Hacking a Jet City Jca2112rc

[shreditup]

Well i haven't checked but i think i heard the seller say 1 buck for small tubes(el84's, 12ax7 and the likes). Bigger tubes are more pricey(a 6l6 costs 3-4 bucks, haven't checked the others), but still that's cheap. Though if i had the money i would def buy most of them then sell em for more.

The main concern right now is that i need a way to attenuate the signal hitting the power tubes because it's just really unpleasant and unplayable if you hit the strings hard.

Basically the sound gets much louder, thin and shrill while the strings that i hit hard are ringing, then when they decay or if i'm playing soft it sounds good, so i guess it's a way too high signal amplitude for the el84(even if i plug my guitar in the fx loop in, i still get the shrill and really overdriven sound when i hit the strings hard, especially considering its only 1 gain stage, pi, and 2 el84's from that point). So how do i attenuate the signal hitting the power tubes?

I've tried lowering the b+ for the pi, didn't work, and tried a larmar ppimv, which helps but dulls the sound. Any other alternatives?

[J M Fahey]

Respectfully I suggest you scope that amp output, and different points along the signal path, to see what you *really* have there.
a) you may have the EL84 grids rectifying signal and blocking themselves with high signal peaks
b) the amp may be marginally stable, and oscillate on parts of the cycle
c) sometimes you may have grounding issues, "negative" feedback becomes "not so negative as you think", it causes some visible (ugly sounding) kinks in the waveform.
d) maybe it's not an *EL84* issue but a *PI* issue .... or even something with the last triode preamp before it.
Since today your only way to judge sound is hearing it through the speaker, you are hearing all 3 stages I just mentioned mixed and at the same time.
It's hard to decide which part of the problem comes from where.
e) you name it. A high gain amp is a complex beast.
Good luck.

[jmaf]

Well i haven't checked but i think i heard the seller say 1 buck for small tubes(el84's, 12ax7 and the likes). Bigger tubes are more pricey(a 6l6 costs 3-4 bucks, haven't checked the others), but still that's cheap. Though if i had the money i would def buy most of them then sell em for more.

One dollar el84's? Sell your car and buy them all.

[shreditup]

Well i dont have a scope, and i guess it's not easy to diy one.
a) Tbh i dont really understand what you mean
b) Strange considering it only happens when i hit strings hard... Or maybe not. Confused
c) Feedback disconnected. But it may very well be a grounding issue, dont know where though.
d) well it is most likely with the last triode because the strange amplification nearly completely disappears if i bypass the fx loop. Though the amp is not loud enough without it.
Then again a larmar ppimv solves that to a (large) degree, so i guess it's most likely a) or the way too high signal swing, which needs less tone sucking attenuation.

[shreditup]

What if i use a oscilloscope software and use my (embedded low quality) sound card? What additional components do i need and how do i do it? And should i do it at all, since i figure it wont be very accurate and all.

[jmaf]

Well i dont have a scope, and i guess it's not easy to diy one.
a) Tbh i dont really understand what you mean
b) Strange considering it only happens when i hit strings hard... Or maybe not. Confused

It's not strange at all, it's all part of the same issue. When your signal swing drives the output tube grid positive, the phase inverter impedance is very high, so the signal can't drive any current. When the output tube grid goes positive, it conducts towards the cathode, grounding your signal. Positive grids are like diodes down to ground.

With a scope you would be able to see the waveform and see if the positive swings are getting clipped.

What you describe sounds like excess gain, driving the output tube grid positive. Getting the distortion when hitting the strings hard is an indicator.

[shreditup]

Wont measuring the dc voltage at the output tube grids with a multimeter when i'm playing help?
And yes, this excess gain is pretty close to what i had in mind. What gain are you speaking of and how do i reduce it to el84 friendly levels?

[J M Fahey]

Yes, it's basically that, coupled to the fact that the now forward biased diode (grid=anode; cathode ... well ... cathode) conducts very well in one way, not at all the opposite , so the coupling capacitor gets charged to a very high *negative* voltage, the peak of the highest signal hitting that grid.
When the signal level lowers somewhat (meaning everything else except the peak) that tube has a *huge* negative bias which turns it off, only passing choppy farting signal peaks.
Horrible.
That is called "grid blocking distortion". It's clearly seen on a scope.
A couple months ago my reliable Kikusui Scope died, obviously on a Saturday afternoon when I needed it most.
In desperation I downloaded some Software Scopes, a couple actually worked very well.
The only *important* precaution is that since it's the direct board input, it has low impedance (think 10K) and needs around 200mV, period, or you blow it.
I kludged a rotary 5 position attenuator which *always* had some resistance in series with the actual board input, and added a couple antiparallel 1N4002 diodes across it, so you never go over safe 700mV peak.
You may use (hot crocodile).022x630V/220K/47K/10K/2K2/470r/100r (to ground crocodile)
I put a .022x630V capacitor in series with the attenuator, to separate it from plate DC (which anyway you can't measure or display)
The other problem is that it's AC coupled, so you can't see DC values, but to see clipping points, waveforms, oscillations (you can't see the actual HF waveform but its envelope) and assorted stuff, it saved the day.
Better than nothing and you can't beat the cost.
I found this one very good: Winscope
You have a similar one here: Oscilloscope
Good luck.
PS: remember to protect your soundcard input, high voltage can *easily* burn it.

[shreditup]

Ok
I have never worked with an oscilloscope before, but i guess the plan is to use signal generator from the software oscilloscope via speaker out into the input of the amp, then instead of a speaker connect a cable to the output jack with the signal side in series with this .022 hv cap then a 220k(or the others that you listed) to from signal to ground, then into a mic/line in on the pc and observe. This is probably wrong, but it's a guess. Obviously i wont try it until you approve(the fact that the amp output has high current that'll go into the mic in on the pc kinda frightens me).

[shreditup]

I forgot to mention that the effects of the strange amplification are apparent only at lower levels. As the volume gets beyond a certain point, about loud conversation levels, the sound loses these loud farty sounds and gets pretty normal, apart from lots of presence, which may be normal. The loud farty sound is more or less constant in volume and and is heard only if the normal sound is not loud enough to overshadow the farty sound. Also i get sound even if i turn the mv down(more volume with the loop, almost none if the loop is bypassed). This i guess sounds like a signal from the ground somewhere in the preamp get's amplified and even more if the loop is engaged, if that is even possible. But then 2 things don't add up:
1) When i plug my guitar in the fx loop in the sound is normal until i crank my guitar volume pot, which is when i hear these strange whistle noises along with the normal sound(which not that normal because it's very gainy).
2) Why does the ppimv lower the effect a lot?

I'm thinking it's a phase inverter issue. I'll check all the components and wiring related to it tomorrow.

Meanwhile on the oscilloscope
http://tone-lizard.com/images/oscilloscope_setup.gif
I found this here pic. So i should use a 16ohm(is 15ohm ok?) 100w(or less, i wont need to crank the thing at first) at the amps output jack, then i don't understand what comes next in between it and the line in. Can you draw a schematic or explain more in detail?

[J M Fahey]

Yes, something like that.
Some slight corrections:
1) suggested attenuator, an image is worth 1000 words.

Not Lab precise, simply to lower the intended signal into something the board can handle safely.
The diodes are there in case you touch something you shouldn't.
2) You still will *always* leave your speakers connected, your amp needs them and they are the "real world " load.
An important side effect is that you'll very soon educate yourself to match what you see and what you hear.
It's one of those OH! moments.
3) Start by using your guitar as usual.
Play the "ugly" chords and see why they sound ugly.
Then, as a diagnostic tool, inject some sinewaves at various levels, which you have previously found using the guitar.
I personally don't use the PC as a signal source and at the same time, as Scope.
To simplify things I download a 1KHz MP3, play it on my MP3 player and set it to loop (repeat endlessly).
Most can produce 200mV signals, which are ample for this.
If you want your hands free, download some clean guitar solo (guitar only, no backing instruments) and use it ... or record one yourself.
IMPORTANT: although I included a 500V setting, it's a very dangerous voltage level, it's there only if the signal at a triode plate is somewhat larger than the screen.
DO NOT USE IT ON POWER TUBE PLATES OR SCREENS.
You have practically the same waveform, at safer levels, across the speaker terminals.
BEWARE THAT A CROCODILE CLIP IS TOO SMALL AND YOUR FINGERS WILL GET TOO CLOSE TO WHAT YOU ARE MEASURING.
Get one of those probes that look like a multimeter test probe, but have a small spring loaded hook on their tip, so you can clip them to component legs or socket pins.
Good luck.

PS: or do what I did when I was a young broke student in the early 70's: I went to the "Universidad de Buenos Aires , Facultad de Ingenierķa" (substitute by any approppriate University or Technical School close to you) where I was studying, asked permission *very* politely to the Lab Manager, and he let me use all those wonderful 500 Kg , trolley mounted Tektronik or HP Scopes they had, plus signal generators and assorted measuring instruments; at idle hours of course.

[shreditup]

I'm still a bit confused by this.
1) These voltage numbers are more or less imaginary, right? Afaik, the output transformer outputs ~15-20vac with el84's, it's not possible to get 500v there.
2) I don't understand the need of all these attenuation combinations in my application. I should attenuate the given 15-20v signal to 500mv(more and my audio card goes dead, right?) And i dont have a any kind of multipole switch(should get one though).
3) I also don't have 1n4002 diodes. I do have some 1n4007 left from my jcm800 clone build, and some 1n914's. The 1n914 suits, right?

[J M Fahey]

The attenuator I suggested is the full version, what I built for my own use, lets me see signals from 1V to 500V *anywhere* on a tube amp.
Don't worry, if you *only* want to see what's getting into your speaker, measuring across its terminals, you can use:

Yes, you can use *any* silicon diode: 1N400x , 1N4148, 1N914, etc.

Now that I notice it, you can capture your screen (search for some small program or add-on to do it) and show us your waveforms.
Good luck

[shreditup]

Should i also use the .022 cap in series before the attenuator?

[J M Fahey]

Not in this case, since you are clipping your test probes across the speaker terminals, where there's no DC.
Even less so in a tube amp.
If you want, you can add a .1 or .22uF x100 or 160V in series, in case you want to test an SS amp someday and it happens to have DC on the output.
In any case, PCScopes do not show DC voltages, only Audio/AC.
Just apply a nice middle of the road frequency such as 1000 or 440 or 400 Hz and test .
*It is not a real scope*, but it saved my bacon many times and the cost is zero.
There are some full function USB digital scopes, but I'd save a little more and get a surplus "obsolete" real one .... even if tubed

EDIT:

I searched and Yahoo says the JCA2112 circuit is a Soldano Astroverb.

soldano_astroverb_16.pdf

Is that an official schematic?
Or some reverse-engineering by a Musician?
Because it has some obvious mistakes.
Please re-check it.
Good luck