Good advices for a newbie (threading tools lathe)

[rendoman]

Thanks to you PJs! It's a pleasure to read your advice!

REW - Room EQ Wizard Room Acoustics Software
I tried this free software, before that my soundcard decided to explode . It's easy to use and so good.
This is the graph of my horns + sub (without tweeter ), that's my last measurement, the day after boooom



I buyed from ebay auction this old rca audio generator, all tube circuit, new rca tube. Maybe I should clean with some dry spray cleaner potentiometer and rotary switches, more i use commands, better the sine and square waves are, maybe some oxide inside.



You touch the problem with flame shape

I tried measure amplifier with oscillator on, it disturbs too much signals, after all it's a strong rf emitter, then I decided to unplug kv and regulator power supply, I only kept the 300v anode for the first part of circuit.
With amazement I discovered that I got troubles with this stage. I havn't found the reason why things happen, but this stage has an incredibile gain, with very few turn of imput potentiometer distorsion raise quickly.I got some doubt about potentiometer, I've never seen a pot running so strange. Circuit should be "easy", I will try triode connection instead of pentode, maybe the gain is really too much.
I'm thinking about how to find the right way to determinate which load can be good to drive oscillator, now, even with 3w output pentode, raising the voltage gain seems not touch the volume of sound too much (over the threshold). Maybe I need more current from amplifier than voltage, current that I can find with el34 tube, or maybe trouble is elsewhere

In this 3 picture little curve is imput, 0,5v\square div, than 2v imput. Nearly full closed pot, 5v/square 2ch at output from pentode. Even with signals under volt distorsion raise quickly and volume doesn't change. sorry I don't remember the frequency, something near 500hz maybe.

In first pic that's the maximum "clean" gain, after this gain raise with dist. Different shape for square wave

[PJs]

Based on what you said above, the first thing I would do is squirt some cleaner on the FG pots and switches, like you said. Then I would hook it to the TEK scope and run it through the full range (slowly) while watching it on the TEK scope. You should get some indication of any issues with the quality of signal out. Also I wouldn't waste time on the Square wave at this point. It will give you some clues later but not necessary to diagnose where the issues are. If it's clean Sine all the way through you should be Ok to step 2. If not, note the frequencies where it distorts or has sudden drops/increases in amplitude. Another thing with older equipment like this is sometimes the solder joints get a little cold or corroded. Just reflow them and try again.

Next is the probe of the scope to the input of you preamp and its output all while sweeping your FG (if its clean from test/repair above). You will probably notice some slight degradation of signal quality as you move through the amps and to the coil...and you can look at your Real Gain levels and Q.

I didn't notice any new Faraday cage in the pictures like we talked about and that may be contributing to your problem hooking the scope...Or it might be an impedance problem loading from the scope/probe. Do your probes have a 1x/10x switch? Might try the 10x if you do. If they are only 1x it could be part of the problem you are seeing. I see no issue with measuring the KV measurement even at the RF with the TEK scope, but I don't know that scope very well. The TEK is a 60mhz...so it should handle your stuff fine. The probe may also be tired or have slight cracks in the housing? Also unloading the KV for measurement may not give you a real world look, being unbalanced.

What kind of Pot is on your amp - wire wound or carbon, and what kind of volts & ma (watts) is across it? A wire wound on a circuit like this may give you some inductor type issues particularly at the bottom end, by taking on some of the spurious emissions going on in proximity. Might try a quality carbon pot to see if it's any better.

You may have some issues with impedance matching throughout or maybe some distortion issues in the circuit, but I thought what I heard on the videos was pretty good over all with only some minor glitches going up the scales.

Like I said way back in the thread...be as methodical as you can, start at the input, look at and solve one thing at a time, Document the problems and fixes as you go, then to the next till you get it the way You want it at the plasma end!!

What you've done here is remarkable and really hasn't been done in years except for a few wild and crazy guys out there/here who like to Finger it out and make it work. Again, Great Job...Diligence and patience will pay off.

~PJ

[rendoman]

Hi!
I tried with tek scope, sine is pretty good and clean, output is pretty high voltage. There is also a "line freq" output asjustable up to 6v (don't know what it's for)

That's a pdf of manual found online
bama.edebris.com/download/rca/wa44c/wa44c.pdf

Inside of rca is pretty new, i clean only with air, for sure cap are vintage, but seem good, I don't want to change every cap now, instrument is running well imho.
I will try tomorrow to check "preamp"only, and maybe connect triode instead of pentode, gain and power will be less. Probes are all 1\10x, checked with trimmer. Unfortunally I never made faraday gage, because I'm always undecided with new frame since I will change 80 or more % of circuit and I will need more space. I'm thinking to try a light alu grid as gage, just for test. Trouble is that oscillator is strong enough to disturb scope even with metal screen ( I tried with a solid brass panel in front of coil, something better than nothing, but circuit is still critical, maybe I will need some bulkheads between sockets.

Potentiometers are 2w carbon, chinese unfortunately. 100k log for audio. One things not so good is the load of pentode, a 20w resistor, instead of choke or output transformer, for sure it converts current in heat like candies!

You're right, it's time to be methodical, I need to check every single part and optimize.
I will use 813 as experiment with external socket, but I don't think I will use in normal operation. Too expensive imho, with the price of one tube I can buy some 20 gu50, maybe I will parallel output... thoughts

Hope to post some new good updates!

[PJs]

Hi Stefano,

The FG looks pretty good on the inside but how did the signals look on the scope? Did they jump in amplitude or distort at any frequencies? Wouldn't worry about the SQ wave stuff as the manual states it isn't that good and slants, but good for quick checks...so you would have to compare before and after to really get any kind of clear picture.

One thing I noticed in the manual for the FG is the impedance and the ("Limited to" ) .25% THD. It's old school and built like a tank, which I love, but for modern audio...its what you have so keep it in mind.

Pg 7 shows it to be for a 600Ω resistive load for DBM. Pages 8-11 have good procedures to find out the impedance of the system, DB/DBM, the freq response and IM distortion...All good simple hookups and tests. The voltage out should be Ok (8V RMS) for your amp but figure max input for your amp and hold it just below that. Also in the manual the multiple warnings about connecting around the B+ is a good Warning...you will pop some components if not careful.

For a temporary Faraday you can use Al screen door material. Keep in mind it may become charged so make sure its grounded well in several places. Copper is always best and should be porous Screen or perf metal. Some older computer cases have some copper shielding if you can find an old case somewhere with it you might be able to salvage enough from it. MY old Dell XPS had a lot inside with nice finger grounding strips. Wish I had saved it now.

The Pot should be ok. As for the output of the pentode...Maybe you could devise an LR for it...but would be much better to use an Xformer if you can scrounge one up.

Personally I think you will be throwing too many variables into your working circuit by building something new with the 813. Again, You may be pretty close to the power you need with the gu50, imho...multiples in // will add other issues in matching impedance's and outputs with out buying matched components.

Please be clear these are Just my opinions and observations...it's your project and you have done Very, Very well so far.

Good luck and may the Force be with you.

~PJ

[rendoman]

bad news today

I tried to measure both first stage gain and second one. I switched between pentode and triode connection, I changed from resistor load for pentode to output transformer.

summing:

- first stage voltage amp (little triode of pcl82 tube) imput 820mV for maximum voltage gain of 50v




- output stage triode connection , imput from generator (I will measure this a 3rd probe even from first stage) max 360mV for 114v output




- output stage pentode connection , 244mV for 136v output



Clean limits, after this threshold curve become bad.

I tried also measure with output transformer and speaker - 9,1 ohm resistor ( I didin't have 8ohm at home ), output voltage is really low, under half watt in "clean" curve, I settled tek scope to ch1-2 rms voltage. Anode voltage is 310v, I'm thinking about something happened to catode resistor. I need to focus and think what happened inside circuit.

I put x10 probe (x10 key on tek display) on pentode point, 1x on low voltage signal
I think I'll check every component, cap, resistor, socket. I will be ignorant, but the p-p power should be max voltage output (undistorted) / load resistor in ohm, maybe something near 0,2w, Who has eaten the other watts?
I will find an answer!

funny picture with ear protection



Here a small noisy video, at the end you can see the sine changing

Little edit, I missed one part of formula, power is 1,5w in triode, not 0,2. power is V2/r or R*I2 or V*i

[PJs]

Hi Stefano,

You definitely have some distortion and clipping going on from the pics and vid. Unfortunately I don't know your schematic, nor what you are measuring and where, so it's hard to make any kind of educated guess.

From the video it appears that anything above ~400mv (input? from FG?) and 120-134V (Output?) is were things go wonky. Is this the preamp stages and the pentode goes to the GU50 as a final stage? It also tells me your gain is way more than you may realize which always leads to distortion and clipping.

One thing that comes to mind is the cap between the triode/pentode may not be happy or proper. Specs for the PCL82 say .02 & .025pf. Also note the Peak Anode voltage and currents for the Pentode section...If your gain is too high it will lead to saturation and cause the plate to glow and deteriorate the filaments. Are you running the filaments on AC or DC?

Here is an old cut sheet I found for the PCL82. I assume you built the amps and coupled them around these specs??

A bit in the dark here to help any, but it does come to mind that your sound is pretty loud & clean even at the levels below clipping/distortion....even with the speaker hook up instead of the plasma. Glad you found an xformer. Also Not clear from above about the 9.1Ω Resistor with the speaker...series-//? Seems to just add a resistive load to the inductance of the speaker??

Thoughts... It's not always about power, but efficiency is king in my book. Look at the efficiency of the old ESS Heil AMT 1b's. 1W input gets 90db at 1m!!! Even though they will handle 375 watts, not sure you could stay in the same neighborhood at full power!! Mo-Power will give you cleaner sound at the clean operating parameters of your amp but anything after 90-100db is a waste in my book. Adding your horn to the plasma will give you even more sound pressure.

Thanks for the pics and vid...

~PJ

[rendoman]

Hi
I'm trying a different amplifier with 6sn7 single ended +el84 pentode. I built the circuit in rapid mode, on the table. Unfortunately I havn't 5k output transformer (only 2,5k), nor local shop has 20w 5k resistor (yes, here the quality of the shop is everyday worse!).

few words about circuit, tomorrow I connect the amp to the plasma and see what happen. Compared to old pcl82 amp this new sound better and louder. I think I need to check every stage but at first sight with same clean 0.2v rms imput I got 5 volts good output, it's more than 2x old amp (clean 2,5\3w). If I close an eye I can run circuit up to 9\10v output with some distortion, but still acceptable (5/6w). Leaving out the measures, connecting a speaker to outpu,t sound is really louder, and that's what I need in order to check spl connected to flame.
That maybe only an experiment, soon I'll try El34 tube, maybe in the same good octal relais socket, maybe the final tube I'll choose

Good thing, despite long cables and chaos, no hum, buzz, or strange sound!

[rendoman]

New driver is better than the last! Power is double, with better sine. Volume is loud from speaker, and even from plasma flame. Maybe a bit more power can help, but even now is good.
I tried just for curiosity this amp with my horn, it sound really good, even with a poor output transformer (frequency range between 100hz - 20khz), but I see too much power for my speaker, I think max 1,5watt can be optimum.

New attempt contemplates EL34, in pentode should give 11watt at 10% distortion, with 0,45v imput, practically double than now!

Single ended 6sn7, only use triode used. 300v
Imput under 0,5v rms, 7v ouput with I think good curve shape


El84 pentode connected, 300v. Good voltage gain, I tried taking signal from output transformer + load it's nearly 2,5 watt very clean, some 5watt with little distortion

[rendoman]

I hate brass, in every shape
Have to sharpen drill bits better. Maybe I should also take exemple from this fantastic forum and build some clamps - boring head, mine are bad

This is what I thought when I said modular frame Yesterday. Actually I need to try many configurations, with different tubes, that's why I prefer to make single solid brass (can be a definitive way) individual sockets. I think I will add a fast mammut connection or something similar, plus a spike each for gnd point.
These single frame can be mounted on columns or just for prototype with long allen bolts. I need octal, gu50 and maybe one noval frame.
In the future, for sure, I'll build a planar chassis drilled as base, good to mount single modules. I'm thinking about building single cells shielded, easier to build and to control, above all RF power unit, that clashes with everything

As planned, old first tubular frame will be reconverted as home for my old ecl82 amp ( 2 noval sockets ready to use, enough space for my big diy output transformer, choke and maybe a regulated Hv psu with gas regulator ( 0a2 and 5651 maybe ). The front nice panel in polished brass have a pot for volume, 2 switch for filaments and Hv and I connect ammeter just for appearence, I don't need to know how much current flows in this circuit ( fortunately ). I have to find good ammeter for plasma, I will connect them on ( - ) rail of psu, last time I connected 1kv (+) at 50ma it burns like thermite

[Toolmaker51]

Fun to read, no clue as to what's going on. Could be Trabant turn signals for all I know. Definitely not one of my languages, that's certain!
Toolmaker 51