KT88 Amplifier 80 Watts Williamson Design Using Triad S-152A Output Transformer

tonight is February the 4th 2020 I've got an amplifier here for you tonight with a discussion of a number of things about it I think you'll enjoy I hope you do you've seen this amplifier in a slightly different form posted a video on it a couple of months back where I had what's called a poseidon board right here it had a 12 ax7 to 12 au 7 and it had 2 1625 s over here and I don't know what I got out of it I don't remember it's posted and it got a lot of views and some comments so I think you guys enjoyed it quite a bit but it just didn't satisfy me so I took out that Poseidon board put it in aluminum plate and and this is this is a Williamson circuit right here driving a pair of kt88 now this one rocks this one will do over 80 watts I'll show you I did change the rectifier out right here I had to put a solid-state rectifier in it because they're a single 5u 4 or a 5 AR 4 they're just not rated high enough to run an 80 watt amplifier I mentioned several things about that in a video earlier that a single 5 u 4 is if you really want to be conservative about it is probably not good for an out for over about 30 watts output because the overall efficiency of the amplifier to be conservative again it's probably about 50% so that means that for 30 watts output you're wasting about 30 watts and heat so your 5 year for your rectifier whatever you're using 5 AR 4 et cetera is going to be having to deliver about 60 watts to the circuit now we know that a 5 AR 4 for example or even a 5 V 4 will have a whole lot less voltage dropped in solid-state rectifiers but unless you're just you know and I completely understand unless you're just completely enamored by a vacuum tube rectifiers are even mercury vapor rectifiers which I seem to be solid-state rectifiers really are the smart way to go the mercury vapor rectifiers are pretty exceptional you know the voltage drop across them is usually constant something like around 10 volts so if you really want to deal with them they can be pretty impressive they can do some very very impressive work you know of course with the addition of the heat the filaments supply etc but anyway just get to the point here this amplifier I am very pleased with I've been listening to it I have a bias balance right here and I have a like say a plug-in type v u4 if you want to call it that a substitute it's actually I can unplug it with power um that doesn't hurt anything it's actually called a a1 in 12:39 if you could see that in case you want to buy one one in 12:39 if you look for that like on eBay this is what you'll find it's rated at 3200 PIV etc anyway ok let's move over to some schematics and I want to show you where I'm going tonight with this number of you all have asked me for schematics I don't do a good job of drawing schematics like I used to but this is a really good one to start with this little heap kit this one is for the Williamson type amplifier w-4 a.m. you can find all this on the internet no no not necessarily input links and there it's just too easy to find this is a an excellent place to start in my opinion you know you don't have to build this exact amplifier component for component but what I do is I like I like this setup right here I don't I don't go to all this trouble right here that's not quite necessary you do need a resistor right here in series with the grid even if you put that into the wiper of a pot so that you have a a gain controller at the front end you do need a grid stopper right there not because yeah well because of stability that's exactly what you need because when you get the when you get to guide pot down at a very very low value maybe just a few ohms they can I honestly it can actually throw this tube into oscillation okay but if you look at the values from from here back here back then just about every Williamson out far you'll see they're all about the same 22 K here 47 K here 47 K there now there's some things in here that I usually don't use I usually don't use this little stabilizing circuit here 4.7 K and a 420 picofarad there's something in that range that's to stabilize the amplifier it just knocks the heck out of high frequency response so if you don't have instability issues and I think it's better to leave it off same thing I hear that see they use a Zobel network out here the point one and a forty seven home resistor across the transformer I don't use those either unless I have stability problems and if I have stability problems I try to figure out a better solution that again will knock the heck out of the high frequency response if you don't use their exact transformer you will have to tinker here with the feedback resistor you may or may not even want to use this capacitor you have to be very careful with that capacitor I've experimented a lot with that you want to get your voltages pretty much what they have in here you know with voltage dividers here with decoupling capacitors by the way I thought I don't think I've ever mentioned this before but I'm not sure everybody knows what a decoupling capacitor is for example let's look at this this is the plate load resistor for this tube we know that don't we and it comes down here and here's where it gets its DC and it goes through a voltage divider right here between this resistor and this resistor so that the voltage is correct but this capacitor off the ground why is that necessary well in case this is not fully understood there's two components we have to deal with in an amplifier one is the DC voltage and the other is the signal voltage now if you didn't have this capacitor here this resistor would be in a series with this resistor it also be a series for this one and weird things could start happening but you wouldn't have a 22 K ohm load resistor for the signals you would have a 22 K ohm resistor and a 22 K ohm resistor and you'd get the correct DC voltage here but once you start putting a signal in and driving it and the AC component is on it then this resistor would be in series with it but when you put this capacitor here it grounds this point right here for the signal voltage so the signal is developed across this resistor basically to ground at this point because a 20 micro farad capacitor off the ground is going to take all a signal component to ground so the this DC signal stays constant and the load resistor is only this one right here I hope that makes sense it's actually pretty obvious that you have to deal with the AC component and the DC component but anyway I generally build this circuit right here backwards I don't always but this is actually a quite a nice set up this is a balance this is the balanced of cathode currents you can also put one in right there like UT c does to actually set the cathode current to what you won't you could make that one variable so this one sets it and this one right here makes them equal and they've got their little jacks there too to measure it see this is the same thing when you put this capacitor right here thank all of you know that when you put a capacitor life from here to ground you've got a lot more gain out of it and that's because this the AC signal is grounded at that point the the output power will drop significantly if you leave this capacitor off because now the AC component coming in on the grids' and being on the cathode is having to work itself back through this resistor network to ground okay enough of that okay so when you build that ply like this you don't have to build every component in it exactly how the schematic is you notice I use one 5v for it's a pretty low powered a little amplifier oh don't worry they put these capacitors in series right here is just a raise of voltage okay so this out fire that we're looking at right now is basically built like this right here I'm using a fixed bias supply and it comes in to a 10k pot these hundred caves of these 47 case allows me to adjust the bias on it and balance it see this one's gonna go up and down so I'm balancing it now in mine and the one I just showed you we're gonna focus on it entirely tonight I've also got a bias adjust well this one does to see how the bias adjustment right there so you could set the bias and here's where you can balance the bias and the cathode currents I'll show you how to do that look at what they use they use to 5 V fours the plates are in parallel for a pair of 65 50s that will work but if you're gonna use anything over about 30 or 40 watts you're gonna have to have 2 5 u fours or - 5 V 4s or - 5 AR 4s unless you want to go solid-state alright you go see here's how they do it here they do the same bias balancing over here and in this case whether using 65 50s - they rate this amplifier 50 watts okay now the let's see the exact one here is here so here's a beautiful design right here just component for component this is the UTC w20 as you can see this is another Williamson circuit just drawn a little bit different but if you follow it out you'll see it's exactly the same thing and see and there's that resistor for setting the cathode current and there's a resistor right there for balancing it here's another thing if you're going to use parallel push-pull tube with like a quad right here like they use a quarter 6l sixes you really don't want to strap the plate to the plate directly with a piece of wire you don't want to strap that the screen to the screen with a piece of wire I don't know exactly why but you need to give the screens of the tubes and the plates of the tubes I guess for lack of any other way that I can explain it you just got to give them a little wiggle room there and all they do is put 47 ohm resistors here and hundred ohm resistors there and I found that that absolutely stabilized that quad 16:25 modulator that I built made a solid as a rock up until then it was squirrely so be very careful and a quad how you how you connect to the plates and the screens together hi that's my advice don't do it directly let's see I believe that's all of the the schematics we need to look at right now this is one I believe that uses hope goodness well that's okay let's go back to the amplifier before we get to totally until the aqua fire again I guess if you go out and you look for 100 amplifiers part 1 1916 to 1945 part 1 part 2 1945 to 54 part 3 55 to 59 and part 4 59 to 82 I find these schematics so valuable that I did exactly this I printed every one of them even in color they are so important to me I don't know if I can show you color photos in here or not but there's some really nice ones I think I think some of these oldest ones have some really nice color photos that's not what this is or you'll see there well right you got to have these they're free I mean it'll cost you the the ink or the the toner and the paper and the binders but and you don't have to have them in hard hard copy of course you can have them in the good file format but these things are good this Lillith all engineering yeah I'm gonna put these back I just brought them out here to show you show you to show them to you and how to find him you just look at 105 part one two three four and you'll find them all so that is such a valuable resource I didn't want to forget to tell you about it okay let's look at how the amplifier is set up I am using first of all these are their stamp Telefunken kt88 but all of you that are familiar with Ike if you know that these are actually JJ's but I have found that they're actually quite expensive to these Telefunken but today just like it was you know 75 years ago tubes all tubes are not made the same they may be made the same but they don't perform the same and I can say here that I've had these for a couple of years they they perform extremely well there they must be the very top of the line that JJ makes and Telefunken buys them and puts their label on them GE and RCA and Sylvania and Mullard and the whole nine yards of them did this 60 years ago you know they did it too they made I guess RCA said we're gonna make a million 12ax7 s and Muller it says well we'll make a million 1280 you service I don't know and they sold them to each other and put each other's name on them and then they graded them too so anyway I'm not trying to sell you these tubes I'm just tell you that they perform extremely well and not only that they have I want to show you the balance this out before I've been using it now for two three months and I want to show you the balance huh I'm actually shocked not shocked and who not only shocked no electrocution zero but I want to show you that the balance between these two I use a 1 ohm resistor in the cathode circuit that's a zoom this guy in right there often times you can see it better in the in the camera I can see that but point 0 67 that's 67 milliamps across a 1 ohm resistor I'm gonna move it to the other side and the other side is 67 point 0 that's even kind of shocking again no pun intended to be 67 point 0 you know they fluctuate slightly where I started from Wow no I promise you I'm moving it left to right so the balance is just I'm moving it from here here and back and forth so anyway that's that's the balance pot and the one in the back the bias it sets the the overall level so they're they're perfect you want on set about 70 or so milliamps I run the vias on my tubes actually quite hot I run them up at 80 or 90 percent of the maximum plate dissipation they run quite hot I'll show you how hot they are i meter over here I'm trying not to forget anything because there's some little things that that I really that I've learned and really enjoy see this tube can you read that it's about 300 and about 370 degrees 360 70 let's pull this off this relation both of them and since the current is just essentially exactly this time see this one is 360 70 that's the same I hope you can read those numbers it's reading about 300 and it fluctuates between 300 well there's sending them about 360 that's somewhere but a 60 and 80 and the same over there very very good I am extremely impressed with these tubes I don't have another set of them to anyway that's what that is now the city sky upon its uh on the side and take a look underneath and I'll show you what my thoughts are Lam okay well some of the things I want to point out in this amplifier that I think are important that I like to use and I'm using a non-conducting rod here I know some of you have reminded me that I have to be careful with that as I use it in my pencil because it is too alone this is I this is the way I put my capacitors in there I don't these are like 100 microfarad I think which is actually quite large you don't need to make these things monstrous five and six hundred a thousand microfarad and just isn't necessary but I put them in with just a little bit of goop on the bottom goop is a material that I've shown my plate before and it's just it's a miracle stuff for me that's what it looks like you could put a little bit on there and it'll hold just fantastic but you can pull it off anyway in series in parallel whatever you want to do same thing over here and this is a high voltage supply this is some decoupling and I'm not sure what oh yeah this is the bias over here here's the bias applied decoupling capacitors high-voltage capacitors I had a much smaller token here this is a 500 million amp one and a half Henry that changed things dramatically now as far as like the the component values as I said if you're going to build this circuit you want to use these types of voltage dividers but the one that you're gonna have to tinker with the most or the only one you're really going to think of is that one right there because your supply voltage may be slightly different right here if you want 450 in your end up with 500 that's actually okay but you might want to play with this one you might have you might see this is one and a half K one watt I put one in here this one ends up being I can see it five K for whatever reason that gave me the creek neat that gave me the correct voltages on this part of the circuit and that is fairly important as far as the biased supply here's the bias control right here there's the balance see there's those 100k pots and 400 K resistors and 47 K ohm resistors I use orange drops I love them here's feedback right here I do I am using a capacitor in this one you can put in some suggested value unless you've got some you know unless you've got the equipment to really deal with it you may be better off just leaving that capacitor out here that are putting it in pretty darn small 100 Pico farad's or so they suggested some some stabilizing stuff off the screens of the tubes so that's where we measure them I think that's it I'll just put a piece of aluminum in there to mount those tubes on I do not that the success in sevens this is the Williamson driver feeding us a different my design bias circuit so that I can feed a negative voltage to the to the kt88 I use a small capacitor right here on the output of the buy supply that actually feeds the grids and one time I put a big one in there because I wanted to make sure my bias voltage was very very very clean and when I say a big what I'm talking about 100 microfarads but I found out that by is monitoring and it takes it like forever to build up because you got some series resistors in there I should think there was a good idea so I think this one's like 1 micro farad doesn't take much that's pretty much yet underneath not too exciting I don't suppose let's let's run some some power tests on it I do want to show you something here before we get too far away because this might be a reasonable question to ask what do I set the plate dissipation for well here's our plate voltage let me see let me zoom in on its RAF to do all this but 577 volts and if you remember 577 577 and we were running at 67 million at point zero six seven that's thirty eight point six watts absolute maximum power rating excellent maximum absolute plate dissipation is 42 watts and if I divide that by 42 I'm running it at 92 percent please that point zero down to so I do run them hot that's just my choice I'm not saying that's how you should do it you will get you probably for sure will get better performance running them really hot will it shorten the life of the tube oh I suppose so but I'm not too worried about it I'd rather have performance out of it and replace the tubes there once in a while then have lower performance and I have them last forever although quite honestly I don't think we're going to hear the difference but anyway that's what it is so then 577 volts is not terribly high for 4k t-80u 6550 thought I'd throw that in there before I forgot it okay just to start the test here we'll start it out at the rating of the transform 65 watts there's the 65 watt right there that's an average power point seven percent THD here's its voltage across 8 ohms because where they divided by eight and you'll get this number down here there is a kilohertz so there's what it looks like that's a 65 watts at one kilohertz so it does actually quite well absolutely back to the power that I call my absolute maximum power at 1% THD we'll watch this guy right up here can we get it about 1% and we'll also see some nice symmetrical clipping in there to everyone point there it is there's trying to split hair but there you go that's pretty 181 watts so it does a good 80 watts I don't see any clipping in there I can see the tiniest tiniest amount we're starting the clip if I turn it up a little bit more you can see the clipping there the clipping is a nice and symmetrical at 87 but that's up to 4% so it should be quite a rock and roll guy here okay so Lew 80 watts before any clipping occurs so it seems to be performing quite well to me that speaks extremely well for the Williamson driver so you know just make sure that you don't have rectifiers where the voltage sags terribly because it's gonna work great at 30 40 maybe even 50 Watts probably does 40 Watts below it's probably going to perform pretty well but once you start getting above about 40 Watts it's gonna start performing pretty bad you're gonna go well what's the problem well your voltage is sagging - pretty bad it's gonna drop it's gonna drop because of well I've got the line voltage just stabilized at 120 volts input so our line voltage coming in is not dropping but just causing internal resistance of the transformer it's going to start dropping some even though that's a pretty hefty transformer over there it doesn't matter how big it just gonna drop some so you see what I mean solid-state rectifiers will give you a lot better performance let's do some frequency scans of it and then we'll let it go okay now we're going to do some scans on it I'm gonna show this for my any newcomers I might have has never seen this equipment this is an HP 89 o3 audio analyzer and it'll do scans it has an HP IB interface on the back with a with another interface connector on it that interfaces it to a USB port on a computer it's all controlled by the computer and it's down here in this little PC and what we're gonna do is run to teach the scan on it so I'm gonna start that up and they then ask us right here we start setting the parameter I'm gonna set it we're gonna start at 30 Hertz we'll go to 20 kilohertz we'll go to 20 kilohertz 32:23 kilohertz five points per decade that gives us a decent smooth sea get something clear out there there you go and then we go set the generator level to 2 volts now even though it's set to 2 volts we're not going to be pumping 2 volts into it because we're gonna let me let me turn out another lie here all the glare off we can and then what I'm gonna do even though I've got the generator level set to 2 volts again 30 Hertz to 20 kilohertz 5 points per decade and then we say down here we say start and then it says okay it says reference level okay he wants to know what our output voltage is we're gonna have to look back over here and what I've done is I've adjusted the level right here and the level of the input as you can see this cable going up here the output of the 80 903 and I've got it set for 50 watts it's dropped ever so slightly make sure real full 50 and the reference level has killed hers so we say yep everything's fine and we can monitor it right here I want you to see it there's our kilohertz and then we say okay now what twice don't start do it it's gonna start out 30 Hertz there's our line right there starting at 30 Hertz it's about point about point eight five percent THD so this is the 1% line right here so it's staying well below it's 10 below 1% we're running at 50 watts remember that's basically what the designs are set for it is about 50 watts at least that's the one that I have built it by with that in consideration I'm also very conservative about amplifiers if I say it will do 50 I expect it to do at least about 70 that goes up a bit here at the end I'm trying to stop it at 15 kilohertz but these numbers don't exactly this stop frequency doesn't exactly stop it at that so we just got a deal with what we what we can get we can change it around but anyway it's about one percent until we get out to about 8 kilohertz and then it starts to 1% then starts going up so 15 kilohertz would be there's 10 there's 18 so 15 kilohertz would be around around there be around about two and a half percent THD at the high ends now you gonna say wall what about that transformer that transformer up Darius says he says right there seven Hertz the 50 kilohertz can you see that power level 65 was what good luck you doesn't work that way and there's no other design that's gonna make it work that way either now is this gonna sound bad not at all I think it sounds marvelous but that's that is the way it performs and that is that's the way it is if 50 was sounds great got a lot of power to it I mean it really I listened to it I hearing the little paradigms they're called focus it's a little bit dark in here if you can see it I got one on the right side and I just got a parallel and one on the left side over here and rock and roll in here so that is a Williamson design into a pair of kt88 built about as good as I can do it into quite a nice transformer the only transformers that I have and that I've seen in the built with that actually do 20 to 20 kilohertz below a half percent is either one a Macintosh transformer which does exceptionally well are the a crow sound transformers are the UTC transformers well actually the James transformers perform quite well - I've built amplifiers with all of those I've posted videos on all of those and that is my experience those are the ones other than that I generally think of a output transformer is a 30 to 15 kilohertz transformer and now if you actually look at the real specs on some of the amplifiers that we have a lot of respect for like the Dyna codes here actually rated 30 Hertz to a 15 kilohertz so there you go I'm not putting him down I like him I think they're actually quite quite good devices I want to run one more test for you though okay well we have here is this is the box this is where the output from these resistors by the way this uh this amplifier down here it's uh wire right here this cable right here goes up to these dummy loads these are precision resistors goodness it's good windy out there opening my door these are precision resistors let me see you might I got something on the front of this thing there we go 1% non-inductive anyway they do a great job I can switch between left and right channel but we just usually get on one channel right now and it all goes over to this breakout box this breakout box is what feeds everything so what I've got set up right now is basically the same thing there's our sine wave there's our output power I'll drop it down to 50 so that we can stay about the same okay add a kilohertz or there abouts I set two kilohertz up here so I'm not you know to have to be exactly 1,000 point zero there's it's THD right now sorry for the glare because it's th the again and what we're running over here is a program called spectra plus I'm gonna hit run there it is yeah it's running you can you can you can see everything wiggling I'm gonna turn out some more lights oh but I'll put us in to bunch of dark I think you can see it running there's our fundamental one kilohertz there's two kilohertz three four five I've got a lot of harmonics in it vacuum tube amplifiers are gonna have plenty of harmonics the second harmonic this is a 10 DB per 10 20 30 40 all about 43 DB down or so that's quit running but at least we still got the displace table that's a harmonic profile of that amplifier sorry for that glare it's got a terrible amount of glare on it anyway I can't do the video over again here think just take me too much I can only apologize for it now one of the things that's really nice about this program is see the 60 Hertz over here and see there's 60 right there if you've got like a hum balance pot in it which I do you could you could tweak that down up and down you can see a real-time occurring almost started running again yeah so you can see the lines wiggling it only runs a few seconds and then it causes and freezes but you can you can adjust the hum balance right here see the hum this is a about minus 100 and this one up here is about minus 20 so there's about 80 DB difference between here and there so you could not quite 80 76 or 77 you might actually want you could estimate that as the signal-to-noise ratio because that's about what it is the fundamental frequency - it's a 60 Hertz component down here so that's a marvelous little program to have along with this one over here I've made a lot of videos on this stuff this one this program is made by a good gentleman by the name of Pete millet you can go to Pete millet calm and find all this stuff it's not easy you can figure this this audio analyzer up here to work with it but it's worth the trouble in my opinion so there we go missed you're running it at 50 watts I'll tell you what let's do it let's just for the heck of it see see just trying to power up a little bit okay there it's running by the way here's THD see about 0.7 which agrees with our Tektronix and our hv stuff let's see what happens if we run it up to about maybe 75 watts let's don't go over the edge here 75 watts analyst look at it you start seeing a lot of spurs in here my ears hear all that that doesn't mean that we don't like it I'm actually convinced that but for some reason or another that I can't quite explain we seem to like a lot of the artifacts of of a vacuum tube amplifier I mean it's it's got the full range out there if you look at a really really high quality on before like again a mac app this will be quite a bit lower it likes to be lower than the third this right here a triangular wave has a descending set of harmonics right here a harmonic profile if you look at a single-ended amplifier that's exactly what you'll see now it'll look like a sine wave on an oscilloscope but it's practically a triangular wave the second harmonic if this was an SC triode SCT single any trial amplifier this was the second harmonic could be up here the third to be there at fourth fifth six seven you could just you know bosz draw a straight line right across the tops of them I don't build se up far as I'm not saying you're bad to each their own but that's it right there so there you get the harmonic profile here you get the frequency sweeps now these things will do more than that but that's the equipment that it's always fun to play with power meter you can also make the hewlett-packard Reed power you can say clear nineteen point zero eight special well no this thing's got control of it clear nineteen point zero eight special and there's and then it reads power C it says it's seventy four point seven seventy four point seven this is a jewel to have if you could get one of those in good shape and you'd like to do audio work you'll love it but see there's it's THD you got a kilohertz it's only point seven eight percent it 75 watts I'm actually quite pleased with it like I say it you know I mentioned a while ago how the the a croissant transformer is a James transformer the UTC transformers performed really well but pretty much pretty much that maximum power is about 20 watts you could make an amplifier for 20 watts whether Williamson design I think that's about as good as has ever existed on earth well when you start getting higher than that things get a little bit more complicated so there you go ladies and gentlemen I hope you enjoy this this is the first audio video I've made in the audio equipment video made quite a while I want to show you something very quick this is one of my next projects I got in terms of lights back up this is a push-pull amplifier with plug-in coils from 1948 that I hope to have one am here pretty soon I've been spending a lot of time it's just so darn dark in here I'm not quite sure why but there you go this one was made by James milling down here as you can see I've got a plate current meter at a grid current meter uses a pair of not eighth eighth 11th but hmm I can't remember what they're called anyway they look like ADA levels but they're they're a lower game Tube uses link coupling this is how you a couple a year and tell them this is the tuned circuit here and then you cut me into this way I've also been into my strana me a lot lately but that's for another video well thanks for watching stay safe

Channel: ElPaso TubeAmps

Views: 6,032

Rating: 4.8969073 out of 5

Keywords: Triad, output transformer, opt, vacuum tube, 807, 1625, audio, guitar, music, power supply, 8t88, kt88, 6l6, 5u4, stereo, williamson amplifier

Id: OwA3hkzBPaQ

Length: 40min 40sec (2440 seconds)

Published: Tue Feb 04 2020