• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

The Fever Bile eBay 6J1 preamp everyone loves (Smart people pls help!)

Im trying to get a little 6j1 tube preamp (the fever/bile one, lol) i built from an ebay kit to sound right through an (I think) 22k ohm input impedance (generic class D, tpa3255). The output impedance of the tube preamp is 5k ohm =/- so far as i can determine with research. It has rising output impedance toward the low end of the frequency spectrum but quite low near the high end. And as far as im aware, the rule of thumb goes, it wants to see no less then 50k input.

My obviously wrong logic (as well as several youtube videos and forum guides) would dictate i should be experiencing very rolled of low-end running it into an impedance that's too low. And i can fix that by changing the included coupling capacitors from 1uf 50v ceramic to 1-2uf 50v film caps.

Well when i plug the preamp into the power amp, the opposite is my problem, all i get is low end. Id guess its rolling the high end off no higher then 5khz. Went ahead and did the coupling capacitor mod anyway, and as the problem would dictate, it made the issue got worse. At that point i assumed there was some components functioning improperly so i ordered the kit again, exact same issue.

For the hell of it i plugged the preamp out of my little kit tube headphone amplifier into the power amp. Exact same issue, but plugging in the headphone output sounds great. I do not have this issue with my DAC, my tube phono preamp, or any other line level source of audio i have.

To add to the confusion, they all sound just fine plugged into the 3.5mm aux input on the back of the phono preamp (Fosi x4), but fosi is not being cooperative on telling me what the input impedance of the preamp when used as a buffer only so i can actually understand anything i'm observing, i don't have enough info on my own. very much a beginner, and for whatever reason google has been all but neutered. Obviously it's going to be very high, it's a buffer. But no clue the actual value.

What am i missing? I'm definitely missing something obvious here, i've been searching off and on all day and cant figure out how to fix my issue. Should i drop the value of the coupling capacitors and see how that sounds? Im just learning as far as this circuitry goes and i dont want to damage anything.

Schematic of the amp in picture. And i am aware of the output resistor to ground error in it as well. I was able to remove the hum with a modification to the filter cap circuits, and into my fosi buffer and the yamaha s801 in the den, it sound phenomenal. So i'm pretty motivated to get the thing right for my bedroom/desktop power amp. The fosi x4 handles tubey imaging duties in the den just fine already.

download.png
 
Last edited by a moderator:
I apologize, i cant either, i must've downloaded a thumbnail by mistake. And let me pre-thank you for any help whether it pays off or not and any responses to my likely incredibly ignorant questions. It's massively appreciated. And i honestly really, REALLY like the way this preamp sounds into a 47k load on my yamaha with an extra filter cap bridging r9 to ground for noise. I really want to get it working on my desktop amplifier in the bedroom. The quality of sound when its properly loaded is truly phenomenal for what it is.

First pic is preamp schematic, i do believe you've input the correct values. Second pic is what i plan on doing with it currently, im under the impression doing so will have this preamp adequately driving loads as low as 4-5k as well as cut it's bandwidth from roughly 5-500k to roughly 20-100k.

Third pic is the analog input of the power amplifier as schematic. The manufacturer says any preamp will see 20k as the load, but my results seem more on par with a 1k load. Which is obviously abysmally low. I'm not super experienced here, yet anyway, but i would think 20k would only be the load if referenced to ground with a multimeter and the amp powered off? (total guess)

And its a dual tpa3255 chip amp, one chip is bridged to provide subwoofer output, and one chip is stereo with a 100z high pass.
Would the fact it's two chips running in parallel maybe cut the 20k assumed load in half? (i have o idea how ill-informed this observation may be)

The modded preamp schematic design was made by TheStuffMade on youtube for credit where it's due.
The lowering as frequency rises curve on your first graph is a pretty damn good representation of the response curve im getting plugged into the class D amp honestly.
 

Attachments

  • download (1)fever bile.png
    download (1)fever bile.png
    54.3 KB · Views: 177
  • Modded preamp schematic.PNG
    Modded preamp schematic.PNG
    191.5 KB · Views: 272
  • circuit diagram.jpg
    circuit diagram.jpg
    53.8 KB · Views: 172
Last edited:
I did actually have that thought too, based on my knowledge of wiring up guitar tone pots
Would removing or changing the value of those capacitors cause adverse reactions in the amplifier? My dac, my interface, as well as my fosi preamp have no issues driving this analog input at all, there is no high frequency loss detectable by ear at least. Same gain and eq settings sound basically the same on the 47k input yamaha, and this class d amp. MAYBE 1 or 2 db lower output, if that.

i believe the lowest i can get the output impedance of this preamp would be about 800 ohms. is there an equation for factoring the cutoff frequency with a particular filter cap? Im fishing around on google, but its terrible these days. Just trying to sell me stuff

Thanks for the last comment though, it got the hamster wheel a spinnin finally i think, and significantly reduced my confusion
 
Last edited:
I have no idea about modern DAC outputs, but I assume some of them are of fairly low output impedance to drive 0.1uF. It could be that the capacitors are part of the DAC integrator, I'm not familiar with modern DAC designs. But if they are the integrator caps, then your DAC will stop working properly, I think.

Maybe somebody here with more experience with DACs can help.
 
The dac is a two way DAC, its a usb audio interface from focusrite. It's output impedance through its 1/4" balanced trs outs is 100 ohms. I have mono trs/rca female adapters on it when i plug it into this class D amp. I used balanced trs to XLR balanced adapters when i plug into the yamaha. That's all i know about the dac im using. I may remove one of the 0.1uf caps and to a quick test on the tube preamp and see what happens. idk.

Only issue, is theyre small surface mount caps, im not confident i could put it back on properly. Got no gun, only an iron. I would think the it wouldnt damage anything to do so, just maybe increase the noise floor, but no confident assumptions here for me though. Im good at soldering pcb's, never messed with surface mount components before though. Pretty sure i would learn something if i pulled em and plugged the preamp in though. And somewhat certain it wont damage anything on the amp at least lol. I can manually lower the opamp gain in an app that controls this amp. Wondering what pulling those caps, then lowering the opamp input gain would result in.

The impedance of the preamp is so high, im guessing those caps are functioning like the tone pot in a guitar bleeding high frequencies of the high impedance signal to ground, in a guitar removing them slightly increases noise floor, but very effectively brightens the sound and you can tune the bbase high freq rolloff by changing the value of the cap even when not messing with the associated potentiometer. I do have some 0.01uf film caps, but no idea how i'd go about putting them on a surface mount board with pads that small. Got a feeling i may end up abandoning this power amp even though i really like it lol.
 
Last edited:
The lowering as frequency rises curve on your first graph is a pretty damn good representation of the response curve im getting plugged into the class D amp honestly.
Actuay there are two curves, amplitude and phase. The one that rises and goes then flat is the output signal level, showing excellent response. The one that goes down is the phase shift, reaching - 180 deg as expected.
 
yea i didnt know what i was looking at, just used it as a metaphor mostly. Bout as beginner as it gets with circuitry and measuring audio equipment both. I just jumped in because my skills at soldering have improved a lot recently. Am looking into some analog discovery equipment and accompanying software.

Im going to complete the modification, and if it doesn't improve enough ill mess with lowering that cap value, maybe raising the 20k to ground resistor value a bit (would this cause more loss through that input filtering cap though?) and dropping the opamp gain. I dont think ill destroy anything, and worst case scenario ill be abandoning this particular tpa3255 amplifier anyway. I'm thinking that feedback cap/resistor loop to the output of the pre will probably have the biggest positive effect on output impedance. But really i have no idea.
 
Last edited:
So i have an old yamaha AVR im not using with a buffered 1.2 k ohm rec out in the very cabinet underneath the gear in question and it feeds the input of this tpa3255 class d amplifier no problem. Just checked. Should be confirmation the mod down to 1.2kohm impedance on the tube pre should work. And if not ill just use the avr to buffer the signal. Its slightly quieter using it to buffer my DAC's signal then dac straight into the class D, but once volume is adjusted i can't tell the difference in frequency response.

Problem solved, sort of.
 
Im sorry for the triple post lol


Problem solved, i finished the mod and it brought a good bit of the high end back in, but not all of it, was still a little pretty dry to my ear, but listenable. I dropped the value of c3 and c16 to 0.022uf from 0.1uf and that sealed the deal. Sounds as fantastic as i would hope, and honestly my phono pre and my DAC/interface sound a little nicer as well now, why the hell they decided to use those values for bypass is beyond me. Seems quite pointless tbh. Definitely improved the airy feel of reverb decay and cymbal strikes. Also got rid of all background ac mains hum in the supply circuit of the preamp by bridging the outer leg of R9 to output ground with a 680uf 50v electrolytic (470 was recommended to me, but i had a 680 and it worked just fine).


Even running the power amp without those capacitors at all seemed to cause no adverse reactions in the system, and installing the 0.022uf films in c3 and c15 didn't seem to change anything about how the amp responds to input or noise compared to no cap at all. Makes me wonder why they bothered putting a bypass filter cap on the inputs at all. A 0.047uf mylar seemed to still pull some sparkle off the preamp, tried it first, 0.022 seems to be the highest value i can run where i suffer no appreciable losses across the wide band with whatever impedance curve i landed on.
 
  • Like
Reactions: jcalvarez
Great, well done, glad to hear everything went well. I have been looking at several tpa3255 schematics on the net, and have never seen one with 100nf in parallel with the input signal path. I have seen RCs with 100pf in some schematics, which of course is a very different proposition. Maybe you could check with the amplifier supplier the reason for those caps.
PS: I would really replace the 0.022uF for 100pF. I added a load capacitor to the simulation. Have a look at the bandwidth with both options. There is still a big impact on the highs with 0.022uf.
1702637776558.png
 
Last edited:
Yea i didnt have the legs of the new caps soldered properly to the surface pads. Tinkered some more today and soon as i got good contact, there goes the high end again. All my sources sound better with those parallel caps being gone completely.
Currently im using the amp with no parallel capacitors at the input as a i type, quite loudly with zero issues, and every source i have sounds fine, far better really. I have some surface mount 100pf caps on the way as well as a heat gun though. Those caps have got to be there for a reason, i dont plan on leaving them out of the circuit


I have not looked into the circuit of the amp beyond the input though, the input schematic was the only schematic the company was even willing to give me. They wont give me a full schematic of the circuit once they fund out my intention was to modify the product, and not all the R and c positions on the board are even marked only some. (weird i know). Told me they cannot provide schematics for the purposes of self-modification. But I did let them know that using a bypass cap of that value on the input is shunting all the high end from a high impedance signal that should pair fine to ground making it completely unusable. And changing the value of that cap to something a lot lower would likely fix a design flaw in the amp (it's a dajunguo zk-as21 dual tpa3255 ic power amplifier that has honestly been quite unreasonably impressive in all regards aside from pairing with this tube preamp lol)
 
Sorry for double post, again, too late to use the edit feature.
Interesting note i meant to mention, the rec out on a yamaha avr i have is 1.2k ohm buffered. Which should be almost exactly what this tube pre is sitting at with the mods i did according to source of the mods, the pre was still all bass going in before the cap removal, but run through the avr's buffer that should be a very similar impedance it sounded fine into the class d's input, as the pre running thru the avr buffer still does post cap removal. Both the preamp kits i built and modified react the same way to everything i've done. Most of which still isn't making much sense to me.

Is there an implication of some kind here i dont have the knowledge to grasp?
 
I completely missed this. Those 0.1uf caps are probably a low-pass filter for tbe subwoffer channel. Makes sense.
Finally figured out how quoting works on this forum lol.

Are you sure that makes sense though? (i just want to make sure i understand if it does) The knob that controls my subwoofer crossover frequency and the level knob both are not responding any differently then before, i still have it set to roughly 100hz and i didnt notice any change to the low end at all removing the capacitors. The signal is split for the two chips somewhere after this input circuit, any low pass filtering on the input affects both chips wouldn't it? Correct me if im wrong, but reintroducing high frequencies after they were removed with analog circuitry would introduce noise to the signal wouldn't it? There is also a 100hz high pass on the non bridged chip, whatever circuit handles that is definitely not at the input. The tone controls on the amp are sub freq, sub level for the bridged chip, and bass and treble for the stereo chip, and master volume. It's spec sheet is essentially identical to the fosi bt30d pro. I wouldn't be surprised if it's the same circuit ripped off, with only a few mild changes, mostly to component quality, id imagine. And personally i think it sounds better then the bt30d pro (most of the time) if i were to be honest. It definitely has a lower noise floor and i prefer the way the high freqs come across, so i think i just prefer whatever's been done with this amps output filtering. And without those caps, i like it even more to be honest.

P.s. I keep coming across posts of yours while trying to (poorly) educate myself. I even have bookmarks saved i occasionally reference from before i joined up, and your posts are in many of them. Thank you for all the information. Genuinely. I totally understand how frustrating it can be to help someone this clueless on subject matter. my memories not what it used to be, and i basically have to beat myself over the head several times with new information before it sticks without notes and references everywhere.
 
Last edited:
P.P.S. A friend just came by. He is swearing up and down my amp is making the same sound a cathode ray tube makes but intermitently, even higher pitch, and kind of ungulating when it does. That cr tube scanning is a sound i havent been able to hear for a few years thanks to all the loud rock music and the age on my ear drums (if i cup my ears forward, and get in the right part of a room, or other side of a door, i can still JUST hear crt scanning but i have to go out of my way, i cannot hear this sound no matter what i do). Apparently my buddy has some lucky golden ears still somehow. He mentioned it immediately upon starting a record, and he assumed it was coming from the hardware in the turntable. But it does it with my phono preamp when used as a buffer too apparently. My focusrite DAC/interface does not do this, but it's got very little analog circutry in it.

Is there a possibility those caps were SUPPOSED to be 100pf? But the 100nf caps may have been used by mistake thanks to an error when designing the schematic/boards? Seems very within the realm of possibility knowing what little i do about chinese manufacturing. I wouldn't be surprised at all if those caps were there to suppress super high frequency noise on resolving enough setups with certain sources? I'm using GRS ribbons that are dead flat in response and impedance both all the way out to nearly 40k. If i had to guess the sound he's hearing was likely between 18-22khz based on what he was describing (Dude's almost 40, what a lucky bastard lol) My focusrite interface that isn't making any high freq noise outputs at like 95ohms, idk if thats useful information or not. the phono preamp is 250, and the new tube pre is 1.2k ish. The new preamp makes the most high freq noise, followed by the phono pre, then none at all on the focusrite, or it's just too high freq to hear. idk if that's coincidence or implies something.

Dajunguo's rep's on amazon have also sent me an email adress and asked if id be willing to correspond directly with their product engineers. idk if thats a good idea since im a total novice with no real information, but they are actually interested in the information i have about that capacitor though apparently once i mentioned it. They wont give me a full schematic, but they're totally down to let me try and fix their product for them hahaha, yay china!
 
Last edited:
To be honest most of my comments about the rest of the system (apart form the tube preamp) are kind of "educated guesses". It is not possible to be sure without looking at full schematics or at least detailed technical information about all modules in the audio path.

Talking to the product designers should definitely help. Good luck!
 
  • Like
Reactions: Tha(t)oneGuy
Overloading one's self with too much information one doesn't understand too quickly does create a tendency to ask dumb questions and extract the wrong take away.

i've hit the part of the dunning kruger curve where i've started to realize how dumb some of my questions are and why. Time for reading and education so i can understand what the hell i'm even trying to accomplish here.

And, honestly, the inductors are the only thing on the bt30d pro i would call higher quality components then this amazon dajunguo amp. Better filtering caps, better quality lower esr big caps, whatever those are for (lol, obsiously i have assumptions) The filter caps on the bt30d are surface mount ceramic looks like, on this cheapo amazon amp they look like giant versions of the orange spragues in my guitars. Definitely metalized poly caps of some kind, probably why my ear prefers the dajunguo most of the time. The only cheap class D i've heard personally and liked more was the v3. But im a sucker for these passive sub outputs, i rarely use my v3. Unreasonably impressive is usually a perfect way to describe the thing. Wish the brand was easier to correspond with effectively
 
Last edited: