Step 7: The Heater Power Supply(s)

Unfortunately, the filament secondary for my power transformer isn't a separate winding, and doesn't have a center tap. Maybe I could disassemble the trannie and see if the coils could be separated...but it's a "potted" transformer (dipped in resin), and I didn't want to ruin it.

The trannie also powered about 12 tubes, and the filament voltage is ~7v, and doesn't drop enough under load to get near 6.3V (load isn't big enough.) In fact, one 12AX7 "went nuclear" and burned out (~$25 "down the tubes".)

Placing two large diodes in parallel but opposite directions in an AC supply limits the voltage by the voltage-drop amount (.5 to .7V), just the same as one diode in a DC supply. That dropped the filament voltage right to 6.3V, and the tubes were happy.

The two-diode trick works only for AC--current flows through one diode at a time, dropping that half of the waveform by the diode's voltage drop amount. One diode would do the trick for DC.

Plan B

However, they weren't quiet. You really need separate windings to setup a false center tap, which can be used to quiet the heaters.

After trying various solutions, I decided to light just the power tubes with the main transformer, and use a cheap "wallwart" for the 12AX7 preamp. Now the preamp has it's own, "dedicated," DC supply This was very quiet, indeed. The wallwart was already on hand.

A ground reference (false center tap) was provided by bridging the 6V by a pair of 180 ohm resistors, tied to the chassis ground. It does make a difference.

For some power transformers, the required 2.7A @ 6.3V is a little much. Many are rated for 2.5A max. Of course, an extra 200 MA might be fine, depending on the transformer. But a separate DC supply for the preamp isn't a bad option.

OK, this is a little, ehem, unconventional, perhaps even ghetto. But it works well.
<p>I have a 1967 gjibson ga 30rvt guitar amp cabinet and speakers. Can I put a 1966 gibson ga 45rvt in it with those speakers?</p>
<p>Don't quite get your question? Are you referring to the 8 in. Weber speakers I'm using here?</p>
<p>I got 2 Greenlee punches (one for preamp tubes and one for power tubes) for about $25 each on Ebay. If I want, I could build a couple of amps with them and resell them for what I got them for. </p>
<p>Cool. I did buy a set of Harbor Freight punches, but haven't used them yet...still gotta drill a pretty large hole to start the punches, so still need a step drill-bit. The HB set probably won't hold up like your Greenlees...</p>
<p>I like your style. I read alot more than I understood, but nice job. And nice sound too. I've been working on alot of ideas of my own the past few weeks. I have noted the trend on the lower watt tube amps, and the high dollars they expect. bleh... gimme a junk pile and a few weeks on the internet. ;) well done. cab reminds me of an old gibson blended with a marshall cutout for the controls. tastey.</p>
<p>Thanks man :-)</p><p>I'd like to rebuild this with a different power transformer; clean it up, update it. Most of these low-wattage amps nowadays don't have enough headroom for me...</p>
I am on a mission myself to try to put together something truly affordable and somewhat easy to assemble, yet usable as a guitar amp.<br> The cost of tube amp kits and projects are just too insane for my taste right now, plus, Many of us have seen this before. They come in vogue, price goes thru the roof and then the bottom drops out when players start to see the cost and limitations of tube amps. Just bring a nice hot one out of a warm building into the freezing winter air up north once.<br> Anyhoo. solid state it is. But there really isnt anything out there put into a package a person could one stop shop and solder together, at least not at a realistic price. like you said, low wattage is pretty common among what is out there.<br> at this point, I am looking at a 60 watt post amp board preassembled and coupling it with some tone pre-amps of my own making or of the effects pedal variety reboxed into a classic style cabinet. alot to learn here for me, I'm no engineer, but i do have a concept and the interweb.<br> So far I have soldered up a tonemender board, using scavenged parts from old power supplies, modems, whatever was around, and had a few 4558 ic's in my stash along with a pack of new resistors. yet to get my pots and hook it up and to be honest, the layout was just too tight for my liking, and suspect i may have overheated and probably have some solder bleeds. so wish me luck there.<br> So jump in this and get us 100 watts of highly affordable amp project with pcbs for sale and a bag o parts.(less transformers and chassis and cabinets, or make those extras)<br> or is it just impossible to source out enough parts to build a good amp for less than a good used amp?
<p>why not start out with the infamous fender bassman aa864??? it has both a guitar channel and a bass channel and you can mod the holy living merciful Jesus out of it.. change the bass channel to a marshal guitar one, add tone stacks... modify the existing ones.... the good part of the schematic is that the power amplification stage is so isolated from the preamp stage that there is so much room for personal modifications </p>
<p>that would actually be a more than good starting place, isnt that where James Marshall started :)</p>
hahaha exactly and since im working on one myself if u have any questions please feel free to ask! and on another note.... anyone have any experience on the Dual CS 741Q LP player?
<p>That's a good question...might be equally applicable to tube amps as well as SS. At this point, I've fixed several multi-channel tube amps; amps that were much cheaper to buy busted (and fix) than it would cost to purchase the parts alone. </p><p>For SS amps, I just gave my young cousin an old Harmony B3500 amp. That amp (working) with the cabinet (not working) only cost me $22 USD. And I recently bought an working PA (100 A) with reverb and EQ that sounds pretty good with guitar and stompboxes. For $50.</p><p>I can't turn my back on DIY, but unless <em>some of</em> the parts can be found used, there isn't much financial incentive. And modular amp building seems pretty inexpensive--finding manufactured power amps, etcs., then slapping a preamp module to it.</p><p>With the &quot;reuse&quot; ethos, it's still possible to make cool projects cheaply. But some of those are a little hard to duplicate...(that's why I'm always looking for junked amps, transformers, etc).</p>
<p>The price of tube amps is just ridiculous these days... back in the 80s &amp; 90s everyone was doing transistor solid state amps... tube amps were fairly inexpensive... as my good buddy Bill said, &quot;now it's a $1000 for a decent 50 watt amp because everyone realized tubes sound better...&quot;</p><p>I'm making my own..... I need a hobby anyway</p>
<p>I believe there's a mistake in the scheme. There's a downshifting resistor after B1 - it's stated that it should be 1k 5w. But if i understand the wattage requirements right it should be at least 190x190(V)\1000(R)=36100\1000=36.1(W). So there should be 50w not 5w. Correct me if i'm wrong.</p>
<p>That resistor drops 10V (190V -&gt; 180V), for a load of approx 10 milliamps (0.01 amps), which is all the preamp tube and the rest of the HV draw.</p><p>The wattage specifications for that resistor are WAY over engineered--the total wattage passed through are approx 0.1 watt. 10V * 0.01A == 0.1 W.</p><p>It's normal on older amps to see a 1/2 watt carbon comp resistor in these places. I probably should spec this for 1 watt, just to be sure...</p>
Is there anyway around using the GE century TV trans. Without to much trouble. So new at this the article makes my head spin. I need that tube sound tho....
<p>Best way to get used &quot;iron&quot; (transformers) for tube amps is to scavenge a thrift store tube organ. Sometimes they are even free on Craigslist, etc. Old tube gear will have both the required transformers, too (power and output).</p>
<p>If you still want to build an amp using two of the 50L6 tubes, they normally work in a series heater string, not in parallel like 6 volt heater tubes. You can use a standard isolation transformer for the heaters and run the heaters of the 50L6s in series, that accounts for 100 volts of the 120 volts. Then you have to calculate a dropping resistor for the other 20 volts or heat another tube, and then use a dropping resistor. All the tube heaters in a series string must have the same heater current value. <br>I'm sure you can find more information on series string heaters on the internet.</p>
<p>I'm not a big fan of AC filaments, although I have several &quot;radio tube&quot; amps. </p><p>You can read more in-depth about those amps (and how to calculate the series resistor) in my &quot;isolation transformer upgrade&quot; instructable...</p><p>Thanks!</p>
Thanks for the reply!<br>The isolation transformer upgrade article is pretty interesting. I have a number of various series heater string audio tubes that I would like to experiment with, and I have several 100 VA isolation transformers that I can use for the power supply.<br>Amber
<p>Nice--sounds like some fun projects! There are a lot of intriguing hacks with those higher voltage filaments.</p><p>Sadly, tubes like the 50L6/6DG6GT use more current than the more common (and expensive) power tubes such as 6V6, EL84, etc, and that makes them less attractive to me (now)...</p>
<p>Interesting project, I've built several hi-fi tube amps in the past, but now only work in silicon.</p><p>Anyway, you might want to look int the 6W6GT as output tubes.</p><p>Same basic specs as 50L6 or 6DG6, but with a plate voltage rating of 300V.</p><p>You might be able to use a higher plate supply voltage with a more common transformer voltage - and get a bit more power output to boot.</p><p>Just some food for thought.</p>
how do you drain the electricity from the amp
<p>use a light bulb on the caps</p>
<p>It is always best to drain the charge off of a capacitor through a properly rated power resistor than to dump it all of a sudden, it could cause unintended consequences.</p>
The current is stored in the &quot;filter caps,&quot; which are connected after the power supply rectifier. Essentially, you need to drain those capacitors; shorting them by connecting the leads together, or by connecting the positive (+) cap terminal to the GND.<br/><br/>There's more info <a rel="nofollow" href="http://www.instructables.com/id/SQWYOA2F9T400IV/">here, on my amp rebuild project.</a><br/>
If you short across the caps with a wire (or tool) you'll get a high current snap as it arcs. If you use a resistor (perhaps 1K, at least 1W) it'll take a little longer to drain but it'll drop the arc intensity.
Thanks! (the link above also mentions using a jumper / resistor combo:)<br/><br/><em>-- OR jumper the positive (+) lead of each large cap to GND for several seconds. A jumper with a built-in resistor (10K or so) will help prevent sparks here...</em><br/>
<p>It is the Amps (current). What I mean is it is the amount of current that you push through your body. If your body resistance is low enough and the voltage you come into contact with is high enough to cause your body to conduct or pass as low as 100mA, it could possibly kill you, or at least do damage. In other words it ends up being the amount of power your body dissipates. generally you can get a pretty good sensation out of as low as 5mA. The Ohms Law relationships should be understood.</p>
<p>It's <em>both</em> amperes and voltage--which is pretty much as I read your comment.</p><p>High-voltage static electricity won't kill you, not enough current. Low-voltage 1.5V batteries can supply well over 1 amp in short circuit, but not enough voltage to overcome skin resistance.</p>
Correct, Amps times Volts= Watts. It is the power dissipated that does the work in any type of body!<br>
<p>Hi gmoon this looks real good</p>
<p>Thanks, mate. It's a real rat's nest from all the changes, but still sounds pretty good. Wanna make some more &quot;original&quot; designs someday, but there's always an old project amp here to fix.</p><p>I think I've commented on your stuff before. It's good stuff.</p>
<p>Here's another question: does the speaker power rating have to be the same as the amp's, like, could i use a 15W speaker even though my amp probably won't output 15W ?</p>
<p>So long as the speaker has a higher power capacity rating than the amp it's OK.</p><p>You shouldn't use a 15W speaker with a 30W amp, that would blowup the speaker (if the amp were turned up). The standard approach is that the speaker have a wattage capacity as large as the <em>peak</em> output power of the amp...for tube amps that's usually twice their normal power rating (so a 15W amp requires <em>at least</em> a 30W speaker).</p><p>For amps with much lower output power than the speaker--no problem. People are always plugging in their little micro &quot;Smokey&quot; amps into 200W Marshall cabs, and that works just fine....</p>
<p>Ok, now i've got a problem: the max plate dissipation watts of the EL84 is 12W, so if i do Va/(Pa/Va) running my amp t 20V, the OT primary impedance should be about 33 ohms. I don't think i can can find anything like that, but i found a 1500 to 600 omhs OT, so do you think i could use that with a 12 ohm speaker, or like a 8 ohm speaker and a 4 ohm resistor in series, or something like that ? (1500:600=30:12)</p>
<p>Yeah--not totally certain if that formula will work at these low voltages, but every time I run the numbers for an amp it's darn close to the spec'ed impedance. So it's certainly worth using as a starting point.</p><p>It's a little crazy but I'm not certain you <em>need</em> an OT--two 16 ohm speakers in series will yield a 32 ohm load resistance. Might work just like that, wired as the load...</p>
<p>Yeah, but i wouldn't get as much volume, right ? And it's certainly cheaper ($2.25)</p>
<p>The purpose of an output transformer is just to match the high output impedance of the tubes to the low impedance speaker. For devices with low output impedance (such as transistors), no OT is needed. If the impedance of an EL84 @ 20V really is ~33 ohms, there isn't much point in using one.</p><p>Certainly the best load would probably be found by experimenting with a various speakers (8,16, 32, 64 ohms) or even w/ the transformer you describe. The formula could be off at these voltages...</p>
<p>I heard that the lower the speaker impedance the more volume you get, but i really don't know (do you?). If it doesn't make volume difference i'll do what you suggested, but speakers are kinda expensive, so if i use, say, an 8 ohm speaker and 24 ohm resistor or something like that, would it work (an sound) as well ?</p><p>And if the formula goes off at low voltage, could it go off enough to damage the amp or other negative effects ?</p>
<p><em>&quot;I heard that the lower the speaker impedance the more volume you get&quot;</em></p><p>This is true for solid state amplifiers, not for tubes. Solid state amps deliver more and more power with lower load impedance until they short out/melt down (or the internal protection, if any, kicks in and silences them until they cool).</p><p>But there's a &quot;sweet spot&quot; for the load resistance of a tube--anything higher or lower still works (to some extent), but doesn't yield as much power. There's a load R vs. power output graph in this very instructable (<a href="http://www.instructables.com/file/FBMZDOIFHEJ16K9" rel="nofollow">see it here</a>). The graph shows the max power/min distortion at ~3500 ohms, and the power drops off above and below that value.</p><p>Of course you should never run a tube amp without a load (or as a dead short), but for a low wattage amp like yours, anything in the general neighborhood would probably work... Stay away from the extremes and you probably won't harm the tube.</p><p>If you place a 24 ohm resistor in series with the speaker it might approach the &quot;ideal&quot; impedance, but it will consume current that you'd want to reach the speaker and act as a voltage divider (speaker will get 1/4 the output voltage), so you'll loose LOTS of power/volume. It would in fact be acting as an attenuator.</p><p>Just dig up some cheap speakers at a thrift store or tossed radios, etc., to test it. You might get lucky and like how they sound.</p>
<p>OK, i'll use speakers to match the impedance. But this brought up to another problem: i wanted to make sure i don't use speakers with a lower wattage then the amps' wattage. So i looked for some of those EL84 loadline things and i found some, but they were all completely different.</p><p>Is there any way to know which one to consider ?</p>
<p>I don't think an EL34 @ 20V is going to produce much more than a watt. And speakers in series share the current, so don't worry about the individual wattage per speaker, look at series resistance.</p><p>I'm not sure you can draw a loadline with the existing graphs, 0 to 20V would only occupy a small sliver of the graph.</p><p>The Load imped. vs. wattage graph I linked to is for a specific voltage (see the little inset--plate V = 200V, screen = 110V). Even at low voltage I'd expect the output graph to have a similar shape. But you'll never find a similar graph for 20V, that's <strong>way</strong> out of the designed operating voltage for almost all &quot;normal&quot; vacuum tubes.</p><p>Most datasheets have examples of operation in typical topologies. If you look at the datasheet for the above tube (6dg6gt), you'll see SE operation at 110V and at 200V. As expected (and you've formulated), the load resistance for those two operational points is 2000 ohms and 4000 ohms. </p><p>I.E., as you've come to suspect, the load resistance drops considerably with the plate V.</p><p>You gotta remember you're in uncharted territory here. No guarantees; the next step probably is experimentation... </p>
<p>Sorry, meant to write EL84, not EL34...</p>
<p>So, is your point that i should just try out speakers without worrying too much about their wattage ?</p><p>Oh, and about the loadline differences, i figured out the problem was that i didn't consider the screen grid voltage (sounded weird that EL84s could output 2 watts at 20v!)</p>
<p>Yep.</p><p>Well, pentodes do have different characteristics than triodes, and might yield more wattage that I originally thought. But wattage is V x A, so don't expect to get <em>too</em> much due to the lower voltage.</p>
<p>Ok. The speaker should be connected from one side to the plate and from the other to the ground, right ?</p>
<p>To the B+, like the way a SE OT functions. </p>

About This Instructable


672 favorites


Bio: Go sit in the Faraday cage and think about what you've done...
More by gmoon: Isolation transformer upgrade for old guitar amps The ValveLiTzer: Low-voltage Tube Booster Guitar Tube Amp
Add instructable to: