Carvin Rebuild Bad! Plexi 50 Build Good!

[silverfox]

I took the Carvin apart last night and after setting the bias to 180 ma for 4 EL34S'the first thing I noticed was- No filament on one of the output tubes. Checked the tube and it's okay, checked the filament voltage and they all read about 5.1 to 5.3 vac.. Then I made some other observations and concluded that this amp was never meant to be anything other than a marketing ploy. It is basically a Belair with different voltages on the circuit. I know, how can the filament voltage be present and the tube is good but no filament glow. Socket? In any event I don't think this amp is worth modding.

So I'm going to build the Plexi 50 using the Carvin trannys and tubes, jacks and misc other parts. I believe also I want to add some of the other mods like the active loop circuit. The Carvin is a 5 tube amp on the preamp and reverb side so it should support the gain tube circuit- I'm thinking of turning this into an equalized tube screamer built into the chassis with foot switch control.

I will definitely require assistance from the forum on this for example- The power transformer on the Carvin is 350vac (each leg) on the secondary with a center tap. I don't know what the secondary voltage is for the as drawn Plexi schematic since it's not listed. Will the Carvin tranny work or do I have to mod the ps to get the right voltage?

I've read there is no tech support for this build and I don't know if that's from the forum or just the catalog.

In the meantime all I've got is a 160 watt stereo mosfet power amp and my Trace tube preamp to practice through.

silverfox

[TIMBO]

Hey silverfox, There is plenty of bits useable for a Plexi build, either a 50 or 100w. I think that both the OT and PT will work well. The plexi uses two preamp tubes and a PI ,so to add an active loop and reverb won't be a problem.

If you are to use Doug's Plexi schem there will be plenty support and help from the forum.

Some pics of the chassis would be a great help to work out a layout. 

Also to use the sch program to draw your schematics makes it easier for the rest of us to follow. 

[jazbo8]

I took the Carvin apart last night and after setting the bias to 180 ma for 4 EL34S'the first thing I noticed was- No filament on one of the output tubes. Checked the tube and it's okay, checked the filament voltage and they all read about 5.1 to 5.3 vac.. Then I made some other observations and concluded that this amp was never meant to be anything other than a marketing ploy. It is basically a Belair with different voltages on the circuit. I know, how can the filament voltage be present and the tube is good but no filament glow. Socket? In any event I don't think this amp is worth modding.

silverfox

I am not familiar with the Carvin, but the schematic says to set the bias to 100mA for the power tubes, why did you set it to 180mA? Also the filament voltages seem low, they should be around 6V. If one of the tubes does not even lit up, then the output is un-balanced, perhaps that's one of the reasons that it does not sound "good" at the moment?

Since the Carvin was designed for 100W, I assume the OPT and PT are rated accordingly, thus if you do go for the Plexi 50, both will be under-powered, so you may not get any B+ sagging or OPT saturation, which also contribute to the unique plexi sound. Just something to think about...

Jaz

[tubenit]

Silverfox,

IF you can be clear on what you're wanting in this Plexi 50w, I can help you with a layout if you would like help.  

Q1: Are you going to parallel V1?  

Q2: How do you feel about using paralleled turret boards like Vox or Hiwatt would use?

Q3: Any other bells and whistles you plan to use?

Q4: You could do this with a relay switchable Hoffman "hot switch" instead of paralleled V1. Would that interest you more?
       OR ............. you could use a mosfet CF and do a "Ripper" style OD section.

This active FX loop does a wonderful job also as a master volume with the recovery pot.

With respect, Tubenit

[tubenit]

And since you have quite a few tubes to work with ....................

Since you're somewhat wanting a Plexi 50w & you also mentioned having overdrive (thru a built in tube screamer),  you could modify this schematic to more closely match a Plexi 50W in the clean & then use a Dumblish OD for the overdrive.

You could do a  D'Mars ODS but with a "normal" Dumblish OD section that doesn't use a 5879.  I actually have a layout already drawn up for this one.

With respect, Tubenit

[silverfox]

I was up late last night pouring over schematics so I've only just started.

Want to address some questions since I'm going to look over everything submitted for a while.

Timbo- Some pics of the chassis would be a great help to work out a layout. I'll try to do that after I figure out how to use a video camera for the shot. Shouldn't be hard. After I posted this thread I saw your build and figured it's philosophically close to what I'm after. Just want it in the Plexi version. Don't know if I can learn to sch soon enough.

Jazbo8- Yes, the output was unbalanced From the beginning. I'd flip the 100 watt setting and no real difference. I made a bad purchase since this is the way it was in the house I tested it in. At the time, on 50 watts that was plenty loud enough to test it out. Something happened between the time I told the person I wanted it and when I came with the money a week later. Don't know what but I suspect this was a bad design from the beginning. I'm probably one of the few people that ever bought the wrong amp in their tone quest. So now, I'm going to do it right. I'll just build it myself.

100 ma bias- In calculating the bias for this amp I multiplied the plate voltage times the bias current and camp up with a severely under powered amp. Hmmm Even so if that was wrong- I went off the suggestions on the Eurotubes website where I bought the Carvin tube set. They state the factory bias doesn't even get the power stage into crossover distortion. 180 ma is only 45 ma per tube.

btw Somebody correct me if my calcs based on bias current * plate voltage is incorrect in this case since I'm new to the Tube amp arena. I suppose an overbias could drag down the filament voltage but don't think that's the case here?

Okay Lets get ready to Rumble

Silverfox,

IF you can be clear on what you're wanting in this Plexi 50w, I can help you with a layout if you would like help. 

Q1: Are you going to parallel V1? 

Q2: How do you feel about using paralleled turret boards like Vox or Hiwatt would use?

Q3: Any other bells and whistles you plan to use?

Q4: You could do this with a relay switchable Hoffman "hot switch" instead of paralleled V1. Would that interest you more?
       OR ............. you could use a mosfet CF and do a "Ripper" style OD section.

This active FX loop does a wonderful job also as a master volume with the recovery pot.

Pretty much on these issues, as long as it results in a neat, great sounding build design I'm not opposed to anything suggested.
I'm going to review the schematics and finalize as much as possible what I hope to obtain.

I don't know what difference paralleling V1 would make. Built in tube based overdrive with tone controls that can be put into the circuit with a foot switch is what I'm after. If the tone control thing is pointless tinkering, then no sense in having it. I have the Ch2 tone control holes already there so I'll just use those. On this subject: In the past I've placed a mixer in between the send and return lines and found that the benefit to the eq prior to the power amp is fantastic. In this case, is there a benefit to placing tone controls into the Active effects loop on the return side instead of the OD? Seems to me I saw a tone control in there somewhere. Is full equalization a good idea or duplicity?

There are some pics of the amp already posted by another builder here. They are in the reply's towards the end in the Carvin Rebuild 1st topic.

Looks like the Plexi 50 with OD and FX with a second set of tone controls somewhere is what I'm after. I definitely tend to play more Hendrix, Nugent, Randy Rhoades, Think 70's Metal Madness with a clean sound turned down and I'll just plug the modeling pedal into the power amp as I've done  in the past to get the Fumble tones when I play that way.


The only remaining question at this point regards the FET. Is that a Tube Screamer type mod? If not, what is the purpose of the FET?

Yes a layout will be great when I make up my mind very soon. I just don't want to waste your valuable time.

I'll look at all this and reply again towards the end of the day.

Well, I'm going a little crazy here trying to get all this done before the Mayan Calendar runs out on the 21st, but it will keep me busy until then.

Really appreciate all this help,

silverfox.

[jazbo8]

100 ma bias- In calculating the bias for this amp I multiplied the plate voltage times the bias current and camp up with a severely under powered amp. Hmmm Even so if that was wrong- I went off the suggestions on the Eurotubes website where I bought the Carvin tube set. They state the factory bias doesn't even get the power stage into crossover distortion. 180 ma is only 45 ma per tube.

silverfox.

I know you are pretty much set on the Plexi 50 build, and with Tubenit and other experts' help, I think you are in good hands.

But I would like to clarify a few things in your post (mainly for my own learning):

"They state the factory bias doesn't even get the power stage into crossover distortion."
I think they meant that properly biased amp should NOT have crossover distortion - may be some of the factory amps have their bias set on the low side, resulting in crossover distortion.

"In calculating the bias for this amp I multiplied the plate voltage times the bias current and camp up with a severely under powered amp."
You are correct in calculating the maximum dissipation for Class A operation, but the MTS is a Class AB push-pull design, so the bias are usually closer to Class B, if you refer to the datasheet for Mullard's EL34 (JJ's "datasheet" is useless for this), you can see that the bias is 30mA per tube, so the 25mA (100mA for the quad) of the factory recommended bias is not too far off (also bear in mind that your plate voltage is at 455V, not 400V as in the Mullard datasheet).



Having said that, the amp may still sound bad - but I don't think it was poorly designed - at least not by looking at the power amp section...

Jaz

[tubenit]

I don't know what difference paralleling V1 would make.

30% increase in gain with NO increase in floor noise.

regards the FET. Is that a Tube Screamer type mod?

I presume you're referring to the mosfet cathode follower???   No, it's not a TS mod at all. It's a cathode follower.  The mosfet is used instead of a 12A_7 triode in the CF position.  Mosfets sound great in the CF.  I don't envision ever using them anyplace else in a tube amp
(except for perhaps the FET design for acoustic guitar in Dumble amps).

Let me know when you decide on what you're wanting & I'll try to help out.  I've got dozens upon dozens of layout in SCH format. I can probably grab one and edit it to what you want.

With respect, Tubenit

[silverfox]

To make a long story short I went back and reread the text. You're right. The amp should be biased around 95 ma. Fortunately I didn't play through it biased that high. However I've been using it at about the 160 ma bias level.

Like I said, the amp has never sounded good to me even after I replaced all the tubes.

In the process I found the source of the error I made. They were talking about the v3.

Something interesting about the V3- First, they block out all the tone networks so you can't see that stuff. The rest of the schematic is there. The power amp is the same design and the bias is set for 100 ma.

Big mistake on my part. What are the consequences of having set the bias so high? What kind of damage. I might be depressed about this situation but now I've got the build bug so it don't matter. Just a lesson learned.

Now, About the V3- www.carvinservice.com/crg/schematics/33101ri_v3_blocked_full.pdf  This is what I want in a tone.

Here is a youtube of the amp Carvin V3M V3MC official demo video - 15min version

Tubenit:

Do you think the Plexi build with OD ad FX has the potential to get somewhere close to this sound? I have a sneaking suspicion it has potential to do that. Another interesting point is the Master Equalizer placed after the Return line and prior to the power amp.

Basically this amp would eliminate my modeler.

Silverfox.

[tubenit]

Do you think the Plexi build with OD ad FX has the potential to get somewhere close to this sound?

Probably not. The Carvin has a singificantly more compressed and perhaps crunchy tone to it.  That amp reportedly is using EL84's.   The Carvins are often using James tone stacks also which have a very different sound than the typical Marshall tone stack.  And please note that the speaker cab looked liked it maybe had a 15" speaker in it when using the clean channel?  Later, it looked maybe like a 12" or a pair of 12" speakers.

I would anticipate the Plexi with OD to be smoother and not have as much "crunchy" gain.

If I am understanding you correctly, your amp does NOT use EL84's, so you should expect a tone different than that demo.

I do think you can have a very touch sensitive amp with sweet harmonics and great sustain.  Perhaps using a distortion pedal or Tube Screamer would add some "crunch" to it.

However, if that is the tone you're looking for ............. maybe the Plexi 50 w/ EL34's approach isn't going to get "that" tone?

With respect, Tubenit

[jazbo8]

To make a long story short I went back and reread the text. You're right. The amp should be biased around 95 ma. Fortunately I didn't play through it biased that high. However I've been using it at about the 160 ma bias level.

Like I said, the amp has never sounded good to me even after I replaced all the tubes.

In the process I found the source of the error I made. They were talking about the v3.

Something interesting about the V3- First, they block out all the tone networks so you can't see that stuff. The rest of the schematic is there. The power amp is the same design and the bias is set for 100 ma.

Big mistake on my part. What are the consequences of having set the bias so high? What kind of damage. I might be depressed about this situation but now I've got the build bug so it don't matter. Just a lesson learned.

Now, About the V3- www.carvinservice.com/crg/schematics/33101ri_v3_blocked_full.pdf  This is what I want in a tone.

Here is a youtube of the amp Carvin V3M V3MC official demo video - 15min version

Tubenit:

Do you think the Plexi build with OD ad FX has the potential to get somewhere close to this sound? I have a sneaking suspicion it has potential to do that. Another interesting point is the Master Equalizer placed after the Return line and prior to the power amp.

Basically this amp would eliminate my modeler.

Silverfox.

No worries, as long as you did not red-plate the output tubes, you are probably ok - all you did was to bias them closer to Class A and in the process increase the idle power consumption - some people even like to run them that way...

"As far as other characteristics of tube guitar amps are concerned, I have found that the pre-clipping frequency equalization and post-clipping EQ are absolutely critical adjustments. Once you have a well-behaved clipper-even if it's just simple diodes, as in the stomp boxes-it is the precise combination of pre- and post-clipping EQ that mostly determines how an amp sounds. The 'secret' of the best sounding guitar amps lies in the pre-clipping EQ response curve." - John Murphy, Carvin Chief Designer

Read the full EETimes article at: http://www.trueaudio.com/at_eetjlm.htm

That's the secret sauce and why the parts were blocked out in the schematic. Anyway, by inserting the equalizer into the effect loop, you are already on the right track. For your new build, I guess once you find the tone that you like with the eq and/or pedals, you can incorporate the general frequency response with a tone stack like Carvin or get a V3 and rip it apart to see what's inside .

Jaz

[silverfox]

Thanks Jazbo8.

Tubnit, for the experience and likely the tone, I'll go with the Hoffman Plexi 50 with OD and FXs. Do you need a pic of the Carvin Chassis? There is lots of room- 23 x 7.25 x 2.25.

As for customization, I'll add a Foot switch operated relay to the OD circuit. Once I get it built I'll experiment with it from there. Is there a parts list or should I build one from the schematic and the existing Plexi 50 parts list?

I've got funds set aside for this and I'm at the starting gate.

Silverfox.

[tubenit]

Yes, seeing a gut shot of the chassis would be very helpful.

Q1)  You're going to use two power tubes for 50w instead of 100w?

Q2)  I personally do NOT use the local negative feedback on the active FX loop.  I am going to recommend leaving that off.  My D'lator
       has a switch to turn that off and it stays off. This would also simplify the layout board wiring.

Q3)  Do you want the single triode and CF overdrive design. This works. I did it on my Double C amp design.   OR ........... would you
        like more of a dumblish OD with two triodes in series for more gain (which is what I'd recommend).

Q4)  Since you have 5 12A_7 preamp tubes .............. let's just go ahead and use a 12A_7 triode for the CF instead of a mosfet to
        simplify the CF design.

Q5)  You have a tone stack in the clean channel  AND a tone stack in the OD channel.  I think a tone stack in the FX is redundant.
        What do you think?

Q6)   How many holes do you have for pots on the front of the amp?  I need to know this to help determine where to locate the send
        & return for the FX.  

Q7)   Have you considered using cathode biased instead of fixed bias to simplify things?

NO, there is NOT a parts list.

Please respond to each question.

EDIT:  I added a SCH version of a paralleled turret layout. It needs editing after you answer some questions, but I thought I'd attach it to look at for now.

With respect, Tubenit

[silverfox]

Thanks for the designs. Here are the answers:

The gut shot is very similar to the chassis you rebuilt for the other customer. Just two trannys at the moment. Same kind of room.

2 power tubes. If I ever play in a club, I'll mic the cabinet if 50 watts don't cut it.

Leave the feedback off.

I like the last design #2 if that is the recommended. If not, I'll go with the recommended.

The tube instead of a mosfet is good for me.

I don't need tone controls in the FX loop.

There are 11 holes for the pots up front, an FX loop in the back, speaker jacks. I'm going to patch in a rotary impedance switch in the back.

Adjustable cathode bias is fine.

Parts list: I at least need to know what is not a 1/4 resistor. I'll look at the Plexi 50 parts list to get an idea of the tolerances. If I have any questions I'll just ask.

The layout you drew up will work. I have plenty of room inside the chassis. The two trannys are separated at either end. I may make up a daughter board for the power tubes.

What size choke? Bandmaster or larger amps as found on the catalog?

Thanks for the designs,

Silverfox.

[tubenit]

Silverfox,

I can't imagine needing more than 50w of power for a small to medium size club.

OK, take a look at this design.  I made a few mods to bring them closer to matching up.  I'd print off both the schematic and layout and highlight them to make sure they match up.  REMEMBER, there are SCH versions attached so you can modify things as you wish.  Any discrepency, you can edit them to match each other.  

I drew the layout using just V1a.  Reason for that is you then can VERY easily parallel V1b with V1a for 30% more preamp gain with no increase in floor/idle noise.  This allows a very quick easy simple mod that you can do at a later time if you want.

I'd probably build this with the clean channel volume pot being 500k instead of the traditional 1M.  Having said that, maybe use the 1M first. If it's not exactly what you want ............ parallel a 1M resistor on the outer pot lugs to make the pot into a 500k.

I would use a 82k to ground for the "return" on the FX.  You don't need the 250k there, IMO. I used a 82k in a built in FX loop in my Tweed BluezMeister.  Works great.

As far as components and values .........................  look at Hoffman's parts list.  However, you will need to make a parts list specifically for this amp.  This should be a fairly lowcost conversion.  

You can build your own Hoffman parallel turret board,  buy one, or use a tagboard which works OK  IF you don't mod much.  

I think Mallory 150's would make decent caps for all the Marshall parts (meaning everything but the OD and FX).  I think you can use Orange Drop PS series on the OD, or use Xicons.  IF you use Orange Drop PS caps on the OD section,  I think doing the cap "polarization" thing is worthwhile but not critical.

Not sure what choke to tell you to use?    I'd use 1/2 watt resistors for preamp, OD, FX and LTPI or even 1 watt resistors on the triode triode plates.

I personally would NOT have the FX switchable in and out.  Having said that a small mini-toggle on the back would easily attend to that if you want it switchable.  You're going to be using the FX return as a Master volume so I think you'd always want it in?

The only real footswitchable relay would be for the clean to OD  (unless you want a midboost or PAB design).  Build it first with DPDT mini-toggle and then you can add the relay later.  Just leave enough room between the layout board and pots to add the relay.  Doug's relays are easy to use and quiet.

I personally prefer .03 caps for the mid pot of the Marshall tone stack instead of .02.

The "R" node on the schematic should actually go to the FX loop where it says "B+".  

I am presuming that you have some amp building experience and can work on them safely?  IF not, this would be an ambitious build for a true novice & I would not recommend it for a true novice.  

Look this over, post any questions and concerns & I'll try to help.  

With respect, Tubenit

[silverfox]

Thanks again, my spouse is very surprised that the board members here are willing to design a $1000 guitar amp for free, but then we don't get the shoe thing either.

As for my prior experience, 56 yo male technoid, radio and electronics books in the late 60's, electronics tech trainee in the late 70's, computer service tech by the 80's. Performed rework and modding of multilayer automated test system boards for Tektronix. Operated computer consulting business- Marooned by the export of jobs to 3rd world countries- working on retraining... again!

Prior projects:  Heathkit Oscilloscope, design, build and program data acquisition system interfaced with 150K lb/hr steam boiler and turbine.

I've worked on some amps at a very basic level. Observe one hand rule, tinkered with television circuits.

How this will turn out, I don't know. Feel confident of eventual success. My soldering is good but it won't surprise me if I assemble the amp and it doesn't work. I'll be sure and let you know how it all turns out and seek help if needed. Aiming for sometime early January to complete the project.

Only other question I have is: next to some of the parts is- "uf". I think you forgot the o ...or is that something important electically speaking?

Can't thank you all enough for all the work you all put into making this forum what it is. I hope someday to contribute- Perhaps this will be my first contribution.

Silverfox.

[sluckey]

uf = µF = micro Farad = unit of capacitance. It's the value of a capacitor.

[tubenit]

Gut shot of the chassis would still be helpful.  I'd like to see exactly where the tubes are located in the chassis.  Layout is important in a high gain amp to avoid oscillation problems.

Sounds like you have reasonable electrical experience & know enough to be safe. So if you're patient .......... AND you're willing to tweak and mod the amp to suit your personal tone values ........... I think you can have a great amp.  There is ALOT of help available here on the forum.

IF there is plenty of room,  I would be inclined to put the OD trim on the front instead of using a trim pot on the layout board. I have it on the front on both of my amps & find that very useful.  In fact, I'd be more inclined to put the presence on the back if needed.

In order to simplify things for you so you can use a "standard" paralleled turret board or tagboard, ......... I redrew the 220k/500p to V4-7 (FX) to be on a terminal strip.  Easy to do.  So with the trim pot on the front & that moved, you can now use a standard layout board.
This also eliminated another row of turrets to shorten the board.

OK,  print these off and see if they match now.  I redid a few things to make this easier for you. The schematic shows a DPDT to switch the FX in and out, but I personally would not use that & did not draw it on the layout for you. You want the FX engaged all the time to have the FX "return" be used as a master volume.

With respect, Tubenit

[silverfox]

First I want to clarify- The switching is for the OD in/out. You don't have to redraw the circuit as I can figure that one out. Here is a link to the shots. I won't need any switching int FX Loop.

http://min.us/mfBOYUmDE3Bwf

Is the standard layout board a catalog item? That would simplify things.



Program Note: In the past my humorous comments have caused various misunderstandings etc... The reference to the uf was humor, not representative of my technical knowledge base.

[tubenit]

OK, the chassis has the same quasi-dilemma as the Marshall JCM900.  Preamp tubes in the middle and NOT much space between the preamp tubes and the pots. 

Having said that ..................   I think you can make it work. You may have to make a turret board that is narrower or even use good terminal strips. I buy Doug's turret board material and cut it with a hacksaw. Cuts easily and quickly for me doing that. Then I drill my holes and install my own turrets. So that is a choice. In "tools" forum there is a thread on installing turrets that I posted that does not involve a drill press.

You can also look for a turret board that is already made that may be narrow enough. Can't tell from the photos how much space there is between preamp tubes and pots?

Switching on the layout is strictly for clean and OD only.

With respect, Tubenit

[guitardude57]

For this chassis, I think mounting the tube sockets on the chassis, and doing the Fowler( ampgarage) style tube cut outs
on the board, may simplify tube pin hook ups to parts and wiring.  

So you would stand off the board a little closer to the chassis, cut holes about a quarter inch larger than the tube sockets... in the board.  Socket pins sit above the face of the board to hook it up.  Never cared for the mid line chassis - tube set up, this solves the problem.

[tubenit]

Wow!   That's quite impressive!  Great idea.   I appreciate your sharing that photo.  I'm gonna save a copy.

With respect, Tubenit

[guitardude57]

I've seen a few amps with this board feature, and Mark's recent Express build incorporated this.
He did have some good pix demonstrating this.

It is a good idea, for those wacky tube layouts, that you wouldn't want to cover/redrill.

[silverfox]

I've hit some rough going. I'm trying to finish the parts list but don't know enough about specifying the voltage level and type for the coupling and bias caps throughout the circuit. The power supply and filter caps are pretty much good as they are I think. I'm trying to stick with a Fat Mid tone Marshall sound. I want to get past the thin treble sound that some of the early JCM 800 builds had.

I have already used the values that are clearly expressed on the parts list for the Plexi 50/100.

If you can suggest cap type and values feel free also to suggest an experimenters lot so I can adjust the tone later.

I am using the last sch. and layout by Tubenit as the design, and building this as a 50 watt version.

The other major problem I have relates to the power transformer and output transformer. I have found voltage levels for test points on the schematic that relate well to the Plexi 50 here> http://www.amparchives.com/album/Marshall/Service%20Manuals/index.html as well as lots of other information. I think I got the site off here in another area. It is wonderful with lots of build pics.

The Carvin power supply here: http://www.el34world.com/charts/Schematics/files/carvin/carvin_mts3200.pdf indicates the voltage levels available from the existing transformers. The amp is designed as a 100 watt.

Do I have to modify the power supply on the build schematic or can I just wire the power and output trannys into the amp as desinged? Will parts smoke etc..

Another question I have relates to bleeding the HT off the supply caps. I'm not sure it will be bled off. If not, should I add a 220K 3 watt bleeder resistor to ground across one of the caps.

As for layouts- I think I'm going to try something like this: http://www.amparchives.com/album/Hiwatt/71%20Hiwatt%20DR103%20100W%20SN%201489/slides/inside2.html

I'm definitely serious about building the amp now and trying to get the parts list together for ordering tomorrow.

I wasn't sure how much assistance I would be able to get but then I saw the favorite amp build pics and know I can get the assistance here.

I will also add- I'm considering reentering the electronics trade as an amp builder in the future. The amount of knowledge I've gained from this board and the project is astounding. Much thanks and appreciation to all the contributors.


Silverfox.

[tubenit]

I typically use 600v or higher coupling caps.  Any of the 600v or higher coupling caps that Doug sells should work fine. I think his coupling caps are 600v or 630v. In a preamp section, you can probably even use a 400v cap.  

You can use any 50uf/500v cap can.  However, I typically use individual filter caps and mount them on paralleled  terminal strips (being sure to avoid the ground terminal) ......... or build a small turret board for the caps.   You can use 47uf/500v caps.  Given your first node is 455v on the Carvin PT, I might use (only on that first node) two 100uf/350v caps in series which will give you a 50uf/700v rating.

Often, I will mount 3 or 4 filter caps across parallel terminal strips, then mount two or 3 more on top of the bottom row. I usually silicon in place and also use a plastic tie to keep them from going anywhere.  The reality is that the wiring and soldering would probably hold that top row in place anyway.

Having said all of that,  I have found that I like a lower value filter cap on V1 preamp tubes. I feel like it genuinely helps with a more articulate and expressive tone.  So, I would be inclined to use a 10uf/500v or 16uf/500v on V1 node.

So I might be inclined to have 50uf/700 , 47uf, 47uf, 20uf, 16uf or 10uf as an example.

Use a 220k/3w across the node A filter cap to bleed voltage as you described.

You got a couple of options regarding the bias supply.  You can use the Carvin design and see if that works OK.  OR ...... use the design that Hoffman specifies which is probably what I would do.   Bias filter caps are lower value voltages.  Look on Hoffman's part's list to see what he recommends.  I typically build cathode biased amps so I'm not real familiar with what those need to be?  Maybe 100v or 150v ?

I like the HiWatt design.  How much space is there between the tubes and the front panel pots?  Will a standard size turret board fit there?   If not, you can "narrow" one of Doug's boards with a hacksaw or use good terminal strips.

With respect, Tubenit

[PRR]

Coupling cap voltages "can" be determined by test. Build it without the coupling caps (also tone-caps, but you do need filter caps, maybe). Use a battery-power voltmeter to probe where each cap will go. That's what it has to stand.

Problem. In most circuits the maximum voltage is the few seconds after turn-on, before the tubes warm-up and pass current and pull-down their plate resistors. Also the cold-no-load voltage of the power supply may be high.

For a very conservative treatment, determine the NO-load voltage of the power supply (you need the first filter-cap to charge-up normally). Pick most of your coupling caps higher than this.

Vacuum-rectifier amps, the HV does not come on until nearly when the other tubes hot-up and start to suck. So while the NO-load voltage may be 460V, in real life you "never" get over 400V before things sag back down. 400V caps are very popular.

Note that fix-bias power tube coupling caps must also stand the negative bias. (And this usually pops-up instantly.) So 400V supply on driver, -45V on power grids, needs a 445V cap.

OTOH, cathode caps "can't" ever see more than a few volts. Even if the tube goes dead-short! The 100K and 1.5K resistors will split say 400V down to 5.9V. A 3V electro could be toast, but mostly we find 6 10 25 50V electros. For the 0.68uFd cap in Marshall tune, you can't hardly find one under 100V because they don't make film that thin.

Most film caps will stand 20% overvoltage and die in a few years. 20% under-voltage will last decades. There's no shame in buying 600V caps to work at 400V max 300V normal. For a few bucks more you probably NEVER have to replace them. Considering that finding, buying, and swapping one cap later is over an hour of time-sink when I could probably find $5 worth of fun instead, a buck or two more is cheap insurance.