Question about dissipation vs output and OT impedance mismatch

[munkeyboy]

I've built a SE amp from some spare transformers laying about from a small reel to reel.   The PT only puts out about 250v and powered a 12BH7a and 12ax7.   At first, I try the micro-champ build, but I didn't like the sound of that amp.  Then I built a EL84/12AY7 thing with volume, framus mid, boost switch.   I put all this in a 6x4x2 enclosure.  It all turned out good except a bit of 60Hz hum.  Things are so tight in there, this hum was kindof expected.  I may increase the first filter to something greater than 47uf if I can fit it.

With the PT, I can only bias the el84 to get 6w. 93v plate to cathode, 9v on cathode, 130R resistor.  I like the amp because I was going for a very bedroom friendly amp and don't want to change it minus a few tweaks.  My question is why does this 6w dissipation only sound like 2w output?  Is it because of the OT mismatch.  The OT is a 12k primary, 4ohm secondary.  The el84 expects 4k and the speaker is 8ohm (10").  Am I going to kill the amp with this mismatch?

If feel like I'm on to something with this build (a small watt amp that sounds good), but I'd like to know why it sounds good at such low output.

[DummyLoad]

6W in SE would result in 3W ideally (50% of power delivered to load) typically that is around 40-45%, so <2W output with 6W applied in this case is not unreasonable with such a gross mis-matched load. the output tube is loafing and OT core probably not saturating. SE sounds better when pushed to maximum if you are not saturating the OT core.


--pete

[PRR]

> PT only puts out about 250v

So how come you only have 102V (93v plate to cathode, 9v on cathode) going toward the tube?

> The el84 expects 4k

The tube has no fixed expectation. It will probably be near-optimum when load is similar to V/I. 93v/0.069A is 1,340 ohms. Way-way off from 12k. You may really have a part-Watt. (which really can be plenty at home.)

[munkeyboy]

> PT only puts out about 250v

So how come you only have 102V (93v plate to cathode, 9v on cathode) going toward the tube?

> The el84 expects 4k

The tube has no fixed expectation. It will probably be near-optimum when load is similar to V/I. 93v/0.069A is 1,340 ohms. Way-way off from 12k. You may really have a part-Watt. (which really can be plenty at home.)

Good question.  I'm not sure, but I assumed it was because of the PT: 250v @ .015A (15mA).   If I'm reading it right, 15mA seems pretty low.  See attached info on the old tape machine.

My first dropping resistor is only 470 which is where g2 it tied to.  The second dropping resistor is 10k which goes to the preamp.  I think that is left over from the first try, but sounded good, so I left it.   I haven't worked up the schematic yet.  But attached is the layout.

After playing thru high output pickups on neck position, I think I need to rethink.  It sounded good on tele, but my jazzmaster and humbucker pickups are producing tons flab and fizz.

[tubeswell]

I've built a SE amp ...  The PT only puts out about 250v ... I built a EL84/12AY7 thing with volume, framus mid, boost switch.   ...

With the PT, I can only bias the el84 to get 6w. 93v plate to cathode, 9v on cathode, 130R resistor. 

250V B+? or 250VAC on the secondary?


Either way, you should get more than 93V plate-to-cathode, unless the PT heater can't handle the 1060mA heater filament draw - in which case you could try putting a 6V LED lamp in there to reduce the lamp current draw. https://www.aliexpress.com/item/32813994977.html

{EDIT- tracking info removed from over-long URL-- PRR}

[munkeyboy]

250VAC.  Sorry left out the original power section of the schematic

I do have a little red LED I'm using (with a resistor and diode).

I can get more than 93, but that means increasing the Rk which in turn makes the tube bias colder.

[munkeyboy]

Well, I don't know what I did, but I recheck my values. And now I'm at 164.6v plate/5.6v cathode.   Still  6watts.

The only change I made was adding 5.6k screen resistor to the el84 grid.

[Williamblake]

You changed the rectifier?

[tubeswell]

Well, I don't know what I did, but I recheck my values. And now I'm at 164.6v plate/5.6v cathode.   Still  6watts.

The only change I made was adding 5.6k screen resistor to the el84 grid.

What is the VAC on the High Tension secondary?

[munkeyboy]

BTW,  here is a lesson in safety.  Make sure your power switch and and components next to it is secure before turning on the switch.  The other day, I forgot to bolt down my pot and switch before turning it on to test voltages.   The hot side of the switch was touching the pot.  The attached pic shows the result.  Nearly melted the pot shaft in half.  For sure the threads are gone.   Thank goodness I always put one hand behind my back when working with live voltage.

Also, don't put a switch right next to a pot. 

[munkeyboy]

You changed the rectifier?

No.

Well, I don't know what I did, but I recheck my values. And now I'm at 164.6v plate/5.6v cathode.   Still  6watts.

The only change I made was adding 5.6k screen resistor to the el84 grid.

What is the VAC on the High Tension secondary?

VAC is 250.    Volts after the bridge rectifier unloaded is 377vdc, loaded around 203vdc

[tubeswell]

What is the VAC on the High Tension secondary?

VAC is 250.    Volts after the bridge rectifier unloaded is 377vdc, loaded around 203vdc

That is quite a big drop.

Given you got a higher voltages by adding 5k6 Rg2, there may be a problem with the screen in that EL84. Have you got another one you can try?


And what happens to the voltages if you take the lamp out?

[munkeyboy]

Sorry, the actual VAC of the HT is 287 unloaded and 187 loaded.

[tubeswell]

Sorry, the actual VAC of the HT is 287 unloaded and 187 loaded.

Still a big drop. Indicates something wrong with the PT load regulation or a component is loading down the voltages way too much, or overtaxing one of the windings. see my earlier comment about the lamp on the heater winding

[munkeyboy]

Hmm.  Ok, I'll start digging tomorrow.  I did try a new tube with the same result.   It could be I have something loading it down or grounding out. 

[tubeswell]

...  It could be I have something loading it down or grounding out.

On that chart you posted, the heater winding is 6.3V @ 1A. You have 1.06A with just the EL84 and the 12A_7. The lamp could be 150mA, 300mA or 500mA on top of that, depending on the lamp

[sluckey]

It could be I have something loading it down or grounding out.

In addition to the under-rated filament winding, the HT winding is grossly under-rated. Your chart says .015 amps. That's only 15mA! Way too small for an EL84. This is why your B+ changes so much when you plug in the tubes. That PT probably gets very warm. I doubt it will last much longer.

For example, you said "9v on cathode, 130R resistor". That means there is .069A flowing through that resistor. And the PT is only rated for .015A!

[munkeyboy]

Good question.  I'm not sure, but I assumed it was because of the PT: 250v @ .015A (15mA).   If I'm reading it right, 15mA seems pretty low.  See attached info on the old tape machine.

I thought that 15mA could be the cause, I just didn't know the math.  Now i do sort of.   Oddly the PT does not get that hot.   

Looking at the original schematic voltage readings.   Does that mean that 1/2 of the 12bh7a was biased at 21% for about 1.5watts (7.6mA)?   The other half at 2mA, with the 12ax7 at .5mA.   So about 10mA plus heaters.   

[munkeyboy]

Either way, I'll give up this el84 and see if I can make the 12bh7a or 12au7 work within .015A.   Or sub the PT.

[sluckey]

Either way, I'll give up this el84 and see if I can make the 12bh7a or 12au7 work within .015A.   Or sub the PT.

Good idea. You may just come up with a nice bedroom amp. I'd be interested to see the final schematic and hear your thoughts about the sound.

BTW, I calculated 1.6W dissipated by the output tube using 8V for cathode voltage. So, yeah, this was well within the capabilities of that little PT.

[munkeyboy]

I did try Rob's Micro champ circuit with these transformers and a 12bh7a.  I did not like the sound, but I also had a 12ax7 instead of 12ay7.   The overdrive felt very fuzzy and unpleasant.  I'm going to try the parallel power section again, but keep my existing preamp which is still champish. 

EDIT: nevermind on the micro champ.  That 12bh7a will draw too much current in that config I think.

[munkeyboy]

Ok, so I don't have a 12AU7 on hand, but do have a 6CG7.  I believe they are similar, but with different heater pins and the 6CG7 has more heater current.  I am using LED, not pilot light.  So I am right at .922A I believe.
6CG7: 0.6A
12AY7: 0.3A
LED: 0.022A

Using this in parallel output, I believe is it using about 12mA.   That puts me super close to the 15mA.   I think the 12ay7 is around 1.5mA.  So I think I'm ok right?

278v on plate
9.8 on cathode
268.2 plate-to-cathode
820ohm Rk

Still getting that static/fizz sound on breakup.  I'm guessing blocking distortion.

[tubeswell]

Ok, so I don't have a 12AU7 on hand, but do have a 6CG7.  I believe they are similar, but with different heater pins and the 6CG7 has more heater current.  I am using LED, not pilot light.  So I am right at .922A I believe.
6CG7: 0.6A
12AY7: 0.3A
LED: 0.022A

Using this in parallel output, I believe is it using about 12mA.   That puts me super close to the 15mA.   I think the 12ay7 is around 1.5mA.  So I think I'm ok right?

278v on plate
9.8 on cathode
268.2 plate-to-cathode
820ohm Rk

Still getting that static/fizz sound on breakup.  I'm guessing blocking distortion.

The PT is looking healthier. The 6CG7 is idling at 3.2W. The datasheet Pmax is 5W (for combined triodes in parallel), so you won't get the most out of the tube, but for 3.2W, in SE centre-bias Class A, Zout wants to be 278/(3.2/278) = 24,151R (24k) to get the most out of that operating point - or something close to it.


The static overdrive fizz could just be what a 6CG7 delivers when run as a parallel SE output stage. Triode output stages tend to be like that. 12AU7s have a similar fizzy characteristic. The way to 'minimise' it would be to set the output tube up for maximum output power and centre-bias with the optimum SE reflected load*. That way, any possibility of distortion occurring is kept to 'a minimum' (although there will still be some distortion, and mucking around with the coupling cap value or adding a grid stopper might keep that to a minimum too, although you may want to be cognisant of managing the HF roll-off point that is caused by the grid stopper resistance value interacting with the Miller Capacitance of the parallel triode)

*Zout = Vp/(Pmax/Vp), where Pmax = the power dissipated by the whole stage (i.e. both triodes in parallel totalling 3.2W in your case), and Vp = Plate voltage at idle.


YMMV

[tubeswell]

24k load line for above calculation attached. (You can probably go comfortably to 12k for the load resistance, because you're running the tube at only 64% of it's rated maximum dissipation, but you will get more harmonic distortion, so over to you. If your OT is configured to reflect 12k into a 4R speaker, its simple to use an 8R speaker if you want 24k for more of a 'linear' output)

[PRR]

> see if I can make the 12bh7a or 12au7 work within .015A.

It was a perfectly good, if teeny, Power Amplifier before, working at 8mA, well within the PT's clear (in #8) spec of 15mA.

With this PT and this OT, you can not do a lot better. A lower load impedance tends to suck-out more power, but the increased demand on the teeny PT means less power. A pentode might raise efficiency from 25% to 40%, 400mW to 640mW, which is not a big change on the ear.

A 3-stage preamp may just be too much here. Early Fenders, Leo had to degenerate two stage 12AX7 to be sensible; later Champs added frills to lose excess gain. Yes, tape head is that much lower than guitar.

[munkeyboy]

That is some seamless stitching with that schematic PRR!

[munkeyboy]

Yes, I think this may be a futile effort.   I liked how that el84 was starting to sound, but wasn't feasible long-term.  And I don't like how these dual triodes are sounding.  I'll go back to the original single triode to see what it sounds like.   Otherwise, I'll just start fresh with the correctly sized pt/ot.