Parallel single ended 6550 amp question

[Jarrodthebobo]

Hello everybody,

A few months ago I built a single ended plexi that has gone through so many modifications that it's now not only able to switch from a jcm800 to a superlead, to a friedman, and to the #34 SIR amp, but has two 6550 tubes in parallel through a huge beefy output tranny.

I was just curious as to how much wattage this thing should be outputting. Last time I checked the amp using el34s, I got plate dissipation of 28 watts (hence why I started using 6550s). The plate voltage is 390v, 348v from plate to cathode. Cathode bias resistor is 250ohms, and the output tranny uses a 2.5kohm tap @ 8 ohms.

This amp has no volume difference to my ear versus my dean markely t-60 signature series amp which is a push pull 6l6gc jcm800 clone for the most part supposedly outputting 60 watts.

Although I highly doubt this amp is hitting anywhere close to that, being that the output tranny is rated for 30w maximum, I'd assume it'd half to be somewhere close to that maximum in order to be not noticeably different in volume through the same cab.

I also don't really notice much of a difference between headroom between the two amps either.

Any thoughts?

Thanks!

[shooter]

348v from plate to cathode. Cathode bias resistor is 250ohms

measure Vdc at top of cathode R (Rk), use I = E/R, then * 348. walla, Pdiss (Idle Power in Watts), or, measure Vac-rms at speaker, square it, then divide by speaker R, walla, Audio power out.

[Jarrodthebobo]

348v from plate to cathode. Cathode bias resistor is 250ohms

measure Vdc at top of cathode R (Rk), use I = E/R, then * 348. walla, Pdiss (Idle Power in Watts), or, measure Vac-rms at speaker, square it, then divide by speaker R, walla, Audio power out.

So if I'm doing my math right, the Idle power is around 29 watts? Voltage at the cathode is 21.3v; so 21.3/250 =.0852 * 348 = 29?

I'm not too sure how to go about measuring output at the speaker on the other hand; all I really have for testing gear is a multitude of digital and analogue multi-meters, haha.

Thanks for the help!

EDIT: Since I have a plate voltage of 390v, and I'm measuring 21.3v at the cathode, shouldn't plate to cathode be closer to 368.7v?

[sluckey]

First you said...

The plate voltage is 390v, 348v from plate to cathode.

...which would mean the cathode actually has 42V on it.

Now you say...

Since I have a plate voltage of 390v, and I'm measuring 21.3v at the cathode, shouldn't plate to cathode be closer to 368.7v

...So which numbers are correct?

[Jarrodthebobo]

Hmmm, you're right. This is strange. These are numbers I wrote down from a few weeks ago so I'm going to have to crack her open again and recheck my work.

With 42v on the cathode I'm now looking at 54 watts idle, which seems quite high

[PRR]

It's around 20 Watts, say 16W clean and 25W pretty bent (there's a broad zone where THD rises from small to large).

You are not stressing 6550s. OTOH with the parts you have, there is no significant power increase available, nothing to move you from the club to the mega-stadium, so no point running them hotter.

[shooter]

54 watts idle

IF you have 1 R and 2 tubes - divide by 2

Like PRR said!

I build PSE, 20-25W Idle dissipation/tube is about all there is, unless you're DummyLoad 

clean audio power ~ < 1/2 Pdiss
use a sig-gen, start ~ 1khz 100mV rms AC in, scope speaker, adjust sig in, volume, tone etc till you get max clean sinewave, math and you're close enough

[Jarrodthebobo]

54 watts idle

IF you have 1 R and 2 tubes - divide by 2

But does that still hold true if I'm using two separate bias resistors? I have a 250ohm going to the cathode of each tube individually

[PRR]

I'm very confused now.

Measure and report all your numbers accurately.

[Jarrodthebobo]

I'm very confused now.

Measure and report all your numbers accurately.

I'm sorry;

The amp has 2 6550s; each 6550 has its own seperate 250ohm bias resister which is bypassed by a 1000mfd electrolytic cap.

When I last measured voltages, I took a measure from cathode to ground (on both tubes) and got 21.3v on my meter. The plate was at 390v, whereas cathode + plate was 348v (on both tubes).  When I return home for the weekend I'll double check the measure on the cathode to see what the value actually is.

[pdf64]

...clean audio power ~ < 1/2 Pdiss...

Just to clarify, the idle dissipation takes no account of the plate load value suitability, so the actual signal power output could be a small fraction of it.
Merlin explains the theory of operation here http://www.valvewizard.co.uk/se.html and nickb's interactive valve data sheet will allow you to plug in the numbers and obtain a theoretical max (ie assuming bogey tube and lossless OT) power output and plate dissipation http://bmamps.com/ivds.html

[DummyLoad]

...clean audio power ~ < 1/2 Pdiss...

Just to clarify, the idle dissipation takes no account of the plate load value suitability, so the actual signal power output could be a small fraction of it.
Merlin explains the theory of operation here http://www.valvewizard.co.uk/se.html and nickb's interactive valve data sheet will allow you to plug in the numbers and obtain a theoretical max (ie assuming bogey tube and lossless OT) power output and plate dissipation http://bmamps.com/ivds.html

bmamps.com is a cool site. thanks for sharing. 


--pete

[Jarrodthebobo]

...clean audio power ~ < 1/2 Pdiss...

Just to clarify, the idle dissipation takes no account of the plate load value suitability, so the actual signal power output could be a small fraction of it.
Merlin explains the theory of operation here http://www.valvewizard.co.uk/se.html and nickb's interactive valve data sheet will allow you to plug in the numbers and obtain a theoretical max (ie assuming bogey tube and lossless OT) power output and plate dissipation http://bmamps.com/ivds.html

I've looked over the Merlin sites before in the past but never really delved too far into them as I always found myself somewhat confused from some of the concepts presented. However, I made myself sit down and digest all the information this time and I think I understand a bit more than I did prior.

The only real question I have now is this; is my primary impedance on the output tranny of 2.5k appropriate? Using the given formula to bias for center operation I'd get 390^2/28 = 5432ohms (28v is just what I have the amp currently biased at; I'm sure I could increase this quite a bit since 6550s can handle quite a bit more than that). I'd take a guess that parallel tubes you'd divide by 2 to get the appropriate impedance rating  of close enough to 2.5k?

I have to take a look at all my voltages again anyways; but I'm sure I could probably tweak the screen voltage as well; I've just been using stock values when building the power supply so I'm sure it's not optimized. I'm pretty sure I'm using a 1k dropping resistor currently going to the screen, along with a 1.5k grid suppressor.

EDIT: Alternatively, would there be any added benefit with dropping the cathode bias resistor closer to 200ohms, or perhaps even lower? If I'm truly only at 28w plate dissipation and the 6550s can handle around 40 watts, wouldn't I get quite a bit more volume and headroom biasing the amp hotter?

[shooter]

any added benefit with dropping the cathode bias resistor closer to 200ohms

It's always worth tweaking, now that you have a handle on the numbers, you'll find as we  xSE builders have, there's a diminishing return on investment, lowering Rk increases current, drops plate volts, and you're back where you started.  You'll find the tweaks are more about sound than tube sweating, so each tweak should be played once you know the #'s are below max disipation

[Jarrodthebobo]

any added benefit with dropping the cathode bias resistor closer to 200ohms

It's always worth tweaking, now that you have a handle on the numbers, you'll find as we  xSE builders have, there's a diminishing return on investment, lowering Rk increases current, drops plate volts, and you're back where you started.  You'll find the tweaks are more about sound than tube sweating, so each tweak should be played once you know the #'s are below max disipation

Honestly I took on this project to push what you could do with a Single Ended design, and I keep finding that I probably should have just made a PP amp  .

I am happy with the amp by all means! It's WAY loud enough for my purposes, sounds really good, and is a fairly cool/odd design... a one of a kind? Perhaps not; but a one of not many.

Really all I want to get out of this amp is perhaps slightly louder clean volume; right before breakup the amp can get 'considerably' loud, ie it'll shake various things in the basement and throw things off the top of the cabinet when they're forgotten there... but I'm not certain that it's loud enough to play with a drummer. Granted I havn't played with an actual band in quite a while, so that's not a real issue either atm.

Distorted playing is definitely loud enough to play with a whole band however. My ears ring rather horribly after playing the amp pushed for more than a couple of minutes. I've gotta get myself a decibel meter; I've tried measuring the amp using my phone but it seems the mic's in built compressor stops at around 109dbs

EDIT: I currently have a Hammond 272fx installed; would it be worth it to get a 273bx for the extra 50v, giving me around 493v after rectification? Or would the performance difference not really be enough to notice?

[shooter]

would it be worth

perhaps slightly louder clean volume

You're at the point in a build where a scope and sig-gen come in real handy, they allow you to "fine tune" each stage.  by adjusting voltage dividers, you can set how much drive, clean OD, get moved on to help get where you want.
and stay with the SE world, the PP world is way over rated 

[PRR]

> I'd get 390^2/28 = 5432ohm

OK...

EDIT: Alternatively, ...the 6550s can handle around 40 watts, wouldn't I get quite a bit more volume and headroom biasing the amp hotter?

Same load, same B+, you'd get more heat but no more power. You need to keep the V/I about equal to the Z.

For 5k load/tube and 40W Pdiss it looks like 447V plate-cathode. Plus cathode bias, plus OT losses.

The difference is 1.2dB which is NOT "get quite a bit more".

With cheap bottles I may endorse running ALL the Pdiss rated. But I have paid $100 (and more) for a couple 6550; I'd rather stay short of the max rating.