several tranny questions

[DRH1958]

I'd like to have the proper primary just for peace of mind, but do 6V6 require 5k or 6.6k to be optimal? I thought i saw 6.6k but not sure.

Or we could look at the data sheet: http://www.mif.pg.gda.pl/homepages/f...93/6/6V6GT.pdf Look on page two under "push-pull class AB1" then "Effective Load Resistance". Normally (if there is such a thing) a pair of 6V6s use an 8K P-P load. So using your in-hand OT, if you double the connected load that the tap is rated for, you will get 7.6k, close enough.

[daz]

8k? Does this mean 2X6V6 would be 4k? That would put it right where my OT is, yes?

[DRH1958]

8k? Does this mean 2X6V6 would be 4k? That would put it right where my OT is, yes?

No, when you look on the data sheet under "Push-Pull Class AB1", I abbreviated this. In addition notice it says "for two tubes" So by looking there you have already determined that you have two tubes in total and the load is 8k Plate to Plate, one tube per leg; one push, one pull.

If you had two 6V6s PER OUTPUT LEG in parallel, then you would cut the value in half and have 4k P-P. Another way to say this is to say you have four 6V6s in total.

Just to be sure, you do have two tubes total? If so, then the data sheet says 8K. But as we have determined before, most times a one degree mismatch is OK. I have heard Marshalls don't like this but I can't speak personally about this.

So we are back to using a 16 ohm speaker since you have an 8 and 16 ohm tap. You need to double the connected load to get 7.6k reflected back to the primary. So as Chuck stated in post #11, you really have a 16 and 32 ohm tap since your primary is 3.8k and not 8k. You only have half the required data sheet value with your 3.8k.

[daz]

Yes, 2X6V6 and a 8 ohm speaker. Then i take it i would be well advised to get a more proper OT, and if so how far off from a 8k primary can go with no observable tonal/volume difference from 8k? I linked a classictone 6.6k OT above, (post #12) would that be a good choice?

[Chuck H]

Or...

You could get another speaker! Probably a little more expensive, but what will prove more useful in your future, another OT or another speaker?

There's no dramatic issue using the OT you have. In fact, you can plug it into 4x12 Marshall cab now and be at the correct impedance

Do you have a high-ish watt 10R resistor? Even a 10W at anything between, say, 6R to 12R would do. You'd need to mentally account for the amp being a little lower in output power, but you could get an idea of how the "proper" impedance "feels".

On the subject of proper impedance... The actual ideal impedance is a bit of a misnomer. There will be an ideal impedance for maximum output power and an ideal impedance for minimum distortion. The data sheets don't specify the criteria (but they do show a distortion figure). The two criteria won't produce the same primary impedance. Another view might be that a compromise impedance is ideal. One that has the greatest amount of power for distortion ratio. So that's three possible "ideal" impedance numbers. And they will all depend on your specific tubes, your plate voltage and somewhat on your bias and the dynamic character of your load. In other words... Close is fine and the actual difference between ideal and close is very small. That's why Fender amps have an extension speaker jack that's simply paralleled with the stock load. Doing so with a Fender amp renders it in the same operating conditions as what you are doing now. It's ok, really! But if you MUST hear it, buy a resistor first instead of an OT before spending on an OT. On that matter...

Shipping charges have become pretty stupid. OT's are heavy. If you MUST buy an OT, look up the shipping charge too because it's going to be a considerable percentage of your purchase.

[DRH1958]

Yes, 2X6V6 and a 8 ohm speaker. Then i take it i would be well advised to get a more proper OT, and if so how far off from a 8k primary can go with no observable tonal/volume difference from 8k? I linked a classictone 6.6k OT above, (post #12) would that be a good choice?

Sorry, I forgot to look at that link as I was going to respond to it. Trainwreck amps use this 6.6k OTs because it can run 6V6s and EL34s. That 6.6k impedance value is halfway between the 3.4-4k for the EL34s and the 8K for the 6V6s, so it works for both since it's only a half step mismatch. So yes go ahead and use that 6.6k for your pair of 6V6s. And I like that it has the multi-tap secondary for other speaker options for later. Also, since it has a 30 watt rating, it will have a good amount of iron for the lower output of a pair of 6V6s and give you good bass response.

EDIT: Simulpost with Chuck. Great info there. I just wanted to show the data sheet info way to find the value. Then things can be adjusted as Chuck outlined. That's why this is such an open subject. There's many ways to make things work and much of the time it's "close enough for rock & roll."

[daz]

Well, i'll look at graigs list and see if there are any 16 ohm speakers i might wanna try. Only thing is unless i can find the same model i'm using used/broken in, it's be hard to tell what is making for the difference i hear. As to load resistors, i have some aluminum chassis mounts but i'll have to see if any fall within 6 to 12 ohms. But that would be series your talking about to get 16 ohms....won't series resistance affect the tone a lot? Anyways i'll see what i have.

[R.G.]

Yes, 2X6V6 and a 8 ohm speaker. ... how far off from a 8k primary can go with no observable tonal/volume difference from 8k?

Yeouch!!

This is a large, economy size can of worms.

Back in the Golden Age, the tube designer-guys would design a new tube. They'd then go tinker with the thing so as to find out what the best OT ratio was for it (them?). They would set up a series of "victim" amplifiers and go tinker with loads and then measure things.

It turns out that for every tube, there is a maximum-power loading for a given amplifier setup (i.e. class AB, etc.) If you have, for instance, a pair of 6L6s, the max-power point is near 4000 to 4400 ohms plate to plate. They would then go test the distortion and look at how distortion varied with loading. No surprise, they found that there was a place where the sum of second harmonic and third harmonic (those being the biggest contributors) was a minimum. For the 6L6, this is something like 6K to 6.6K.

So the hifi guys would go load them at 6K-ish for low distortion. The guitar amp guys would fight over power claims, so they'd load them at 4K-ish for better advertising.

Both curves are very, very broad. There is no sharp peak of power at the supposed perfect loading, and no sharp minimum of distortion at the supposed ideal load. But they are differences. It's really, really hard to say how far away from either optimum, or just any other loading point you can go without an observable tone or volume difference. As you know, even very slight tone and volume differences can get noticed.

[daz]

ok, so i have more info that will fit into my tiny brain already and it just keeps coming making for more confusion. So let me make this easy.....is my 3.8k OT using the 8 ohm tap with my 8 ohm speaker with two 6V6's pushing things too far in one direction or another ***based on general consensus*** as you see it, that i would be better off getting a 16 ohm speaker or a new OT? And if i get a new OT what spec would definitely be agreed upon by all. I appreciate all the replies, i really do. But so many things have been said I'm really not sure what the safest bet is. To maybe make it easier to answer, i want enough clean headroom to use it in a band situation for typical small to medium bar levels and have a somewhat clean sound when i roll the guitar back. On the other hand if it's loud and clean as can be, maybe too optimal a spec would be too bright and harsh so maybe consider that.

[Chuck H]

OT at 6.6k for max power. That number will vary with your specific plate voltage, etc. but since you won't find a 6k or 6.8k transformer the 6.6k "standard" value will have to do. That max power thing DOES make a concession for distortion figure BUT THAT IS FOR HI FI!!! For a guitar amp it's the little harmonic anomalies that make the difference between a workman like tone and a more musically expressive tone. We're not talking about clipping distortion here. The stuff that guitar players associate with "clean" vs. "distorted". We're talking about insinuated harmonics. Like the root tone has a (very slightly) increased ratio of the third harmonic in the content of the final output. This would be measured in percentages in the few, or even decimals. The greatest output is the typical choice for guitar amps, so... This is the sound that our ears have become acclimated to and the sound that guitar pickups have been designed to work with and the sound that studios expect and the sound that sound engineers monitoring stages have become familiar with... Get what I'm sayin'??? If you consciously veer from the norm you'll be into less explored and less accessible territory. So... 6.6k

Yet you like the tone at 1.8k primary!!! This is where tone becomes subjective. You absolutely WILL hear and feel the tone move further to one direction as you go from 1800 to 3800 to 7600 (or 6600). In this light 6600 goes further away from the tone you already preferred! Now, 1800 is an unreasonable load for a pair of 6V6's. Just play the rig at 3800 primary, which is reasonable, and be happy.

[daz]

Will do Chuck. Thats sounded reaffirming enough for me. It does sound great so i'll just leave it. I just can't leave something alone if i'm unsure whether it could be robbing tone or in this case volume and tone. But i just ran it w/o the delay in the loop and it gets freakin loud and cleans up well. It won't get loud enough to gig with the delay (it starts blocking in the DSP well before loud) tho so i gotta get that loop working. On that note Chuck, can you think of a amp similar to mine with a single AX7 powered loop that would work where you put the SS loop into my schematic? Gotta be before the master so that would be the only place i can imagine.

By the way, i just noticed the name of the jpg you posted of my schematic with your loop idea and i actually laughed out loud !

[Dave H]

Just play the rig at 3800 primary, which is reasonable, and be happy.

If it's loud and clean enough I'd run it like Chuck says. Anyway the speaker impedance is only 8ohms over a narrow band of frequencies. It will be greater than 8ohms above and below that so the 6V6s will also see an impedance greater than 3800 for higher and lower frequencies (see graph below)

[Chuck H]

Indeed!

I test a little peripheral stuff whenever I have an amp on the bench and I'm often surprised at how little difference a 100% mismatch makes in output power with an actual speaker load. Usually it'll be virtually NO difference or a small increase mismatching down and only a few percent loss mismatching up. An interesting observation is that with a resistive load there is sometimes more power with a 100% mismatch up and a few percent loss mismatching down.?. I speculate this is because the white coats that designed the tube, transformer and speaker relationships long ago were paying attention to the fact that an actual loud speaker presents a higher average impedance at frequency than a resistive load, so, the efficient ratios we mimic today are grandfathered in. Testing with a purely resistive load is more about detecting operational faults than actual musical performance. The fact that RMS watts are standardized from a resistive load not withstanding.

So, daz, in my humble observation you may be getting more power with the 3800 primary than you would with a "proper" 6600 primary when connected to an actual speaker (at some relative increase in harmonic distortion).

EDIT: Hey daz, for more peace of mind consider that a very popular Marshall mod is to swap the el34's for 6V6's for. Typical primary impedance in these Marshall's is pretty much right where you're amp is.