Amp

[printer2]

I have been taking part in a challenge on another site to make a bedroom amp for under a hundred bucks. I messed around with a few ideas and will be revisiting them at a future date. Somehow I took a detour and decided to do this build. In part because of the schematic I drew up switching between a paraphrase and cathodyne PI in another thread here. I did a schematic earlier switching between those two and the long tailed pair sparked by a inquisitive bunch, especially a bright fella named merlin. I finally decided to wring it out in a tabletop radio I found in a thrift store.

I found that an 8" speaker would just fit in it, I cut up the front and mounted the tuning dial along the side where I will be putting the controls. The hole on the bottom is for the input jack, the knobs will be volume, tone, PI type, still deciding on the next knob, maybe master volume maybe power light, and then a power switch. The chassis is a nightmare with compound angles to fit the box and as to mount all the components to take up the least amount of space from the speaker. I want the speaker to be in a sealed box as with a cabinet this size I think it will sound thin if it were open back. Hopefully it will give a 40's flavor with the paraphase PI, a 50's with the cathodyne, and a early 60's with the LTP.















Because I want the speaker sealed in the box, well maybe ported, I am making the chassis in the side of the box. It is in three pieces and will be screwed together. All kinds of crazy angles on the sheet metal to fit the enclosure and take up the minimum of speaker box space. The grill cloth material is a open weave placemat that I stained to give it more of a period appropriate look, along with discoloration to imitate the years gone by. I will probably miss the deadline this is going to be a crazy build, no mater this is more an experiment and an amp I wanted to build for myself. With one triode before the PI sections it will not be a scorcher, hopefully hitting some mild tones and enough volume to drive the inefficient 8". I will get back to the other designs as time permits, I really want to play with the 6AU6 some more. Some pictures.

[P Rails]

Where's the rabbit ears ?.. Nice build, I've been itching to build a low watt PP amp. Hope you post clips.

Does the challenge call for a chassis and cab or combo, or just a chassis build?
I've worked up a 3w 50L6 /6BQ7A single ended chassis build that uses new parts for under $100.

[Telenut62]

A good percentage of TDPRI dudes and dudettes wouldn't know what you mean by rabbit ears...lol There a fellow here in Australia that builds these kind of amps using 50's and 60's art deco radio cabinets etc. Buggered if I can remember his website though

[printer2]

Originally Posted by P Rails

Where's the rabbit ears ?.. Nice build, I've been itching to build a low watt PP amp. Hope you post clips.

Does the challenge call for a chassis and cab or combo, or just a chassis build?
I've worked up a 3w 50L6 /6BQ7A single ended chassis build that uses new parts for under $100.

Just that the circuit has to be assembled safely in a box, I was just going to do a bare chassis with the tubes hanging out. I was doing a build with a 6AU6 in the middle of the preamp but I decided I wanted to have a circuit that took me over the $100 U.S. mark. Mind you I probably constrained myself more than others in my selection of part suppliers.

Search for 'hundred buck amp challenge diy' and you should find it. Better hurry though as the deadline is the end of this month.

[printer2]

Did a preliminary general layout, I will be using radial caps so it will be a little different. Dropped the NFB for now, included a tube rectifier on the diagram along with two 100 ohm resistors for heater elevation. The software I was using did not have a suitable 4PTP switch so I sort of made my own. Hope it does not look too scary.

[printer2]

So far so good.

[telex76]

Cool.

[lckyjcky]

Much admiration for your mad skills!

[printer2]

Originally Posted by lckyjcky

Much admiration for your mad skills!

More like just plain stubbornness. Try something, did not work, redo till it does. I noticed that the layout and the schematic is missing a switch that I put on the chassis. When in LTP mode the cathode bypass cap can not be connected to across the resistor otherwise the second triode does not get a signal. I will fix that on the final schematic. The second switch on the back is to parallel the output cathode resistor's capacitor with a 1000uF cap to tighten up the amp and give it a fixed bias feel.

This amp is more an experiment as I do not know how much effect the change in PI's will have on the sound. Also I am using a third rate output transformer, more to see what it would do. If the results are promising I will use the circuit with a better transformer and a couple of more ideas I have. I drew out a preamp where I can switch from a Tweed type tone control to a bass/treble Blackface circuit. Hopefully it will give the amp a variety of sounds.

[printer2]

Made it a little more complicated, fool that I am. Decided to add a Baackface tone circuit along with the above Tweed circuit. Dual pots. The PI selector switch is moved over, hope the lead dress does not cause it to squeal. The schematic had a few changes, found the odd mistake along with the addition of the tone selector switch and dropping the final volume control. Boy I hope this thing is going to work.

[limbe]

I am curious how the 1000 µF cathode capacitor will affect the sound.

[milocj]

Originally Posted by printer2

Made it a little more complicated, fool that I am. Decided to add a Baackface tone circuit along with the above Tweed circuit. Dual pots. The PI selector switch is moved over, hope the lead dress does not cause it to squeal. The schematic had a few changes, found the odd mistake along with the addition of the tone selector switch and dropping the final volume control. Boy I hope this thing is going to work.

I'm not sure you can add the blackface to the project...it may be a little more "color" to the tone than that box can handle.

Nice build and out of the box thinking.

[printer2]

Thanks.

As it stands now, threw together a jig in order to wire it up and shake down the amp. Just got power to it last night and am chasing down mistakes. Missed cutting one trace and soldering one wire so far, anticipating a few more issues. I soldered the parts in and then changed the design a few times moving parts around. Left cutting traces close to the end to accommodate my (what seems quite normal now) changes of direction. The switch was to be in the center of the board and now I have longer leads, hope it does not cause me issues.

Way too complicated a build for the size and method of construction. It might be different with a proven design but as an experimental design it is just a pain. Also does not help spending a couple of hours on it and then not getting back to it for a week at a time. Plan to just get it going and then put it aside as I have house renovations to complete.

[printer2]

Well it is sort of alive. Got the cathodyne and paraphase PI's working. Not sure what is going on with the LTP but that will be for another time. Boy does the Fender tone stack really suck the gain out of an amp. The Tweed tone control really rocks as far as overdriving the amp. Can only get a clean output with the tone stack. Mind you it is a pretty sound.

Not running any NFB yet, wondering if it is worth it, mind you I set up for it so I can play with it later. Tried it with the 8", 10" and with the 12". Surprised there is no squeallies and it is quiet even without a case around it. Mind you the sheet metal the controls and the tubes are on probably works as a ground plane. The two PI's do sound different. can't wait to get the LTP going so I could tell how that one is compared to the two. This is going to be a fun little amp when I am done.

[printer2]

Silly me, right from the start of doing the switchable PI thing I missed the connection to tie the cathodes together for the LTP. I am short a pair of contacts to do it with the rotary switch, could do it with the addition of the switch I put in to switch in and out the bypass capacitor one the second triode. Not sure which way I will go though.

The cathodyne sounds thinner that the paraphase at moderate levels, I probably have quite a bit of unbalance going on, I put in a pot in place of the 56k resistor so I could adjust the paraphase balance. At more gain (more distortion) the paraphase starts sounding a little ragged. The switching in the 1000uF cap on the output cathodes does have a little effect on the bottom end, makes it a little stiffer but you really have to be paying attention to notice it. Might be different on other amps.

I probably should redraw this thing and simplify it so that it is a basic single channel Tweed type amp. Maybe something like this. If anybody is interested I threw together a parts list.



It was to be an amp that anyone can make for under $100 bucks, think I got in the $80 range with 12V tubes, 6V would push it up to $100. You could also sub out the 6AQ5's that I used for 6V6's, mind you doing that would call for a beefier output transformer.



I'll have to play with this one some more before I finally decide what to do with it.

[printer2]

Boy this amp has been a long time coming. I do not know how many times I have rewired it. Originally the board started in a different chassis with a pentode front end, then with a switched PI section along with the ability to switch between a Tweed and Blackface tone control. Strip that out and use the above schematic but with the Blackface tone stack before the Cathodyne stage with a switch to take it out of circuit. Found there was not enough voltage swing to drive the output enough with it in that position. Finally settled on the current schematic, I think the sound is 95% there, will live with it for a while before I decide if I want to tweak it any further.



Right now it is still on the jig and I hope to start putting it into the cabinet tomorrow. The switch to bypass the tone stack sure shows how much loss it has. Barely breaks up with it in. The next position is straight Tweed flavored but turned up too much using the 0.022uF coupling capacitor has the bass starting to sound bad. The third position has a 0.0015uF cap and it tames the bass. Not sure if it is the final value but it sure makes the gain usable. The cathode resistor value is still up in the air, would not mind more current through the tubes but since I am running them at a lower voltage I may end up with a Class A amp rather than a Class AB. The 1000uF cap does not make enough of a difference to include it but it is there so I will leave it be.

[onenotetom]

Keep us posted. Clips would be great. What are you using for a speaker?

Thanks for the honesty on your mistakes. Got my DRRI ready to bench test the other night and found a wire that I had not soldered. Turned it on and I have other issues. Had to leave it to go to work. This is where I try to go from amp assembler to amp troubleshooter.

[printer2]

Originally Posted by onenotetom

Keep us posted. Clips would be great. What are you using for a speaker?

Thanks for the honesty on your mistakes. Got my DRRI ready to bench test the other night and found a wire that I had not soldered. Turned it on and I have other issues. Had to leave it to go to work. This is where I try to go from amp assembler to amp troubleshooter.

Mistakes? Heck that is just a fraction of the wrong turns. With the old eyes not being as good as they used to be even with glasses and all the time the board has been soldered you miss stuff. Strip some parts and put a new arrangement in and you miss a bridge between traces that is suppose to be a ground path. And then there is the case of the output tubes. Two NOS tubes, trying to figure out why some of the voltages are not where they should be, the output sounds weak and distorted, why there is a greater voltage drop across one side of the primary than the other, finally touched one of the tubes and it was much hotter than the other. Was it due to using a transformer not built for amplifier service or did I use one of the wrong taps? Finally replaced one of the tubes and the amp started acting like it should.

I will run down some of the things I found with the amp when I am done, I will do a more detailed look at the collection of parts with the next one I am anticipating building. I am quite happy that a reasonably decent sounding push-pull amp for home use can be built with inexpensive parts. That was the whole purpose of the build, to see if it was feasible.

[telex76]

You are really sticking to it.

[printer2]

Well I didn't put the amp together like I said I would, instead I have been poking around doing some measurements. I wanted to change the bias to warm up the tubes some. The data sheet gives values of 250V plate voltage, grid -15V and 80mA per pair in Class AB. Taking the -15V divided by the plate current gives 190 ohms, actually a bit more since you have the screen current through the resistor also. So values of cathode resistor I have seen for 6AQ5 amps of 250 ohms seemed in line. Since I am running at a bit lower voltage I may need a lower value but best to start off a little higher, and since the parts store in town does not have the greatest selection I started with 330 ohms. I added resistors in parallel to bring down the value since the current seemed quite low. Never put a signal through the amp with 330 ohms, wish I did, just looked at the idle current and bias voltage. Eventually put some signals through it, changed the resistor to a 220 ohm and did the parallel thing again to get the current close to the datasheet value.

I have a variac to adjust the voltage to the power transformer and a second transformer to feed the heaters so I can play with the supply voltage. I made some measurements with 120V going into the PT which gave me a voltage to the output transformer of about 235V. I would have liked some more voltage but the limited room in the chassis determined my choice in power supply circuit (that was when the board had more parts, now there is enough room if I decided to make a new board, not bloody likely).

Being a cheap build I am not using a regular power transformer with high voltage and heater windings. Much cheaper using of the shelf 12v transformer for the heaters (in my case I grabbed a beefy 12VDC wall adapter and used the transformer from it, $2.00 from a thrift store). The high voltage is a standard 1 to 1 power transformer, put 120V in, get 120V out. Well in this case I bought one with a dual coil primary that can run on 120V or 240V. Just flip the transformer around and you should put 120V into the secondary and get 240V out of the original primary, right? Well not quite. Hammond, in this case, steps up the winding ratio so that the secondary has a few more windings to make up for the transformer losses. So when you use the transformer backwards you end up putting in 120V and get out 200V (or there about). With filtering and dropping resistor I have 235V.

Now if I wanted more voltage I would use the same transformer but used in the direction it was meant to run. Funny thing about Hammond transformers is they put out more voltage than so that you get at least 120V at the rated current. This transformer puts out 135V open circuit, so if I use it the proper way and use a voltage doubler on it I could get in the vicinity of 280VDC when all said and done. Maybe next time.

So anyway the tubes are warm, they have the 235V on them, the current through the tubes is low, try dropping the cathode resistance some more, down to 115 ohms. A heck of a ways down from the 250 ohms Gibson used (and the Pushy Croaker, an interesting 6AQ5 design). So in the interest of, what can this thing do, I did some measurements at the normal voltage with the 220 ohm and the 115 ohm values and I also ran the two with the power supply at 260VDC.


215 ohm, Supply Voltage, Bias Voltage, mA, Watts, Dissipation %
idle
260V - 15v , 70mA , 17.4W, 73%
cranked
255V - 17.3V , 80mA , 19W, 79%

115 ohm, Supply Voltage, Bias Voltage, mA, Watts, Dissipation %
idle -
260V - 12.3V, 100mA , 24.8W , 103%
cranked
262V - 12V , 104mA , 25.4W - 106%


215 ohm, Supply Voltage, Bias Voltage, mA, Watts, Dissipation %
idle
236V - 13.5V , 72mA , 14W , 58%
cranked
230V - 15.5V , 63mA , 15.9W , 58.%


Did not adjust supply up, voltage drop due to increased load.
115 ohm, Supply Voltage, Bias Voltage, mA, Watts, Dissipation %
idle
225V - 10.4V , 90mA , 19.4W , 81%
cranked
229V - 9.7V , 84mA , 18.4W, 77%


235V - 215 ohm - 6.2VAC, 115 ohm - 6.2VAC but 0.5V more gain, 8VAC distorted
260V - 115 ohm - 8VAC,


Plate Voltage Swing- one plate to ground
260V - 115 ohm, 130VAC


235V - 6.2VAC 4.8W 215 ohm
236V - 13.5V , 72mA , 14W , 58%


Yeah I know, what the heck was that. Well I took the measurements at idle and then I turned up the guitar with all the controls pegged and beat the strings as much as they could be beaten with a pick. For the technically curious there is the data, for the rest, to get the current flowing in the output tubes I ended up in Class A, the higher voltages with the higher value resistor ended up slightly in Class AB. Now I wanted to have the tubes around the 80% dissipation level which is 9.6W of the 12W limit of the tubes. I went back and looked at the data sheet and found the tube dissipation at idle was less than 65% if you assumed the tube was using up all the voltage.

So either I could have some cold running tubes (in guitar terms) in Class AB, or some warm biased tubes but an amp that is running Class A. Ironic that there are a lot of companies out there advertising Class A amps when the tubes would burn up if they really were and here I am with one. Well that is what you get with a low voltage.

So when everything is said and done, the amp puts out about 5W clean power and maxes out about 8W. Push it up to 260V and you gain another watt or so. Never tried playing with the screen resistor, might get a bit there also. Scoped the plates and had 130V before clipping, multiplied by the current and wattage matches up with what is coming out of the secondary. I was concerned about the transformer not being able to take the power and getting yucky transformer distortion but no such case. Had one probe on the primary and one on the secondary and the sine wave mirrored each other through clipping. The transformer I have has a 4 and a 8 ohm tap and I switched back and forth giving the tubes around a load of 5k and 10 k. The 10k load had a little bit more peaking of the leading and trailing edge of the clipped signal while the 5k load had more rounder transition. Almost forgot, with the 115 ohm cathode you had a touch more gain, maybe 5%, but the output voltage clipped at the same level as the 215 ohm value.

I also did something strange to some (some are saying to themselves, all this seems a little strange if you really think about it), I ran the amp as a single ended amp. Basically I turned off the signal to one of the tubes and only ran the other. You can do that if the amp is running in Class A. The shut off tube sits passing the idle current while the other amplifies the signal sent to it. Heck, since I was feeling playful I checked out what the amp will put out in SE at lower plate voltages.

Single Ended, one output tube driven

Clean output -
Supply - Volts out

75V - 1.2VAC
120V - 2.25VAC
170V - 4VAC, 4.5VAC distorted

Both tubes driven 235V - 6.2VAC
SE 235V 215 ohm - 2.25VAC
SE 235V 115 ohm - 3.1VAC
SE 260V 115 ohm - 3.6VAC
SE 260V 115 ohm - 4VAC distorted


It didn't sound bad at all with guitar, guess I never mentioned all the voltage outputs measure were done with a sine wave and into an 8 ohm speaker. Since the 1000uF capacitor did not make an audible difference to me I will be removing it and using the switch to go into SE mode. Cuts the watts by a bit better than half. That's it as far as the power amp is concerned.

And just a little bit on the preamp. You know what I said about the preamp only having enough gain to just get the amp into distortion? Yeah well put in a fresh tube you dummy. It got rid of most of the hiss also. So that was the 12AT6 (one triode with the gain about the same as a 12AY7), with a 12AV6 (one triode 12AX7), well more gain. Good thing I have switches for the cathode bypass on the two stages.


Now, may try some negative feedback. I am also going to remove the 1.5nF coupling cap on the switch and have some screw terminals external from the amp. That way I can tweak the value, maybe add a resistor or two to it. Just got to get this thing buttoned up so I can move it off the bench.


I should say something about the speakers I have been using. I have an 8" that is going into the amp. It is naturally bright being a small speaker and it has surprising bass for the size of cabinet. Again it will be closed back otherwise for the size of the box there would be no bass if it was open back. The main reason for building the chassis the way I did. I have a 10" speaker from a console stereo that I just picked up this week. Not very efficient, not that bright as it had a midrange and tweeter, but being from the 70's it is a paper surround and has an alnico magnet. Also have been using a WGS 12" ET65. The 8" can be played loud enough clean at practice levels, that's 5w clean on the scope and no audible distortion, and like any Champ get snotty. The 10" sounds more like a midrangey guitar speaker, probably at its best with a healthy dose of distortion. The WGS sounds really nice with this amp, might even handle a mild mannered drummer.

Well that was a lot of typing with two fingers.