Smartest way to 200W?

[trobbins]

No your measurements ARE useful, but the devil is in the detail as usual. You describe the screen voltage measurement was at sagged level/conditions. Was the screen current measured under the same sagged conditions?

Multiplying those voltage and current levels give you an upper limit on screen dissipation, but I reckon the real screen power dissipation level will be less due to waveform distortion (and its affect on power factor). Unless you have a specialist power meter, or digital oscilloscope that can do the right maths functions, then the next best thing would be to capture a waveform of the screen voltage and current waveforms somehow when at max output.

[Gaz]

No your measurements ARE useful, but the devil is in the detail as usual. You describe the screen voltage measurement was at sagged level/conditions. Was the screen current measured under the same sagged conditions?

Yes, the screen current and voltage were measured under the same sagged conditions (wasn't sure how to phrase that before). It's good to know the measurements aren't worthless, and a bit conservative if anything. Thanks for spelling that out.

Out of curiosity, Tim, what would be your approach to getting more power out of this circuit without killing the tubes? Raising the plate voltages, screens, a little of both? I opened up your loadlines in Illustrator and tried to draw my own, but came up with something way over the knee, and of course it just goes higher and higher with increased voltage. I'm just not sure how to factor in the lower screen voltage.

[trobbins]

Alex, if you have MS WORD then I can upload the raw doc format if that is of any help. I just use the draw package in Word and drop in cropped gifs taken from pdf datasheets.

I guess for bass, you're after relatively clean raw power, rather than output limiting due to saggy supply rails. As such, it sounds like just reducing your screen stoppers would achieve 200Wrms, as long as you have confidence about screen dissipation. And B+ sag may be somewhat related to supply filter capacitance - I have 700uF in my half-built 200W, whereas you effectively have 100uF.

With respect to hitting the screen dissipation limit then if in fact you have hit that limit, but still want more Wrms, then all you have to play with is B+ and optimising the knee location. In reality, your knee position could be all over the place depending on your speaker loading in the frequency range of interest (ie. some bass speaker loadings could be varying dramatically, depending on what resonance peaks are in play - but that depends on speaker/box config).

I'd suggest initially detailing and tweaking your sagged screen and plate performance, and look at load matching, rather than initially heading off to make changes to default voltage levels.

Have you got any over-voltage protection on your KT88 plates?

Ciao, Tim

[Gaz]

Alex, if you have MS WORD then I can upload the raw doc format if that is of any help. I just use the draw package in Word and drop in cropped gifs taken from pdf datasheets.

I guess for bass, you're after relatively clean raw power, rather than output limiting due to saggy supply rails. As such, it sounds like just reducing your screen stoppers would achieve 200Wrms, as long as you have confidence about screen dissipation. And B+ sag may be somewhat related to supply filter capacitance - I have 700uF in my half-built 200W, whereas you effectively have 100uF.

With respect to hitting the screen dissipation limit then if in fact you have hit that limit, but still want more Wrms, then all you have to play with is B+ and optimising the knee location. In reality, your knee position could be all over the place depending on your speaker loading in the frequency range of interest (ie. some bass speaker loadings could be varying dramatically, depending on what resonance peaks are in play - but that depends on speaker/box config).

I'd suggest initially detailing and tweaking your sagged screen and plate performance, and look at load matching, rather than initially heading off to make changes to default voltage levels.

Have you got any over-voltage protection on your KT88 plates?

Ciao, Tim

Tim, thanks a lot for the insight, it's appreciated as usual. I could actually move your loadlines in Illustrator, and I assumed they must've been drawn in Illustrator. So thanks a lot, but no need to send the files. I just realized I should probably be using the sagged B+ voltage to draw the loadline, is that right? I think that would put me closer to the knee.

I do have MOVs from each plate to the CT per your suggestion

[trobbins]

Effectively you know the steady-state idle point and the loadline direction close to that point - which you have best represented on a screen=300V plot of curves and a class A line. The other point you are trying to estimate is the Vg=0V point which I reckon would be better represented on say a screen=210V plot of curves with a class B line crossing at the sagged B+ voltage level. In your situation, you could also try screen =250V and 300V curves as well to get a feel for the variation that screen stoppers may present, as well as the degree of 'steady-stateness' (given that B+ may not sag as much for a slap or other short duration peak). The sagged B+ should also be reduced by an amount equal to the OT primary half resistance multiplied by the likely peak total (2x) plate current level.

[Austin]

What about 6 kt88? This would reduce the output impedance even further. Since you are going for clean headroom what about getting another output tranny with ultra linear taps and running the screens from that solid source? You could also use it for your home hifi sub woofer.

[J M Fahey]

Don't forget that a modern KT88 is not the same as an old one by a long way.
Datasheets usually show the old data.

[Tage]

I agree with austin, running 6 tubes is a "smarter" way to get there. although I have a build in the tube with a very similar set up to yours, looking to get 200 watts, but I'm also using a B+ around 620 and mosfet source followers to drive the 4 KT88's into grid conduction.

I have been asked to do another Bass amp and think I will go with 6 tubes and lower B+ and a load of around 1.2k ish

[Steve Conner]

Yes, I guess 6 new KT88s are about equal to 4 original GECs

[J M Fahey]

Yes, I guess 6 new KT88s are about equal to 4 original GECs

Yes, that's the point.
Building a "200W Hiwatt" and hoping it delivers with modern "KT88" (which they are not) is being naïve, simply not the same tube.
Also using the old datasheets on the new ones does not match.
I made my own power tube tester and *measure* the actual plate and screen currents at saturation.
Can't miss.

[GregS]

I would say that your power transformer that's providing the low plate and screen voltages is what's limiting your power to what it is. I think you said that was salvaged. I think that with a transformer set of similar specs and power supply voltages to the Hiwatt, you'd be closer to 200W. I don't see those voltages being any big deal for KT88's old or new provided the amp is biased reasonably and has reasonable screen stopper resistors.

As a real life example I built a bass amp recently with a power supply and power amp section inspired by the Traynor Custom Special. It uses 6550's or KT88's and I'm currently using 6550's. At idle the plates are at 567V and screens at 560V - considerably above the max recommended rating for the screens. However, with 1K screen stopper resistors and biased reasonably this seems to be totally fine.

Also, with typical bass amp usage the power amp is not being pushed much at all as most bass is played clean. There can be large spikes in the signal occasionally which is where the usefulness of 200W or more comes into play - repsonse remains firm and doesn't fart out. Although there may not be too much difference in overall loudness between 170W and 200W it may make some notable difference to being able to maintain a firm response without flabbing or farting out.

Another thing. When I built my amp I initially wasn't getting the power out of it I should have been. It turned out I was running out of headroom at the phase inverter rather than in the power amp. It may be a good idea to scope before the power amp and make sure your not running out of headroom somewhere else before assuming it's the power amp.

Greg

[Gaz]

I would say that your power transformer that's providing the low plate and screen voltages is what's limiting your power to what it is. I think you said that was salvaged. I think that with a transformer set of similar specs and power supply voltages to the Hiwatt, you'd be closer to 200W. I don't see those voltages being any big deal for KT88's old or new provided the amp is biased reasonably and has reasonable screen stopper resistors.

As a real life example I built a bass amp recently with a power supply and power amp section inspired by the Traynor Custom Special. It uses 6550's or KT88's and I'm currently using 6550's. At idle the plates are at 567V and screens at 560V - considerably above the max recommended rating for the screens. However, with 1K screen stopper resistors and biased reasonably this seems to be totally fine.

Also, with typical bass amp usage the power amp is not being pushed much at all as most bass is played clean. There can be large spikes in the signal occasionally which is where the usefulness of 200W or more comes into play - repsonse remains firm and doesn't fart out. Although there may not be too much difference in overall loudness between 170W and 200W it may make some notable difference to being able to maintain a firm response without flabbing or farting out.

Another thing. When I built my amp I initially wasn't getting the power out of it I should have been. It turned out I was running out of headroom at the phase inverter rather than in the power amp. It may be a good idea to scope before the power amp and make sure your not running out of headroom somewhere else before assuming it's the power amp.

Greg

Thanks, Greg. I tool your advice and scope'd the PI. It's does seem like it's running out headroom just slightly before the output. Nothing drastic, but the waveform does begin to round off first. Any suggestions to raise the headroom of the PI? I can't raise the B+ high due to the "stacked" power supply. I'm using a 12AT7 currently.

Thanks!

[es345]

For a test:

change the common 22K tail resistor to 18K. The PI will run a little bit hotter.
The result is that you have more headroom at the "positive half" of the signal at the plates of the PI.

[es345]

double posting ...

[GregS]

Thanks, Greg. I tool your advice and scope'd the PI. It's does seem like it's running out headroom just slightly before the output. Nothing drastic, but the waveform does begin to round off first. Any suggestions to raise the headroom of the PI? I can't raise the B+ high due to the "stacked" power supply. I'm using a 12AT7 currently.

First of all, it would be good to confirm where it is running out of headroom first - at the PI or the power amp. If you have a dual trace scope you can scope the output of the PI and the power amp simultaneoulsy. As you turn the volume up if the output of the power amp clips while the PI is still clean then the headroom problem is in the power amp. If they distort at the same time and the two wave forms are identical in shape then it's likely running out of headroom at the PI.

Is there not a dropping resistor in your power supply between the screen and PI nodes? If so then raising the supply voltage to the PI is simply a matter of changing that resistor for one of less resistance. Changing that one resistor will of course also raise the voltages downstream in the pre-amp. If you wanted them to be unaffected you'd have to change the resistor after the PI node to one of higher value. If you determined the headroom problem is not at your PI then there's no reason to do any of this of course.

Hope this helps,

Greg