Does valve watts relate to amp output watts?

jpfamps

ecc83 said:

"Surely the THD goes up as you approach clipping, i.e. by the time you're clipping THD is high (by definition)? If I understand correctly (which may not be the case), valve power will continue to go up as THD increases while solid state designs typically are operating near their ceiling power before THD starts to rise (and then it rises fast), due to the higher level of negative feedback in most solid state designs. Or do you mean clipping because they've hit protection?"

Yes, sorry! My point should have been that it might be difficult to FIND a 10% THD point for many sstate designs. They tend to produce buggerall distortion then whammo! Off the scale. A valve amp with little or no feedback can be pushed progressively, 1,2,3,....10% and so on.

Err? What did that Service man Terminator SOUND  like J?

Dave.

Amp sounded loud!

It's really a PA amp not a guitar amp, and was being used as such.

pickerg

ICBM said:

5881s should be OK anyhow, but a 5V4 will help. You do need to get the power down a bit even for a 5V4 to take the current though - it will be at the upper limit (175mA). Not sure it will drop to as low as 60% of full power though.

For illustration, in my Mesa Trem-o-verb the plate voltage is 445V in solid-state diode mode and the power output is 100W - I didn't measure it with GZ34s but it will only be a bit less than this. With 5V4s (which are right at their limit, at 350mA with two valves) I get 420V and 80W.

Ok thanks, I follow everything apart from the current limits of the different rectifiers.  Regarding this limit for the rectifier (175mA in this case), is this something that is physically measured, or worked out using the Ohm's law calculations when working out the bias settings?

Cheers.

ICBM

You need to measure it with the amp running at full power - you can't work it out directly from the current at idle.

The problem with trying to calculate it is that the current draw is non-linear and depends on things that are hard to measure directly, eg the internal resistance of the power transformer winding. As the current draw of the power valves rises when the amp is driven harder, the B+ voltage falls due to the internal resistance of the PT and the forward resistance of the rectifier (which is itself a non-linear function of the current, doubling the complexity of the calculation), which in turn then reduces the maximum power available and hence eventually limits the maximum current draw.

So a rectifier with lower forward resistance like a GZ34 will produce more power from the amp and hence must be rated for a higher current draw as well, if that makes sense! (They're 250mA.) Without trying it in the Bassman - measuring the current draw while slowly increasing the power, you'll probably want a dummy load for that - I don't know whether the maximum draw with a 5V4 in will be less or more than 175mA, although it will certainly be *around* that...

ecc83

Or use 1N4007s and forget all that bollocks.

Ok! I do not like thermionic rectifiers, right? (but I know there's them as do.)

Dave.

ICBM

I kind-of wish I didn't, because they are an apparently unnecessary extra potential source of unreliability, but the unavoidable fact is that they do change the sound and dynamics of the amp, and I much prefer them*. I even prefer a real valve to a valve-emulating solid-state one (ie diodes with series resistors), although the difference is more subtle… I think it must come down to the non-linear forward resistance curve of a real valve, which are quite interesting especially when approaching the current limit.

*Usually. There are amps which I prefer with solid-state - a good example is the Orange AD30, which needs one anyway due to the same standby-switch issue, and actually sounds best (in my opinion) with a straight non-emulating SS plug-in, much tighter and clearer than it does with the valve.

What I find really bizarre is that I come across quite a lot of old amps which have been modified by techs in the past, by not only fitting solid-state diodes soldered to the valve socket inside but sometimes actually disconnecting the valve filament supply! Why?! There's absolutely no need, you can simply fit the diodes. Best of all, make a plug-in - I put them inside an old valve base filled with resin to keep fingers away from the live bits - so you have both options. Fender actually used to ship the Bassman reissue like this, with a solid-state rectifier in an otherwise fully-functional valve socket. Best of both worlds and gives the user the option.

ecc83

Well, there you go IC!

You are what, 75% of the way to an acceptable solid state rect "sound"? There must be a way to "bend" things that little bit more to get the final polish?

Do that, patent it (well don't actually just encapsulate it so it cannot be copied!) and you have your pension sorted!

Re disconnecting the heaters when fitting diodes across the valve. I suppose this is a bit of a con? They want to appear "tubified" but are not. One could be generous and say that the modder may have upped the res' cap beyond the valve's rating?

FYI I found the TAD GZ34 pretty indestructible.

Dave.

ICBM

ecc83 said:

You are what, 75% of the way to an acceptable solid state rect "sound"? There must be a way to "bend" things that little bit more to get the final polish?

I would think so, although it would take a bit of extra complexity to simulate accurately because a valve rectifier has an odd forward resistance characteristic - it actually decreases with increasing current until close the upper limit, then rises sharply. I think this is what causes the compression and then so-called "bloom" (sorry for the silly term, but it seems to be commonly used and is at least reasonably descriptive!) of a valve-rectified amp. Even using series resistors with diodes doesn't quite get it - although it's so close that without being able to compare them in the same amp, it's debatable whether anyone would really be able to tell.

To be honest though, as long as you're using valves at all - certainly power valves - I don't see much point in going to a lot of trouble to eliminate the rectifier... it doesn't really solve many problems - apart from where designers seem to have forgotten (or never found out) how to implement a standby switch correctly! There's nothing really wrong with either, it's just when you put them together in the wrong way that it causes trouble.

ecc83

I just dislike the added cost I suppose.

Not only are they many, many times the price of a bridge rectifier with 5 times the rating but they call for two HT windings on the mains traff with a consequent loss of efficiency. Takes about ten watts to heat the bloody things AND yet another traff winding!

Dave.

ICBM

ecc83 said:

I just dislike the added cost I suppose.

Not only are they many, many times the price of a bridge rectifier with 5 times the rating but they call for two HT windings on the mains traff with a consequent loss of efficiency. Takes about ten watts to heat the bloody things AND yet another traff winding!

You could argue that about using valves at all, though .

ecc83

ICBM said:

ecc83 said:

I just dislike the added cost I suppose.

Not only are they many, many times the price of a bridge rectifier with 5 times the rating but they call for two HT windings on the mains traff with a consequent loss of efficiency. Takes about ten watts to heat the bloody things AND yet another traff winding!

You could argue that about using valves at all, though .

Ooo! I wouldn't got THAT far!

Dave.