Video

I understand what you say. I have a Williamson amplifier that I put 7591 in and now they are $52 each - just too much - I will probably change them out (have to make some socket wiring changes) to 5881 that I have lots of. Of all the solid state amplifiers I have, I like the McIntosh 2100 the best. 105W per channel and for some reason like it better than the huge 2205 at 210W per channel. I doubt I will get rid of any of my tube amps but tubes, along with everything else, and taxes, is just going thru the room.

Totally agree - the comment at the top, transcribed from the Mullard document agrees with us also. Thanks for your comments and confirmation.

You have a good point. I like to spice up the equipment some time. I joke that most of my equipment looks like surplus equipment from a cold war underground bunker with the grey chassis. When I get really fancy, I paint my chassis blue like the headphone amplifier I built for my grandson. 🙂 I do like symmetry try to dress the under-chassis wires along the sides of the chassis rather than direct - although sometimes short, direct connections can be the best. I do hope to build some more audio amplifiers soon. Thanks for your comments.

One thing I would like to say about this amplifier is, in the ARRL schematic, they directly wire all the screens together. I found that separating the screen connections from one another is a much better way to go. Everything works better when each screen has a little wiggle-room. I don't see it in this video but I believe I added 470 ohm 5W resistors in each screen lead right at the socket. It did wonders for stabilizing it and improving the output.

My most serious interest does always come down to amateur radio and RF but my YT channel was initially intended to be primarily for audio equipment so I thought it would be more appropriate to have a name not specifically for ham radio.

This amplifier does have screen taps and are used. It evolved into a much more powerful 6550/KT88 amplifier. Not sure but I think I made a video of that version using the 5U4 rectifier and then a solid state rectifier. A good bit more power with the higher voltage from the SS rectifiers. Being very conservative with vacuum tube rectifiers, a 5U4 is really only good for an amplifier of about 30 watts output. I do still have this amplifier.

This was my first modulator to modulate a pair of push-pull 812A tubes in a 1947 James Millen amplifier. It was too small and I ended up with, and still use in my AM transmitter, a modulator using Eimac 4-125A and a BC-610 modulation transformer. This one does a great job of modulating the Millen amplifier. I mostly keep it at 7293 KHz so maybe we will meet on the air some time for a QSO.

🙂 Absolutely... that solder resin keeps the mustache under control and comb the cats with it too ... 🙂

@ElPasoTubeAmps Well played, my friend. Make it a great day . Til next time. 73. DESPERADO 56 north of Tampa FL. When conditions are right.

I use an external USB sound card like, in this case the Behringer or maybe a Focusrite or whatever you might have that the ARTA software can connect to thru the USB port. I am confident it can find any USB unit. If the external sound unit has mic input you can use that for high gain, if you need it, or use the line level mode for higher voltage levels from the device you are measuring plus some attenuation network, etc. to get the input signal to a level the sound unit can handle.

@@ElPasoTubeAmps ok, can i use such thing for measure the gain of smps ? I develop resonant converter whixh need to explore the mag gain

@@endremurti Sure, if you can measure the input accurately and then the same for the output or you can do 10*log of the power ratios with a volt meter. I am not sure the dynamic range of this FFT device is adequate to measure input accurately but I may be wrong. You will just have to try. I can give you an example. Suppose your driving voltage is 1.5 volts across an input load of 500K (you convert this to input power and get 4.5 microwatts) and your power output is 50 watts, the gain would be 10*log (50/4.5E-6) = 70.46 dB. Hope this helps.

Tom, per all the specs on Hg rectifiers, only a choke input filter is recommended but I have used these rectifiers with capacitor input filters and in voltage doubler circuits which by design has capacitors at the output to the rectifier - running them at 4KV output - and never had any issue. I would keep the size of the input capacitors relative small, as you mention, at 40 uF or below. You are right, these Hg rectifiers are tough. I like to call them the solid-state rectifiers of 1930... voltage drop is constant at about 15 volts regardless of the load. As for the 3B22, it is a full-wave rectifier so you need only one. Thank you for your kind comments. I learn something from you all every time I post a video and have met many very nice people over the years on CZcams. Stay safe...

@@ElPasoTubeAmps Thanks again for such a thorough response. Regarding the xenon gas 3B22 would you say it also could tolerate a 40uf first capacitor in the power supply? Since it only is rated for 725V PIV have you run it past that? I suppose a 1N4007 in series with it would help.

I was inspired by this video to run a test. I ran two tests, one with dual 649/5834 mercury vapor rectifier tubes and then dual 3B22 xenon rectifier tubes in a full wave CT arrangement on a Hammond 278cx power transformer. I coupled them to both a 2000 ohm load and then a 1000 ohm load, representing approx 250ma and 500 ma loads as might be expected in higher power amplifiers. The filter network I used was a 100uf capacitor, then a .5mh choke then a 120uf capacitor. I certainly exceeded the piv rating as I had over 1000V piv. Neither the relatively large first cap nor the piv spec busting did I notice any problems. They even withstood switching (after warmup) into completely discharged caps without arcing. They also, as you show in your video, have much less voltage drop than vacuum rectifiers. The 3B22 xenon tubes had less voltage drop than the mercury tubes. I intend to use these tubes in my next tube amp.

@@tomlang6480 I have experienced similar performance. The PIV rating is very conservative and so is the current rating. I used 866A rectifiers in a voltage doubler circuit raising 1800 VAC from a 700 mA transformer to 4KV for an 833A RF amplifier and never had an issue. The ratings of Hg rectifiers are always into a choke input filter and the voltage doubler by design is capacitor input to the filter. I turn the HV on during transmission of off during receive so there are many cycles of on/off and I never had an issue with the Hg rectifiers charging a 40 uF capacitor at the input of the Pi filter network. I don't have as much experience with the 3B22 but I figure it is just as tough. Thanks for your information and comments. Stay safe. Here is a link for vintage components you might like. n4trb.com/AmateurRadio/RCA_Ham_Tips/rca_ham_tips.htm

@@ElPasoTubeAmps No kidding about staying safe! I notice after all these years (decades) I automatically put my left hand in my pocket around this stuff.

Thank you for your comments. Been a while since I have used this device but still have it in good condition. I don't know if I mentioned it in the video, the cables from the amplifier socket to the test jig may introduce instabilities in some amplifiers if the amplifier is swept with a sine wave. I bought ferrite cores to put around the grid an plate wires but never got around to it. Also some small bypass capacitors in the 100 pF range might help quell any parasitic events but for static measurements it is really handy for matching tubes. Stay safe...

I think you need another base and switch assembly that you can take parts off of to save your heirloom microphone. They are not too expensive on Ebay. I have one similar that was bought for me in 1963 when I got my general class license - 61 years old - and is special to me. I keep them going and have acquired several heads to experiment with different mic elements. I have also found that the inexpensive condenser mics sold on Ebay and Amazon have a QC problem. Some work much better than others so a simple method of evaluating them would be good. I just ordered some more... 🙂 and will work on a method to sort the good ones from the weak ones and plan on making a video. Thanks for your comments. I hope you can save your 70 yo microphone for another 100 years. 73 WA4QGA

Thank you. You are very kind. Here are a couple of additional videos when I got more serious about it. Yes, the antennas are very directional and need a good size tripod to mount them on. czcams.com/video/KLhxggZbsVc/video.html czcams.com/video/SDLm1N10fTI/video.html

I can't remember all these measurements without reviewing the video or remeasuring but I would say 3 mV out of, 450-500 volts for the plate supply is excellent. I don't know if I mentioned it in this video but I have found an important and fundamental design to be most effective in reducing hum in audio amplifiers and that is to keep all power supply components, including the choke and especially the main electrolytic capacitors in one corner of the chassis away from the input connections. There is indeed, current flowing thru the chassis between the CT (or negative chassis ground) of the power transformer and the electrolytic power supply capacitors that will be picked up by the input circuits if the currents are flowing thru the input portion of the chassis. If the chassis layout is not proper with the PS isolated from the inputs, no amount of adding capacitors or anything else will fix it. Thanks for your comments.

Thank you for your answer @@ElPasoTubeAmps

Thank you for your comments. I emailed Doug the other day and he is doing good - but getting old like me. I think he is rebuilding something exotic and will, hopefully, post soon.

When I started exploring inductance measurement, I realized there was a steep learning curve, and I'm still learning. Certain inductors, like the 1000 Henry UTC audio chokes, are designed for minimal current flow, primarily to generate grid voltage for the subsequent stage. Power supply inductors need to be tested at relatively higher currents, such as 500 mA, aligning with the ratings of power supply chokes. When measuring vacuum tube output transformers with inductance values around 20H, I found that a small current of about 10mA was crucial. This is because at microamp levels, mere movement of the transformer on the bench can alter the surrounding magnetic fields, drastically affecting the measurements. As for the testing frequency, power supply chokes should be measured at the line frequency or twice that for full-wave rectification, whereas 1kHz is more suitable for audio inductors. While inductance theoretically isn’t affected by frequency, practical measurement prove otherwise. Measurements of an inductor without specifying the frequency and current level involved might end up being unreliable. I hope this explanation is helpful. Thanks for your comments.

That is a big discussion you have going there. As for global negative feed back (GNFB) often times just called NFB, is taking some of the output of the amplifier right off the speaker terminals of the amplifier and feeding it back, out of phase, to the first stage of the amplifier. I have found over and over that more than about 16 dB of NFB can start to cause instabilities in the amplifier like motor-boating at a very low frequency. As for S/N ratio in dB, anything better than about -75 dB will likely not be heard coming out of the most efficient speakers with our ear right at the speaker. High quality amplifiers usually boast -90 dB S/N compared at their rated output. 90 dB is a ratio of one billion to one... so, if you have 100 watts of power and the noise is -90 dB down, the noise is 10 nano watts or 0.1 microwatts, however you want to look at it. At say, -60 dB down from 100 watts, the noise is 0.1 milli watts and you actually may be able to hear that out of high efficienty speakers. Remember that for every 10 dB, the power is changed by a factor of 10 so the difference between 90 and 60 dB is 30 dB or a power factor of 1000 times - louder or softer - depending if the power is increased or decreased. Hope this helps.

@@ElPasoTubeAmps OK, you are talking about power and not gain. Shouldn't you consider the difference as gain ? The level of detection is basically only compared in voltage or amplitude and not power. Attenuation in a volume control is never specified in power, only as a voltage divider. A spectrum analyzer only analyze voltage and amplitude not current or power. The NFB is also specified as a amplitude and voltage level and then this is valid: 20dB = Gain of 10 40dB = Gain of 100 60dB = Gain of 1000 Noise is also quantified in amplitude and voltage. So, as I see it every change of 20dB is a factor of 10 and not 100. 16dB of NFB is not critical in it self for oscillation. You just happend to run into phase problem at that level of negative feedback. Your feedback loop have a higher frequency pole than the first gain stage so the feedback loop that are supposed to come back 180 degree out of phase have a pole creating more phase shift and when hitting more than 90 degree phase shift it comes back more IN phase than OUT of phase and then you have a negative feedback loop that will make the first gain stage to add low frequency gain, and then it come into low frequency oscillation. Many transistor amplifiers can have NFB of 40dB or more. Just have control over low and high frequency poles of the first gain stage and you will be able to have more feedback without the risk of run into oscillation. Personally I do not like negative feedback and avoid it as much as possible. My own designs most often ends with a little bit of feedback like 3-9dB just to even out amplitude over the frequency response I try to have. To much global negative feedback kills and changes harmonics to much which make sound monotonic and with a collapsed soundstage. IMHO.

I agree the 16 dB of feedback is not any fixed critical level for oscillation but that seems to be a good limit for vacuum tube amplifiers. I think the amplifiers that have the 40 dB level of NFB are solid state without an output inductor. It is that OPT in tube amps that causes some instabilities in many cases and some also need RC circuits with a pole around 75KHz to keep parasitic suppressed. I see the RC circuits in different places, usually around the OPT or in the case of Williamson amplifiers at the plate of the first stage. I use a power meter to adjust NFB - usually setting the power to a clean level, say 5W, with no NFB, then adjusting NFB to a determined level, in my case of 16 dB of FB to about 130 mW. I do this with a pot and then measure the pot and install a fixed resistor. Converting those figures to voltage gives approximately 6.32 volts for 5W into 8 ohms and 1.02 volts for 130 mW. The results turn out the same. Lots of NFB does indeed give an amplifier a compressed sound when compared side-by-side with an amplifier with no NFB, but the gain in power can't be ignored when NFB is employed. A clean 5W amplifier with no NFB can (maybe) be turned into a very nice 30W amplifier with, in my case, around 16 dB of NFB. I have one stereo SET with 300B and as you mentioned, I have a switch for (I can't remember exactly) about 5-6 dB of NFB and the other position of the switch for no NFB. I like the sound of SET with no NFB but, as already stated, the power is often very low.

That's a different video... 🙂