Going to change the power ecaps in my 5f6a.

[Mike_J]

I was watching a video that some fellow with two original Bassman amps posted, one a '59 and the other a '60. He was comparing the internals of the amps to the original Fender schematic. He identified a few differences between the schematic and both amps. He mentioned the differences were on the reissue schematic but not the the original 5f6a schematic.


Since Doug has a schematics library I found the reissue schematic and went through it. The first thing that jumped out at me is the reissue has two 100uF @ 350VDC in series for the first B+ stage with 220K resistors across them. The screen supply was fed by two 47uF @ 350VDC caps in series with the 220K resistors across them. It then went through a 4.7K dropping resistor to a 22uF ecap feeding the PI, nothing new there. Followed by a 10K dropping resistor, nothing new, to a pair of 22uF @ 500VDC caps in parallel (11uF) to feed V1 and V2. Remembering jewishjay's post "why did my filter caps go poof" I was somewhat concerned about running two 22uFs @ 500VDC in parallel, as shown in the 5f6a schematic, for the first B+ stage given the current higher line voltages. The reissue amp had a good solution to those concerns plus my ecap board has places for five ecaps and I don't understand why there would be a need for using two 22uF caps in parallel for V1 and V2. Seems like one 8uF a 450VDC would do. I will have a few F+T 22uF @ 500VDC caps left over. Can put one of those in while I am testing the amp and change out to the 8uF after everything else is stable.


Just wanted to pass this along as it made sense to me and thought others making the 5f6a may want to give it some thought.

[sluckey]

to a pair of 22uF @ 500VDC caps in parallel (11uF)

No. Two 22s in parallel is 44.

[Latole]

Caps are not like resistors when in series or parallel. It is exactly the opposite

 

 

[Mike_J]

to a pair of 22uF @ 500VDC caps in parallel (11uF)

No. Two 22s in parallel is 44.

You are right, my brain disappeared again. Probably because I didn't think they would place 44uF where they had 8uF on the 5f6a schematic. They are caps C21 and C24 on the Bassman 59 (Rev E) schematic. The Hoffman board shows 22uF which I thought might be quite a bit.

[Mike_J]

The guy with the two Bassman amps liked the tone of his '60 better than his '59. Two differences between the two amps is the '60 had a PI tail resistor valued at 6.8K and the '59 had a 10K PI tail resistor as per the 5f6a schematic. It looked like it had been replaced so it is possible someone looked at it at one time and changed it to the value on the schematic. Anyway the reissue amps have 6.8K PI tail resistors according to the schematic.


His '59 had a 250KL Bass pot versus the 1MA the '60 had and is on all schematics. He blamed the mistake on the person grabbing what they had on hand and it wasn't a 1MA pot. The '59s presence and middle pots were not of the correct values but they had resistors added to get them in the ballpark of where they needed to be. The presence pot was 25KL and had a 4.7K resistor tied to the tabs on the ends or the pots. This is consistent with the reissue schematic but I am not sure how that is materially different than using a 5KL pot.


Another commonality to both amps were 100K slope resistors and .1 bass tone caps. This is also on the reissue schematic. Something else common to the schematics for the 5f6a and the reissue amp is the use of the 47pF cap across the plates in the PI section. Didn't look to see if either of his amps had this cap but my guess is they did.


One thing on the reissue schematic that I have not seen before is the use of 47 ohm grid-stopper resistors. The 5f6a schematic had no grid-stopper resistors although Fender commonly used 1.5K grid-stopper resistors in later amps. While I am testing my amp I will probably use the 47pF cap across the plates in the PI section. Will start with 1.5K grid-stopper resistors and if all is well lift one end of it to see if I can get away with no grid-stopper resistor and then go up to see what is needed.


Another interesting thing I read about recently, sadly don't remember where. An amp tech kept track of the value of the plate load resistors in the amps he repaired and which amps sounded best with which values. His conclusion was that amps with 110K plate load resistors sounded better than amps with 91K or thereabouts resistors. Kind of makes sense so I ordered enough 110K two watt metal film resistors and one 91K to replace the PI 82K resistor to test the theory. Another thing to try in search of better tone.


I read a book by Gerald Weber many years ago. I seem to recall him saying Fender copied a good sounding '59 Bassman to make their reissue amp. He said it did not sound as good as the original because they didn't use southern yellow pine to build the cabinet as well as other issues. This was corrected when they built their LTD reissue model. I own one of these amps and it is a very nice sounding amp. I read a comment from Leo Fender after he was asked why his '59 Bassman sounded so good. His response was it was probably the dimensions of the cabinet. Everything contributes I guess but I doubt that the size of the cabinet is all there is. I know I went to a hardwood store to buy the southern yellow pine for my cabinet, not a hard wood I know. Thumped a lot of boards to find the one with just the right ring to it. The store owner probably thought I was nuts but did suggest using an African wood bubinga because it had a lot of wood tone. Wish I made a cabinet out of bubinga and the piece of wood with the least ring. Had the tools then to make finger joined cabinets in very little time. 


Another deficiency Weber mentioned was the paper used to make the Jensen P10R speakers. Don't know about that but the reissue Jensen's in the LTD sound good to me. The last thing I recall is the manufacturing method used to make the output transformer. It had to have a certain iron composition in its core which I don't think he ever mentioned, interleaved windings and a paper bobbin were also necessary.


I spoke with Mark at Magnetic Components. He rebuilt their OT to those specs, although I don't know about the iron core, down to the metal used for the bell cover.

[Mike_J]

Here are pictures of the new ecap board. Two 100uF @ 350VDC in series with three watt 220K resistors across them for the B+. Two 47uF @ 350VDC in series with three watt resistors across them for the screen supply. The next cap feeding the PI is still a 22uF @ 500VDC F&T cap. Couldn't use F@Ts for the 350VDC because they wouldn't fit under the cap board. No problem with this arrangement though. Couldn't find anything rated for heat higher than 85C degrees. Bought the caps with the highest ripple current rating. Don't know what difference that makes. Seem to recall reading somewhere it was a good thing. Note: There were some caps they wanted $20.00 each for. They may be the greatest things ever but seemed a little pricey.


I am going to install a 22uF @ 500VDC F&T cap for V1 and V2 under the pots. Does anyone know the best adhesive to attach the cap to the corner of the chassis? I have installed turrets on the main board to secure the cap and will use heat shrink tubing on the leads for safety purposes. Not sure what adhesive to use though.

[Mike_J]

Took more pictures related to ecaps. Decided double-sided foam tape was the way to go for attaching the preamp ecap to the chassis. It will also be supported by turrets on the board as shown.

[Mike_J]

Aligned my wires so they would be in line with where I want them to match up with the tube sockets. Does that make a difference. I don’t know but it gave me something to do for a few months.


You may have noticed my 50K trimmer pot is not on the edge of the board but more towards the middle. Moved it because the way I laid out everything the pot was too close to the grid and screen pins on one of the power tube sockets for my liking . No fault of Doug’s just the way it fell on my layout. Plus I had an extra column to work with on my board that Doug’s doesn’t have.

[Mike_J]

Want to give a thank you to thetragichrro for giving me the calculator to balance my pi and pdf64 for explaining it to me. If I am correct the pi has inverted and non-inverted gain that are 180 degrees out of phase. Basically, by balancing the gains the dB on both sides are equal. Whether they stay equal on down the line who knows. Seems to be a good starting point though.


I know there is a reason Fender showed a tail resistor value of 10K on their schematic and used 6.8K in their reissue amps and at least some of their originals. At least in a limited sample size in their better sounding ones. Anyway it has sent me on another quest.

[Mike_J]

Did some work populating the board today. Some differences from Doug’s board and this one I can think of. First placed the 100 ohm resistors on a vertical terminal strip connected to the same pt bolt other output side grounds are attached. Second as explained earlier moved the bias pot. Third already talked about the possible DC voltage on the heaters. Wouldn’t be populated until I decided I want to go that way. Forth as mentioned earlier spread the board out so wires lined up better. You can see the 27K resistor for the feedback circuit all by its lonesome. Fifth there is no pi tail resistor because the company I ordered it from experienced the same storm in Texas I did. May need to change the 82K and maybe the 100K resistors to get the pi balanced for my experiment. Sixth I changed the bass cap and slope resistor to .1uF and 100K respectively. Seventh you may have noticed there is extra room for longer coupling caps. Would allow me to use some longer caps for a marketing advantage. Members of the forum have attempted to assuage me of this but who knows. One last thing. Heard amps with 110K plate load resistors sound better than amps with 91K plate load resistors do. So I just put 110K resistors in as an experiment. Schematic says components are within 20% so it still meets spec. Also read somewhere when amp components drift the plate load resistors drift up and the coupling caps drift down. Supposedly old amps sound better so I bought some.018uF coupling caps to test that theory out on. Still within the 20%.

[Mike_J]

Happy days are here again. Was waiting for my pair of precision 8K two watt wire wound resistors. To prove I am certifiably nuts, paid $3.03 each for them. By the way the purpose of this exercise is to experiment with different resistor values to balance my pi. Balance my pi means the inverted and non-inverted voltage gain and therefore dBs are the same. Another thing to chase in my never ending quest for a growler.


Received them today along with about 20 each 82K and 100K two watt 5% metal films. Got lucky, one of the 8K resistors measures 7.8K and the other 8.0K on the dot. I am thinking am I fortunate to have two different values because that should give different choices of R1 and R2.


Do some calculations using 99.7 as my R2 value because that is what I already put in the amp. Need an R1 value of 81.5K. So next step is to go through these 82K, five percent resistors to find an 81.5K. So what happens to me?


Start measuring these resistors and they are measuring 82.0K, 82.0K, 82.0K. The one time I don’t want precision resistors and I get them now? Get to the last one and 81.5K. Thankfully I didn’t buy 19 of them.


If anyone cares, final values were RL1 = 81.5K, RL2 = 99.7K, RK = 463r, RT = 8K and RG = 218K. Will it make a difference? Who knows. Will let you know when I know.

[Mike_J]

Just had an unfortunate epiphany. Have 92nF and a 92.5nF caps. Like everything else don’t know what it means. Always have tried to balance the 220K resistors and .1 caps coming out of the pi. Generally would think .5nF would be close enough but with a balanced pi is it? The mystery deepens.

[Mike_J]

Disaster diverted. Found some orange drop 6ps that matched at 91.1nF. They are polyester. Was hoping to use polypropylene but they say no battle plan survives the first shot in war. Kind of where I am here.


Have some blue Ajax molded that match. I don’t mind throwing $10 or more away on an amp for caps but the era isn’t even correct. The orange drops it will be at least for now.

[Mike_J]

Have attached a picture of the oscilloscope I use for identifying the outside foil on caps. Unfortunately these weren’t previously marked so that will be my next task.

[thetragichero]

the rc filter coming off the phase inverter has a corner frequency well below the useable audio range. 92nf = 92.5nf = 91nf = 100nf for this purpose

[Mike_J]

the rc filter coming off the phase inverter has a corner frequency well below the useable audio range. 92nf = 92.5nf = 91nf = 100nf for this purpose

Could you explain that so even a lowly retired CPA could understand it? I very much appreciate your help, especially the calculator that without I couldn’t have the wherewithal to attempt this experiment.


Also, would you repeat the calculator on this post as I fear your other calculator is in timeout with me because of my rants while I was freezing in Texas? Timeout and other topics are the same place.

[Mike_J]

Okay, since I was last with you I have used my little oscilloscope to mark the outside foil on the caps, also populated the board so I can have my heater AC rise on top of DC. Can be taken out of the circuit by removing the resistors from the board. Everything else will be soldered in place. Also installed everything so the pi is balanced.


My war plan should get partial credit since it was made with 1/2 of an extra space for wider caps. So they fit just fine. Didn’t need to use the extra length this time. That effort has not been needed as of yet.

[thetragichero]

anode cap and grid resistor create an rc high pass filter

http://www.muzique.com/schem/filter.htm

long tail pair calculator: https://www.ampbooks.com/mobile/amplifier-calculators/long-tailed-pair/calculator/

there's a number of other good calculators on the amp books page. bookmark em

[Mike_J]

anode cap and grid resistor create an rc high pass filter

http://www.muzique.com/schem/filter.htm

long tail pair calculator: https://www.ampbooks.com/mobile/amplifier-calculators/long-tailed-pair/calculator/

there's a number of other good calculators on the amp books page. bookmark em

Thank you for this advice it is very much appreciated. My Steel String Singer had high pass filter and low pass filter knobs that could be used to get a variety of different sounds. Wish I had taken the time to learn more about my amps while I was building them but I hope to rectify that the second time around.

[Mike_J]

Just did a little exercise. First I learned the human ear is good for about 20 Hz to 20kHz. Using the calculator you just provided to me you have 7.2 Hz and above that can pass with a 220K / .1uF high pass filter. Using the calculator a 150K / .047 high pass filter would pretty much cover what we can hear. Does that mean the difference is wasted energy and if so why would they set it so low. Now I understand why you weren’t terribly concerned with the exact nF value I was using since it doesn’t impact the human hearing range if I understand this correctly.

[Mike_J]

If I understand this correctly then balancing the pi presuming it could be done as simply as I currently view it creates the same dB rating on both sides of the wave. Whereas the RC filter we just discussed determines the frequencies that get to the output and of course we are limited by the frequencies the speakers can reproduce I suppose. Isn’t it a waste of energy to send signal frequencies to a speaker that can’t amplify it? Certainly must be if we can’t hear it. Just some thoughts coming from someone with no EE training.

[thetragichero]

guitar is an even narrower range than human hearing. pull up a few speaker frequency plots: would be terrible for music reproduction but for music production we like that there's little above a few khz because there's a lot of fugly stuff up there (is the same reason that anyone who's played straight into the board through a di with no eq or amp sim can tell you how terrible it sounds... my old church eventually got a sansamp which helped my bass not sound ugly)
so the R part of this section is a design choice based on the output section, don't go bigger than the max grid resistance listed on the datasheet
the C portion is a design decision based on intended tone. some folks like big old coupling caps everywhere (5e3 sounds horrible to my ears but the old orange graphic amps with 68nf coupling caps everywhere sounds decent for a particular stoner/doom tone)

[PRR]

Low-note on guitar is 82Hz.

Low-note on bass guitar is 41Hz.

Take off 10%-20% for drop-tuning.

The "R-C calculator" number is of -3dB, 70.7% response. This is an audible droop. If you want only 1dB (90%) then figure an octave (2:1) extra.

IAC, 7Hz is very generous for ONE stage of a guitar amp.

Multiple stages at similar frequencies, the loss adds-up. Preamp, effects rack, mix console, tape deck, you might easily have a dozen low-cuts. If each one is -3dB @ 7Hz that is -36dB @ 7Hz! Using the octave-up guide, a dozen -1dB at 14Hz is -12dB@14Hz. Probably -6dB@28Hz, and -3dB near 56Hz.

[Mike_J]

so the R part of this section is a design choice based on the output section, don't go bigger than the max grid resistance listed on the datasheet
the C portion is a design decision based on intended tone. some folks like big old coupling caps everywhere (5e3 sounds horrible to my ears but the old orange graphic amps with 68nf coupling caps everywhere sounds decent for a particular stoner/doom tone)

By datasheet are we talking about the datasheet for the speakers I am using?

[thetragichero]

power tube. thought i typed but didn't

[Mike_J]

I had and hopefully still have a book that broke down every stage of the 5f6a circuit. Way over my head then and most likely now. Probably should not finish this amp until I understand how it works. Will help probably when I go through my other builds in an attempt to improve all of them.


I am really happy with my 5E3 build. Didn’t like it until I placed a three way switch so I could change the cathode cap and resistor on either V1a or V1b. Think it was probably the later but it made it a very enjoyable amp to play. Can play it for hours.

[Mike_J]

power tube. thought i typed but didn't
[/quoteMax grid resistance on the datasheets for both the 5881 and 6L6GC fixed bias is 100K. Far less than the 220K Fender used unless I am missing something.

[thetragichero]

they also ran 6v6 well above the max plate voltage 🤷
if it sounds good and doesn't catch fire i guess it be okay?

how's this for bassman circuit analysis? https://www.ampbooks.com/mobile/classic-circuits/fender-bassman-5F6-A/

[Mike_J]

Using a 100K resistor I would need a .22uF cap to get the same 7.2 Hz corner frequency. What effect would the larger cap have on tone? If it would allow more bass why wouldn’t they use those values as it was supposed to be a bass amp.

[Mike_J]

they also ran 6v6 well above the max plate voltage 🤷
if it sounds good and doesn't catch fire i guess it be okay?

how's this for bassman circuit analysis? https://www.ampbooks.com/mobile/classic-circuits/fender-bassman-5F6-A/

That is the book I have or had. The jury is still out.

[thetragichero]

i doubt you'd need to go that low even on a bass guitar. i had a bass preamp/di with the option of a low pass filter at 80hz or 160hz. 80hz *seemed* high (just under twice the frequency of low e) but I'll be damned if it didn't sound better (bass forum guys will tell you to set a variable hpf by ear rather than your eyes). there's a couple things at play here, one being the psychoacoustic phenomenon of "missing fundamental" where our brain "fills in" the low notes based on the distribution of the rest of the harmonics (I'm guessing this is how bass heavy dance music is mixed to sound good even through a little smart phone speaker). the other is that without my bass 'fighting' the kick and maybe floor tom for low end domination, everything is able to be turned up because the mix doesn't get nearly as muddy (i was on stage and not working foh i don't know for sure, but this is what it sounded like to me)
even most speakers designed for bass guitar don't go as low as you'd think, so why bother trying to reproduce frequencies in the power amp (wasting power) that the speaker isn't going to spit out? (to an extent. don't need to go crazy here. 100nf or 47nf will probably be fine here)

leave the grid resistors as they are. datasheets are meant for ideal situations and we don't see many of those. if it's a design that hundreds of thousands have been made (between diy and marshall jtm, etc) you might as well leave enough alone
if you're designing a new build, follow the datasheet

[Mike_J]

Words alone can not describe my despair at this moment. Read the summary section from the 5f6a book you referenced above. It says the 5K presence pot needs to be added to the 10K tail resistor on the board to arrive at the true tail resistor value. Therefore the $3.03 precision 8K resistor no longer is going to balance the pi.


Let’s see what clues I received and disregarded on my way off the cliff on this one. First, the calculator shows 15K as the initial tail resistor value. That didn’t flip a switch. Next PDF64 tells me there will probably be something down stream (probably not exactly what he said but it was how I understood it) of the 10K resistor that would have an impact on the calculation. No still too stubborn to take a little look. If I could change this part of my nature I surely would. Charge!!! Right off the cliff.


Well all may not be lost from this debacle. Looked at the reissue schematic and noticed the presence pot was a 25K with a 4.7K resistor across it. Saw the same setup on an original ‘59 earlier. Attributed it to a woman wanting to get to her coffee break and decided a 25K with a 4.7K resistor across it was close enough for government work and she wanted some coffee.


Wrong again apparently. Go searching for information on this tail resistor mystery and come to someone who says the tail resistor goes to ground through the 4.7K resistor. No explanation why.


Hoping the 4.7K resistor is a more attractive way to ground than the pot with a .1uF cap from its wiper to ccw. Why you may be asking? Well because a 10K pot with a 2.0K resistor across it would again be close enough for government work and a 6.0K cap on the board with a 2.0K resistor across a 10K pot equals 8K and I can still use my $3.03 resistor. Asking for help before I descend off another cliff. Question is do you think the path would go through the 2.0K resistor and not the pot/resistor combination?

[Mike_J]

Wrong again. Would need a 6K resistor . The 8K resistor goes into the bin where it will never be seen from again. Question in the previous post is still valid.

[thetragichero]

that sure was.... a lot of words

it's just an amp, not a precision nuclear accelerator. there are thousands of people playing vintage amps with resistors that have drifted 100% without even knowing it. no need to fret over value differences that fall within the tolerance threshold of 20% components. sure, if you've got the right value, use it. but otherwise use whatcha got, maybe confirming with online calculators that it should still produce sound. even if it sounds horrible no one is likely to die as long as proper safety precautions are taken

i won't waste money on 1% parts besides 1 ohm bias sense resistors

[PRR]

> i doubt you'd need to go that low even on a bass guitar.

Are you talking about my post?

You want to be able to give "full" bass response in case it is needed. As you say, often it is better with less bass. But for several reasons this is better done with a dedicated filter than by going small on essential components.

An additional concern (yes this is complicated) is the recovery time after a LOUD note slams the tubes into cut-off.

[Mike_J]

My plan is disintegrating. Now remember sluckey informing me two resistors in parallel will always be of a lower value than either of the resistors individually. My plan has been defeated by science. A 4.7K across a 25K pot is about 4K. Going to go by Fender values of 6.8K for the tail resistor and a 25KL with a 4.7K. At least the 4.7K resistor is less than the 5.2K pot that is in there now. Have to take what we can get. If it is combined in parallel so much the better because then I am at around 4K if they are added to the tail resistor in parallel.


I very much appreciate the help you are giving me. Pretty sure I will go with Fender values of 220K and .1uF. Won’t be the first one to take that risk.


I have been told by my family and probably everyone else I know that I can get a little dramatic. Don’t believe in using smiley faces because I am pretty sure they will lead to the  ruination of western civilization so you may not know I am kidding. If you presume I am kidding all of the time you will be right most of the time.

[Mike_J]

My plan is to make it like I know Fender did. Positive it will not be like the Fender schematic from the 50s. Will be like the reissue schematic with the exception of minor adjustments like 110K plate load resistors.


Main reason for doing that is I am nearly positive the OT primaries will be reversed and I want to attempt to remove other culprits for the terrible screech it causes so I will know I need to reverse the wires.Then I will bias the amp and check for noise. As I recall the amp was pretty quiet when it was first built. My reissue’59 LTD is dead quiet at idol which I expect from this amp. The one I am building now has maximum one hour of playing time in 22 years. Played it with the LTD and they pretty much sounded the same. Want this one to be a slight bit more aggressive than the other one.


Put the ability to easily put dc under the ac heaters. Can do it and add a humdinger pot in 10 minutes. Eventually will measure the tail resistor to ground and lower the 6.8K resistor to get the total tail resistor to 8K just to see what happens. Easy enough just to change it back if it doesn’t prove to be of benefit. Have other things I want to try as well. Everyone who has provided help to me is greatly appreciated.

Thanks
Mike

[Mike_J]

Got an answer to my question in reply # 31. Starting off we have a 25KL pot that really measures 23.04K with a 4.7K resistor that really measures 4.65K. Placing the 4.65 resistor across the 23.04K pot from cw to ccw and taking a measurement of them in parallel yields a reading of 3.88K. Installed the .1uF cap to see if the guy on the internet was right and the 4.7K resistor is added to the 6.8K (really 6.74K) is right. Measured from both sides of the tail resistor and get 10.62K. Take the 10.62 total tail resistor value and subtract 6.74K from it equals 3.88K. Don’t believe everything you read  on the internet. Now for the big question what value do I need for the first tail resistor to balance the pi? The answer is 10.62K minus 8K less on the first tail resistor or 2.62K. The resistor in there now is 6.74K so 2.62K less is  4.12K. Will have to try it down the road.

[Mike_J]

Think I will do what you people have been telling me to do over and over. Have this scrawny but fairly long piece of G10 with a hole in it so I can install a potentiometer in it. Slap a 10KL in it, put an old black faced era knob on the pot and a couple of wires with alligator clips on both ends. Remove the 6.8K resistor and place alligator clips on both ends of where the 6.8K was and play while my neighbor slowly turns the pot. Record the preferred value both clean and overdrive and hopefully the numbers are the same.

[Mike_J]

Great news. Got an order from Doug today that included three .1uF 630VDC Mojo Dijon caps. I am obsessed with balanced caps in my pi even though it may not make any sense at all. Anyway two of them tested at 91.1nF so they will definitely get some time in my amp. The other one was close but will go in the tone stack where it didn't matter to me anyway.


Glad I made space for larger caps in my build. These caps are not tiny.


Thanks
Mike

[bmccowan]

I am obsessed

I hadn't noticed.

[66Strat]

I am obsessed

I hadn't noticed.

 

[Pwurso]

Hey, where’s the rest of this thread? It was just getting good.
Kidding aside, I was also on a similar journey building my 5F6A last year. Hope yours turned out the way you wanted it. I love mine. It would be great to hear what you liked and didn’t.

[Mike_J]

I do most of my amp building on the back screen porch here in south Florida which can only be done comfortably from about now to around May 15th. Regrettably my memory is not what it used to be so I am thankful for this forum so I can relearn what I was doing in the spring. Will keep all informed when I complete the amp. Have been working on perfecting it for the past 25 years so completing it to my satisfaction is questionable. If I accept that my ‘59 Bassman is going to sound like my LTD then I could finish it today.