Voltage drop/sag under load

[EFK]

I have a 2X6L6 amp based on a Supro which is cathode bias, idle plate voltage is @ 390 from a 5V4 rectifier and it idles at @ 24W dissipation (actually I'm using 7581A right now). Injecting a 1Khz signal, and running into a 16 ohm resistive load, I notice that the voltages throughout the amp drop b/t 35 and 40 volts at full bore. I tried a 5U4 and a GZ34 as well, and while the idle voltages changed a little, all three rectifiers exhibit the same voltage drop at full draw. This is into a 4.2K primary. MEasuring the cathode current, I get 238 mA so I figure the plate current is what - about 200-210 mA? This is from a 'low voltage' 300-0-300 JTM45 PT. I deliberately wanted some sag, and I've sure got it, but does this seem like too much voltage drop? Shouldn't the GZ34 have dropped less voltage under load as it can handle much more current (I realize the 5V4 might be struggling to keep up here)? Just wondering if it sounds like something is "off" here. Maybe the 4.2K primary is too low - tubes are struggling? Would a something like a 5 or 6K OT w/ a little more resistance make a difference here? Just throwing this out there for some ideas - it's really hypothetical. I finally made myself a resistive load and can begin to measure this kind of stuff, so I don't have any base of reference yet. BTW, the AC voltage hitting the grids out of the paraphrase inverter is 63.4 VAC on one tube and 54.9 VAC on the other. AC at the 16 ohm load at full power w/ the 1 khz is 27.2 VAC.

[Jazz P Bass]

That 27 Vac output into a 16 ohm load is pretty impressive!
You will have a voltage drop.
You cannot have current without it.
How much, I don't know.

[redelephant]

Yea, 45watts will drag down the b+ some.
My guess is u have some voltage drop over the power tranny as well, probably enough to be the dominant 'sagster' which is why different rectifiers don't do much difference in sag. Measure the AC output on the tranny's secondary.
10% sag isn't that much is it? How does the amp respond? If u feel the compression is too much, well, guess it sag too much for you. I bet it's just fine.

[guitarmike2107]

You don’t just get sag from the rectifier, by pulling more current through the transformer there is also a voltage loss across the windings of the PT, also what’s the current rating of the PT, if your pulling more current than it is rated at it will certainly sag the voltage.

How much sag is normal? or desired.

Check out this thread from AX84 talking about Marshall load lines and how much they sag under full drive.

AX84.com - The Cooperative Tube Guitar Amp Project

Specifically Mr Aiken said “

It just so happens I have a JVM 100W amp in here that a friend asked me to repair. It has 500V on standby, 480V on the plates at idle, and sags to...get this...408V at full clip! The screens sag from 466V at idle to 340V at full clip!

I am measuring right at 100V p-p into 16 ohms, for an output power of around 80W when the clean channel hits clipping.

I guess that's one way around the redplating issue...

RA”

[pdf64]

'the voltages throughout the amp drop b/t 35 and 40 volts at full bore'

'AC at the 16 ohm load at full power w/ the 1 khz is 27.2 VAC'

That level of sag is in line with what I'd expect.
By full bore, do you mean max unclipped sine wave or max square wave (square I'd assume). If square, your meter needs to be a 'true rms' type for the reading to be valid.
If it sounds ok then one option is not to worry about it, but if you want to investigate then try alternative values of load, eg 24 ohms should be reflected back as 6k, and see how it performs. Pete.

[EFK]

Thanks guys - like I said, this is the first time I've been able to measure these things so I don't have any frame of reference. It doesn't sound weird or anything like that, I simply had no idea what kind of sag to expect via a meter rather than simply audibly. Good point about the sag via the PT rather than solely the rectifier tubes. Yes, full bore is full on volume all the way up and I am using a true rms meter. What I want to experiment with now are different OT. This is a bassman OT and it measures at 4.2K. I did try 8 into 16 for 8K but I only measured the AC output on the tube grids and it went down quite a bit. What I really want to try is something in the 5K-6K range but I'll have to get some additional resistors.

The thread on AX84 linked above was VERY interesting!

BTW, might I see some variation in sag if I were to substitute the current 1K resistor between the first two filter caps with a 'typical' size choke? I realize that the voltages will change slightly at idle, but I'm not clear on whether or not a choke would affect the voltage sag at full volume. I suppose it would allow me to use a bit more preliminary capacitance which might affect the sag when the amp starts really drawing power?

[guitarmike2107]

Interesting, I was recently reading how when you overdrive a fixed bias amp it moves toward class A operation, or the load line moves up the graph, but my understanding from these discussions is that if the power supply sags allot it would counteract this i.e. move the load line down the graph.

Then apparently when overdriving the power amp in a cathode biased amp it moves towards Class B operation, so if the power supply sagged allot it would it exaggerate this effect… or probably actually lesson the effect as the sag would reduce the current through the cathode resistor.

OK, I am confusing myself now…… but it gives the term “power amp dynamics” new meaning to me anyhow.

[pdf64]

'I was recently reading how when you overdrive a fixed bias amp it moves toward class A operation'
Can you confirm that? My understanding is that generally, blocking distortion causes the effective bias voltage to increase, ie move towards / past class B. Overdrive it hard enough and the bias setting becomes irrelevant, as the signal input becomes more significant that the bias and tubes just switch from full off to full on, effectively in class C. Pete.

[guitarmike2107]

Pete It was from Kuehnels power amp book book, I don't have it here to reference, but I will dig it out tonight. He was making the point that fixed bias and cathode bias amps react differently when overdriving, I am sure I got that right, I don't think sag was brought into that discussion, which obviously has a HUGE effect, but I have only read parts of the book ... I need more hours in the day.

BTW I think its bias excursion, not blocking distortion, we want to avoid blocking distortion.

[redelephant]

It makes sense that the fixed biased amp moves towards class-A when overdriven b/c the sag in the power tranny makes the bias voltage shift to a less negative value, which is like biasing the amp hotter. So the tubes conduct thru more of the cycle, and exaggerates the current drawn from the supply and thus the sag.
A cathode biased amp I would think has the opposite effect, the bias voltage being more constant, if anything increasing and so bringing the operation toward class-B or C.
Interesting...So who has tried using a voltage regulator for bias, which will keep the bias voltage constant even if the supply sags? My guess it wont have a huge effect and not too worried about it.

[Steve Conner]

Sure, but there's also the opposite effect, where the power tubes' grid current charges up the coupling capacitors, causing them to bias themselves colder. I think this is what's referred to above as bias excursion, or blocking distortion. I guess it'll be a matter of which effect wins out at a given level of drive.

If you regulate the bias, you have to regulate the screen voltage too, or the bias stability against line voltage variations will get worse. My ancient Ninja Toaster amp had (has? it is still going strong) regulated everything.

[guitarmike2107]

It makes sense that the fixed biased amp moves towards class-A when overdriven b/c the sag in the power tranny makes the bias voltage shift to a less negative value, which is like biasing the amp hotter.
.

Like I said I don’t think he was taking into account the power supply sagging, IIRC he was referring to how the bias circuit capacitors actually charge up as a result of grid conduction.. but before I spread anymore tripe I better go read the section again!

[guitarmike2107]

So was the Ancient Toaster Ninja quite a stiff power section?

[redelephant]

Ah, yes there is that to think about as well I am so used to using either my hybrid PI with direct coupling to the output tubes, or intestage transformer coupling, both of which don't have that problem...therefore didn't think about that.

[Steve Conner]

Guitarmike: Kind of, I guess. You don't get any bounce or squish from the HT rail varying, but the power tubes themselves still compress. Just now I'm using 6550s in it, and they sound much the same as the master volume is turned up, until all of a sudden they lose it and mush out. At that point it's louder than required for any gig I've ever played with it. So I guess it works in a master volume-y kind of way.

Being a master volume design it gets most of its tone from the preamp, but I tried playing the preamp through a solid-state power amp, and it didn't sound nearly as good, even at low volumes. So the power tubes and PI seem to be doing something.

It was somewhat of an experiment. Given the extra trouble, not to mention the fan noise from keeping those MOSFETs cool, I probably wouldn't ever build a fully regulated amp again.

You are welcome to come to Glasgow and try it out any time.