on it. The tolex was tear-free, but I went over it with a soft brush and soft scrub, then coated it with vasolene overnight. Looks brand new!

I got the chassis stripped down, then I blasted it with glass bead as well as the corners. I'm going re-nickel plate them Monday. Although, I may copper plate them first, buff them, then nickel plate (that's how it was originally done).

The handle was OK, I could probably re-use it; but I'm going with a Fender 4-bolt black handle.

I'll start a new thread on the resto when I'm done, I've been taking pics as I go.

I have some nice tubes coming in for it, just a waiting game. Got some NOS Tungsram 6BM8's (Tungsrams are coming from Georgia - the country), 1 RCA and 2 JAN Westinghouse 6x4's, and a choice of the original Telefunken, a late '50s RCA or an early '60s Tung-sol 12ax7 to chose from.

I am extremely anal about how my amps look as well as they sound, and am excited to get this done!

[ron vogel]

Well the feedback loop is actually part of the design, and the orginal u45b schematic is wrong for this amp. I ended up making it switchable in and out though. I tried running the resistor/cap thing in parallel with different values, and it just didn't work well. I ended up using the 10/25 cap and 3.3K resistor to ground, and the sound was much better.

The big problem with this amp is usable headroom. It has way too much gain in the preamp stages, so you get a very compressed and splatty distortion. I switched in a 12au7, and the amp stays clean up to about 7-8 on the dial, and you get a bit more of the output tube distortion. It would be better to use the 12ax7, and just cut the gain between stages a notch except for the last gain stage.

With the 12au7, the amp sounds great...and the distortion is very useable, but not oversaturated like the 12ax7. The only drawback is the vibrato doesn't work unless the 12ax7 is in there.

I'm thinking of getting a Weber copper rectifier, and bringing up the voltage to the output tubes...it should really tighten up the bass and possible get a little more gain from the output tubes.

[imaradiostar]

I guess I was suggesting throwing away gain in specific places so it would be a "volume knob amp." It might take a little tinkering but you should be able to get it to clean up with the guitar volume around 5 and crunch nicely at higher settings. If it gets flubby and bad sounding then it probably just needs a little tone sculpting. I recommended the values I knew because they've worked in other amps. I'm slated to have a u45b on my bench in the next week or two so it'll be interesting to compare notes and see how wrong I was...which is pretty likely!

If you like the 12au7 but want to keep your trem try a 12dw7- if the gain goes back up and the trem still doesn't work you can swap the wires for the sections or get the JJ equivilant that already has the sections reversed.

I wanted to mention- I recapped and modded my u255r this past weekend. I didn't use any particularly special caps, just inexpensive Epcos ones that I ordered from Digi-Key and whatever electrolytics I had laying around. It made a big difference! I modded the tone control and a few other things after that and now it's a very usable small gig amp.

Glad to hear you've got it going now. These things are neat little amps!

jamie

[ron vogel]

This should be the finished schematic. I made a few changes that should allow putting a 12ax7 in the first position. I also changed the resistor/cap in the trem...I'm hoping it will reduce a little intensity. The trem sounds pretty good, just a little strong. I'm hoping this will bring down the initial gain some, and break up the output tubes first. Right now the amp sounds pretty good, but I am getting oversaturation and a little oscillating with a higher output guitar of mine. The cleans are really nice...bright, but scouped and not brittle. I think it's going to be pretty good once I make the new changes. I may try a 5751 in place of the 12ax7 to experiment (one tube I don't have laying around):

[ron vogel]

Oh, and the chassis after I got it copper dipped and triple nickel plated (a local place did it pro bono for me!)

[ron vogel]

Sorry for the multiple posts...but the amp after I went through it: