New build of slacker's low voltage amp

[redryderaus]

G'day all,

Here's some pics of my build of slacker's low voltage amp. It started out as a cheap solid state amp I was given (non working). I got it going but it still didn't sound good so I went looking and stumbled across this design. I was able to reuse the cabinet, 5" speakers and 34V CT transformer.

Original design thread: http://www.diystompboxes.com/smfforum/index.php?topic=70802.0
sebsongs build thread: http://www.diystompboxes.com/smfforum/index.php?topic=89096.0

Apart from my heater supply and DC filtering it is as per slacker's design. I've wired the heaters in series (25.2V) and I am using AC via a 56R 5W dropping resistor, seems to work fine without the cost of a DC regulator. The filtering is a pretty standard bridge rectifier & RC layout, it needs some tuning as I just used what I had to hand and still have a bit of ripple to get rid of. Regulated voltage is 45VDC. I'll draw up the PS if anyone is interested.

I've substituted a 12AX7 as the preamp tube. I found the 12AT7 was ok with low output pickups but got muddy with moderate-high output P90's and humbuckers. I'm using a 12AU7 for the power stage. I get nice clean tones up to about half volume then some gritty overdriven tones from there. At full volume it is perfect for a bedroom amp. Sound through the 5" speakers is about what you would expect, but still good enough that it's fun to play through. I've played it through my 1x12 cab and it's nicely full sounding and a tad louder. I wired the OT & speakers through a 1/4" socket to give me an external speaker jack.

Pics:  http://s186.photobucket.com/user/redryderaus/library/Low%20voltage%20tube%20amp

For those in Australia, you could build this amp using a 30V CT transformer (Altronics M2860L) and 5W OT (Altronics M1112), just change the heater dropping resistor to a 33R 5W.

Cheers,

red

[PRR]

Welcome!

Interesting project.

> still have a bit of ripple to get rid of.

I think as you kill the ripple you will find that the 25V AC running through the audio area will limit your cleanliness. As your heater voltages are in-sight of your plate voltage, you might consider DC heat.

For a full recording-grade power supply I'd think you want something along this plan:



Some of these values look extravagant for a tiny-amp, but they are not expensive enough to use the sharper pencil to shave-out a penny for a one-off build.

For spare-parts build you could leave the resistors as shown but use caps 1/2 or 1/3rd these sizes. That may be clean enough for bedroom work; if not-quite, you know what parts to hunt for.

[redryderaus]

Thanks for the schematic, PRR. I'll give it a try.

The amp is actually pretty quiet as is, I have to put my ear close to the speaker to hear any hum or noise. I'm just one of those people who has to fix it once I know it's there. Overall I'm happy with the build, but DC heat may be necessary as you said.

Cheers,

red

[redryderaus]

Question for anyone...

If you are using DC for the heater, wouldn't you use the DC heating equivalent of 0.707 x 12.6V = 8.9VDC? Would 12.6VDC be driving the heater harder than necessary?

Replacing the 100R resistor with a 150R would drop the total heater voltage to 18VDC (3.3W dissipation). This would get the heater voltage even further away from the plate voltage.

I'm not a newcomer to electronics, but I am a newbie with tubes. I'm having fun learning though 

Cheers,

red

[amptramp]

Tube heaters are rated for RMS voltage which is the equivalent DC voltage to obtain the same heating effect.  12.6 VAC RMS gives equal heat to 12.6 VDC.

[redryderaus]

Thanks amptramp

Cheers,

red

[PRR]

What that tramp said. History:

"6 Volt" tube heaters were _MADE_ for car *battery* heat.

6.3V float, 7V hard-charge, 5.5V if you run the radio without the motor running. All DC.

"12 Volt", same thing. The target is 12.6V DC or AC(rms).

Some time before that, "AC heat" designs were developed. The best and usual form is an insulated heater and separate cathode with so much thermal mass between that the cathode temperature does "not" vary 120 times a second. These "don't care" how the heater voltage dances. All that matters is the RMS voltage. RMS of AC needs thought; however "ALL" AC meters are calibrated in RMS. DC "RMS" is the same as what a DC meter says.

Or just RTFM:

[amptramp]

Let me add something to what PRR was saying.  AC meters may be calibrated in RMS but the typical VOM is average-reading and the typical VTVM or DVM is peak-reading.  My VTVM reads a peak value of 169.7 VAC and interprets this as 120 VAC because a sinewave with a peak of 169.7 volts would have an RMS value of 120 VAC RMS.  A low AC voltage with 169.7 volt spikes would also be shown as 120 VAC RMS even though its RMS value would be much lower.  Similarly, a 169.7 volt square wave would be shown as 120 VAC because the extra energy is not counted.

True RMS meters are expensive because they have to perform analogue calculations to derive the true RMS value.  The RMS value is the square root of the sum of the squares of the instantaneous voltage and only a few expensive meters can actually calculate that.

[PRR]

TMI.....

> The RMS value is the square root of the sum of the squares of the instantaneous voltage and only a few expensive meters can actually calculate that.

The funny thing is that incandescent lamps (or heaters) calculate RMS cheap.

Put a 12V lamp on a 12V DC battery, note the brightness.

Put the 12V lamp on any funny-shape varying or alternating wave (which repeats fast enough to keep flicker low). When it comes to the same brightness, that funny-wave is 12 V RMS. Even if it is -1V much of the time with narrow +100V blips.

Of course the eye is poor precision for light level. You can compare two bulbs with some accuracy (oil spot paper), but are the two bulbs identical? You have to swap them around a bit to check matching.

This WAS the core of a standard technique for RMS. Two resistors, one on steady (DC) power, one with the crap you are measuring. (This crap was always AC; DC is too easy to fool with fancy tricks.) Temperature sensors. When they come to equal temperature, they are equal power. Since the DC reference is known-RMS, you know the crap is the same RMS.

I saw an even cleverer implementation (theory and clever construction). IIRC, the temp sensor sensed package temperature *rise* (thermocouples do that natch). A reference through one resistor caused a known rise. The unknown wave in a second resistor drove temp up, ref resistor power was reduced to get same rise. This gets into some electrical nulling, but a host of temperature problems become insignificant.

[amptramp]

There were some meters like that and I have a Jackson-Bell oscilloscope with an RMS meter in it that may actually operate like that (temperature sensing across a load) but this limits you to a time constant set by rate of temperature rise and fall.  BTW Jackson-Bell was most famous for railway equipment and this scope weighs in like a locomotive.  But thanks for the reminder, I should look at that thing and see what is under the hood.

[PRR]

Jackson (no Bell) seems to be the 'scope company? Good sturdy stuff but I don't see locomotives.

Jackson-Bell was a early 1930s radio company. Bell appears later as Packard-Bell, decent hi-fis in the 1950s and plastic sound systems (and PCs) in a trademark re-birth.

There were a LOT of small special-gear companies who were little-known (and not preserved) outside a few contracts and the odd hulks in dumpsters (now eBay).

[amptramp]

My apologies - I had Jackson-Bell on the brain.  What I have is a Bach-Simpson 2610 scope with an edge meter under the display, but it is calibrated in waveform peak-to-peak, not RMS:



Bach-Simpson still makes event recorders and speedometers for trains in its factory in London ON Canada.  This is a 5" scope and as you can see, it is fairly large for that.  It is 15" x 11.5" x 20.5" and weighs 53 pounds.  There is a listing under Simpson, but it shows an older unit with a standard meter rather than the edge meter as in the picture above.