Bias Voltage on Vacuum Tube Amplifier vs Head Room & Power Output 

Maybe if I show people that I can build an amp it will inspire them... This type of electronics is a dying art.... We need to get the younger generation into this stuff before its too late.... By we I mean more qualified than myself.... But I wonder with all the tubes and transformers I have, Can I buy a board and go almost plug and play with some soldering and minimal gear?????

There seems to be a trade-off between performance and life in many things but I am not through with this yet. I think the amount of head-room will always be somewhat subjective and I may not be able to quantify it with test bench measurements but i am going to try again with a peak reading watt meter while visually observing a dynamic (music) waveform on an oscilloscope. We might be able to see some difference there but I am not sure yet. If there is a difference I suspect it will be small, I am guessing 10% ?? and that is going to be difficult if not down right subjective as the top end of watt meters become numerically compressed and small differences in power become more difficult to read. But it still seems like fun.

@@ElPasoTubeAmps aye, it does. To my non-technician ears with regards to guitar amplifiers, there is definitely a noticeable difference in the headroom when the hot/cold bias has been A/B'd in some of my amps. However, 2 minutes later after my ears have adjusted and the settings on the amp compensated, there's nothing concrete, to me, that shows a significant difference when playing an amp pushed into soft distortion. On cleaner settings, perhaps the same with a single session of playing. Please keep the great content coming and thanks for responding. Best, MM

I think your explanation is great and I can see peak powers as high as you are talking about, percentage wise, from an average power condition. I suppose a test like I ran tonight could be repeated with pulse driven input and these higher "head room" levels could be attained. A similar power output level occurs in SSB amplifiers where an average power may reach 750 watts but PEP power is 950 or even 1000 watts. I very much appreciate your thoughts on this and your posting.

@@ElPasoTubeAmps I can see that with SSB as there is no carrirer signal it's very much like audio. First time I realised this in audio was few years ago a local guitarist brought in his Fender Deluxe with a complaint that "it had no life" in it. I noticed that someone set the bias much too hot. The owner couldn't believe the difference the adjustment of bias made to the sounnd and feel. His playing style is very dynamic and relatively clean, kind of on the verge of clipping. By the way, great video as always. Keep them coming.

@@markg1051 Good analogy between SSB and the audio. Makes the PEP meter seem like the right idea with possibly just an oscilloscope to make some subjective observations of clipping/distortion level. I like it. Got any simple circuits for PEP audio level circuits? Has to be adjustable so that average and PEP calibrate the same for steady state and then possibly an adjustable 'hang time". I don't want to have to figure this circuit out - but I will put one together that someone has already designed. It doesn't even have to be a vacuum tube circuit - SS will be OK... :-) :-)

@@ElPasoTubeAmps: It's been few years since I've had anything to do with RF and paricularly SSB measurements. Istill remember and understand the priciples of it though. Not sure if I understand you correctly, are you after an audio frequency peak level detector or one that works at RF? If AF, I should be able to provide you with some proven working ccts. Will have to do some digging. I suppose if you have a good quality detected AF signal from the RF envelope it will show up any peak limiting or clipping and you should be able to use an AF peak detector and calibrate it against a scope output of the RF envelope.

@@markg1051 I built a LED volt meter a long time ago that I use for PEP RF and it works great. Essential the LED "wattmeter" is simply a LED voltmeter calibrated in watts to a particular load level. I use a diode detector parallel with the coax line. I think that is what I am going to use for PEP level readings to measure the voltage across the 8 ohm load and calibrate the LED's in watts instead of volts like I did in the other meter. Watching the PEP output on the LED meter while observing some level of consistent clipping on an oscilloscope to indicate a maximum level I think will give us a reasonable indication if there is more dynamic power at some distortion level

Each of you guys have described how the steady-state condition does not represent how an amplifier will respond to transient conditions. I am going to see if I can reliably reproduce and measure/document a PEP type measurement. Maybe a couple of tones from oscillators pulsed by a third generator. THD as read on a meter will likely be useless but possibly some level of clipping might represent a maximum. In the end, the results will be subjective but hopefully we can come a little closer to an analysis of head-room vs (fixed) bias settings. As mentioned in another comment, cathode bias vs fixed bias would surely effect a completely different harmonic profile as the bias is dynamically increasing with increasing drive and cathode current.

May be this helps: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-IPqpmbiRlAc.html i did this as a little "how amps work" series, as i wanted to show how an amp behaves at different operating points and topology change ;) Also is you got an all tube amp with a tube rectifier, the overall current consumption also alters the "tone" as the voltage drop within the rectifier is a function of the current. Furtheremore the screen grid voltage is essential for the operating point of the output stage (assuming you´re running pentodes). and if that dropps only by about 25% you´re dealing in fact with a different amp...see the datasheets, to verify this. As in the real world all components remains the same, the Amp will "answer" with more distortion. And we can go a little bit further: may be you know that it is possible and also widely used in RF-stuff.....:you can modulate the screen grid voltage to alter the conductance of a pentode tube. for ex. the EF83 if spefically made for that. Some audio compressors and much more RF-transmitters use this method of varying the screen grid voltage to alter the output power/voltage without touching the input voltage.

It sure is good to hear from you, Sir. I hope you are doing well. Your input and experience is always very valuable and appreciated. I am thinking about an improved method of measuring "head room" by building an analog type PEP meter. Something like the Bird 43 RF watt meter circuit and possibly just playing some heavy-metal music as the input source. I can't play the actual music over YT or it will be banned from some countries but that might be more of a realistic source of signal than a sine wave. I got out my GR sound meter today and, for the first time ever testing this way, stood 9 feet from each speaker which are placed 15 feet apart. So, I was at the apex of a triangle 9 x 15 x 9 (feet). Playing some Def Leppard at 100 dB (average) was certainly loud but tolerable, raising it to 110 dB started getting serious and then up to a sustained level of 114 dB with many spikes going to 118 dB and even a few to 120 dB became serious. My mind and body start to react with slight tinges of discomfort, for lack of a better word. Maybe "concern" for my hearing and just a feeling of loosing the sense of well-being. Afterwards, my ears felt a little "dull" for a while. This was driving the Klipschorns with a McIntosh 2100, C20 preamp volume control at about 3 O'clock. I definitely had headroom above 114 dB... the meter has two more scales that goes to 130 dB where the meter would max at 140 dB. For the first time I think I can truly conceive of the damage that would occur at 130 dB. Stay safe...

Assuming you are thinking about a HiFi stereo type amplifier and not a guitar amplifier... I recommend some version of a Williamson design. UTC catalogs (particularly the UTC W-20 design) performs exceptionally well. dalmura.com.au/static/UTC%20W-10%20and%20W-20.pdf Two of the designs by Heathkit, the W5M with link here... www.mcmlv.org/Archive/HiFi/HeathkitW5M.pdf and the W3 design www.mcmlv.org/Archive/HiFi/HeathkitW3M.pdf Vintage output transformers like the UTC LS-57 and LS-61 are very good. Just as good are Acrosound transformers and the newer James transformers for output stage. Also see these designs... lilienthalengineering.com/100-amplifiers-chapter-1/100-amplifiers-part-2-1945-54 There are four books of amplifiers to be had for free Part 1 thru 4. Just search for them. Study and stick to a proven design. No use to reinvent the wheel when it comes to vacuum tube amplifiers. The challenge is learning the nuances of each circuit and not the reinvention of the vacuum tube amplifier. Other high-performing amplifiers to study are the McIntosh designs. They require a McIntosh output transformer that you are probably not going to find to best to just buy their amplifiers and not try to build them. Hope this helps.

I saw the bias voltage being more negative (running the output tubes "colder") similar to how you are describing it but I have measured that a few volts difference in bias voltage at the output tubes is easily overcome by just a couple of millivolts more input drive voltage, so from that perspective, I can't see much, if anything I can measure, in what we might call extra headroom by making the bias voltage more negative which, if I understand the idea correctly, is the common belief. Of course I could be wrong because what we think makes sense isn't always how things work out. As for the clipping of the output from the phase inverter, certainly that would change things in a way that is going to make it even more difficult to measure. The imbalance to each output tube grid from the phase inverter could cause one side to clip before the other. It becomes daunting to try and keep all these parameters stable and measure them in a way that can be repeated. I appreciate you bringing up the phase inverter as that is something I need to take a look at and see if it can provide more drive than is necessary or if it is actually (one of the...) culprit(s) that can make this measurement process more difficult or even impossible. Lastly, I ultimately respect the opinion of the musician over what we may or may not be able to measure because no matter how good an amplifier is on the test bench, the ultimate question is, "how does it sound", completely subjective. One other thought in this process is, in this case, we are talking about amplifiers with fixed-bias and not cathode bias. Think about the challenge of assessing "head room" with a cathode biased resistor/capacitor combination where the bias is modulated by the input signal. Except for the fact that fixed-bias amplifiers can provide much more raw power than a typical cathode biased amplifier, the cathode biased amplifier is much more harmonically-active or harmonically-rich than a fixed bias amplifier and for that reason, may be more desirable to a creative musician. Think back on some of the cherished amplifiers of decades ago that are SE amplifiers (running a single 6V6, for example) and cathode biased, that from a harmonic profile basically puts out a triangular wave rich is all harmonics, odd and even, in descending amplitude out to the 12th harmonic, or beyond. Just thinking out-loud as I am not a musician and may be all wrong.

Bartosz Pakulski I have often wondered if power amp distortion is a myth, and all you’re hearing is the phase inverter clipping first when the amp is turned up loud? Although you will get crossover distortion eventually which a single ended PI tube can’t supply

@@voxpathfinder15r It is pretty amazing how bad the phase inverter in most guitar amplifiers is distorted and yet the global feedback takes care of so much of it and much of it is cleaned up at the output. The distortion in a guitar amplifier is from the first stage to the last stage - and somehow we love it - but it sure isn't a high-fidelity type amplifier or signal. Then to top is all off, we spend $$$$ for high linearity stereo amplifiers to accurately reproduce the sound we created in the guitar amplifier. I have found that most guitar amplifiers only produce their rated output power at about 10% THD and that is with some clipping of the output. Just the way it works... :-) Thanks for your comments.

I think that is an excellent point. Cathode bias vs fix bias will change the dynamics of the output section of an amplifier considerably. The harder the amplifier section is driven, the more current flows and the more bias is developed. I would think at extreme driving conditions, the harmonic profile and sound would be completely different from a fix biased amplifier section.

@@ElPasoTubeAmps and if you look closely to that resistor in the cathode path...it also alters the feedback, just run some numbers for most extreme points and you´ll get the idea....that resistor shifts the cathode potential *up* when the tube is more conductive up to a point where it´s almost the same as the anode, if the cathode resistor is high enough...and if you´re using just one for all output tubes (like the VOX AC30 does) it also change the bias point of the other half tube section, which is not doing anything otherwhise ;)

Now that you mention adjusting for lowest level of crossover distortion, without exceeding some level of plate dissipation, that may be the simplest single-point adjustment to make for best dynamic range and head-room. I think that is marvelous and simple approach everyone can understand.

I have never thought about using a LED in a cathode bias circuit. It means there would be no current flow until the forward voltage drop of the LED is reached. Why would we want to do that? There may be a good reason - I just don't know. Here is something I found: Typically, the forward voltage of an LED is between 1.8 and 3.3 volts. It varies by the color of the LED. A red LED typically drops around 1.7 to 2.0 volts, but since both voltage drop and light frequency increase with band gap, a blue LED may drop around 3 to 3.3 volts. From my limited experience with LEDs, when I have used them I usually calculate the series resistor to give about 10 mA so it would be simply ohms law for whatever source voltage you have. In the case of a cathode circuit I don't know where I would go with it.

I don't have any guitar amplifiers around anymore. I did that for a while and enjoyed it but it got to a point where I couldn't get into my shop for other people's amplifiers and I had no time for my own projects so I quit working on guitar amplifiers for the most part. I think 1% THD in a guitar amplifier is probably hard to get - maybe at the lowest volume settings you might get it down that low. The amplifier I am using, Dynaco Mark III, is a good high fidelity amplifier up to 40 watts. They rate it at 60 watts but that is often times not attainable at hifi stereo quality distortion levels. I agree that these measurements would have best been done on a good guitar amplifier rather than a hifi type amp. Also, I think you guys that are musicians are going to be able to tell more about "headroom" than I am going to be able to analytically measure but it is fun trying.

ElPaso TubeAmps you had too much work with guitar amplifiers? That’s hilarious, how much were you charging?

ElPaso TubeAmps I read in some old Marshall literature of mid 70s that their JMP superlead 100 watt amplifier was rated at about 100 watts at 1% distortion into a 4 ohm load , 115-117 watts at 4% distortion into 4 ohm load and 170 watts at 10% into a 4 ohm load. I would love to get my hands on of those to verify those numbers

@@voxpathfinder15r Realistically, it is going to take at least four KT88/6550 in parallel-push-pull to deliver even the 100 watts at 1% THD. I have never found it a real problem but all of these manufacturers exaggerate output power levels with the exception of McIntosh. I know absolutely no one other than McIntosh that publishes true average power output so conservative that their amplifiers will usually deliver 20% more than advertised at the rated distortion level.

@@voxpathfinder15r Definitely not enough.

Yep... I know what you mean. I suppose that means I have about 104 watts of head-room in my home theater system... :-)

I understand what you mean and from an engineer's perspective I can see it that way but in the creation of music vs playback/reproduction of music, the dynamics of the amplifier and sound system certainly seems to be two different animals. We all know that when vacuum tube amplifiers clip, the harmonics generated are very different from a SS amplifier. Tubes "round" off the tops of the clipping whereas SS amps have sharp/abrupt clipping characteristics and our ears just don't like those sharp edges of clipping. Over the years and thru many discussions here on YT, it has become obvious to me that we actually like the artifacts introduced into the music whether it be the creation of music or the playback of music by vacuum tube amplifiers vs SS amplifiers. In a nutshell, I think the description of the sound of a vacuum tube amplifier over driven vs a SS amplifier over driven is the vacuum tube amplifier has a much "softer" and pleasing sound. So, even though we may indeed be running out of transformer, it is part of the creative process and I am not sure it can be effectively measured and documented to the satisfaction of a technical point of view. But it is fun trying... :-)

Hi John, I agree that 60% max plate dissipation is a good point. Possibly a little hot for some people that want their tubes to last years but likely a very good spot for good performance.

Probably as long as you dialed out crossover distortion, you’re good. Although I have seen lots of Marshall amps with a good amount of crossover distortion and I guess that’s part of the sound? That extra sizzle bacon fat when the power amp section is turned up

The idea was to measure the amount of input voltage increase (the "headroom" - such as, the amount that the guitar could be played harder) that it would take to get the same (or better/lesser) amount of output when the bias level is more or less negative. As I have had it explained to me, more bias (more negative - colder tubes - less cathode current) will allow for more head-room before objectionable clipping/distortion sets in. The first measurement attempts were at low levels of THD but I changed my thoughts in the middle and decided 10% might be a good spot because that is where most guitar amplifiers produce their rated output. As for the input voltage change, you can see a mV or so change in the unloaded/loaded state of the oscillator but that is not significant as it became the same under load. The other day I did measure a slight amount of input voltage increase needed to drive the amplifier to the same level of power and THD with a more negative bias. Seems to make sense but even then the voltage increase was something like 24 mV and I am not sure other parameters were not affecting that - tubes getting hotter, possibly my line voltage dropped/rose 2 volts and i didn't notice. I think we can safely say, the steady state, average power measurement using a sine wave is not going to be effective here. Just for fun, note my comment to TubiCal above on the sound level meter and my experience with "head room". I am beginning to believe a music driven/dynamic input level might show some form of head-room much better than average power measurement. I think I have to devise a way to measure that consistently before I can say that I can measure a difference in bias levels that would affect that head-room. I am thinking of a PEP meter like the Bird model 43 RF watt meter. Any suggestions for simple audio level PEP meters? I may have a perfect one right in front of me and not be able to see it...happens to me all the time.

ElPaso TubeAmps Bob Cordell has something of the sort: www.cordellaudio.com/instrumentation/power_level_meter.shtml. I get that colder running gives you more positive headroom, because of more-negative bias, so you're further away from saturation, but surely it also gives you less negative headroom as you get down closer to cutoff? A hifi point of view might be that you should bias for equal clipping plus and minus. I think the real reason for more headroom with colder running is the increased HT, as another poster has noted here, which, along with the reduced current drain, has secondary effects like keeping the the PSU capacitor discharge troughs shallower, which in turn considerably increases the instantaneous current available. In tape recording and your VU meter example, 0dB would be just under 1% THD, which is about as good as you can get from tape, and +3dB corresponds to 3% THD, the onset of visible clipping, and when you peg the VU meter hard you can hear the crunch on the tape. The idea is to record as hot as you can without pegging and without long excursions into the red zone, which is your head room. I'm not sure that the guitar guys see it the same way, and I was one myself for a while (but not a distortion fan). I think their idea may be to *reduce* the headroom ;-)

@@EJP286CRSKW I think you summed it up perfectly in your first paragraph and indeed, others have said similar things - I have found even 4-5 volts difference on the AC input line can change output power and THD considerably. And if a fellow wants a way out of this, and I am not looking for that, the bottom line is, head-room is just a concept. Loud is loud and louder is louder and after a while whatever "headroom" there was, there isn't anymore. It could be like grasping at smoke and wondering what happened to it once we grabbed it... surely there must be more smoke - or not. Anyway, I was toying around today watching the peaks start to clip on an oscilloscope (playing music out of the FM tuner) and determined the Dynaco "peaked out" at about 56 watts average. This is 60 Vpp (21.18 volts RMS) or 30 Vp which would be 112.5 watts peak. So, would the head-room be the power between average and peak? If so, I have a factor of two for head-room power. Maybe that is the way to look at it and make it a mathematical fact (twice the power, 3 dB head-room). In SSB operation, some people will say that the PEP power should not exceed the average power. That simply isn't true and indeed, there is "headroom" in a SSB amplifier. I load my 833A amp to 750-800 watts output (depending on line voltage...) and modulate it on a Bird model 43 peak-reading meter to 950 watts. This is a fact and is a well known practice. The Collins 30L-1 amplifier is rated at 1000 watts PEP (input) but it will only load to an average of about 700 watts. What we also may be dealing with here in audio vs RF is audio amplifiers are basically class-A (or AB) amplifiers where driving the grid positive results in terrible distortion and little to no more output. In a SSB linear amplifier it is ruining class-B and grid current is always there - so, a little bit more grid current... bring it on... to a point of course. Your thoughts?

Totally agree with all of you guys. I did consider the fact that this is a HiFi amp and figured it was not the best candidate for this test. i suppose we could call the guitar amplifier a full-bodied, creative amplifier whereas the Dynaco use in this video is a play-back HiFi amp and just doesn't have the same characteristics.

@@ElPasoTubeAmps usually guitar amps use less up to no output over all feedback loop, leaving the output stage running free, while HiFi amps usually have an over all feedback loop with some aditional interstage or stage local feedback loop in order to keep the THD down.

@@ElPasoTubeAmps Also. I know your channel and videos are mostly scientific measurements, but we have to consider ultimately how the bias adjustments change the sound. Not all can be seen through measurements " at least with the measuring equipment we have now" Please keep the videos coming, you are one of the few people out there actually measuring tube amps. Most other people out there like to talk but show no proof. Thanks