Orange OR200 the definitive maintenance thread

[Helmholtz]

Is your meter true RMS? If you measure the RMS screen current and RMS screen voltage then you can simply multiply them to get the screen dissipation.

I see 2 problems with this method:

1) Most "true RMS" meters are AC coupled and ignore the DC content, which contributes to real RMS and power.
2) The different shapes of screen voltage and current probably cause a power factor <1.

[Dave H]

Is your meter true RMS? If you measure the RMS screen current and RMS screen voltage then you can simply multiply them to get the screen dissipation.

I don't think that works with screen current and screen voltage being out of phase. Didn't we decide in that other thread a while back that using DC average values for current and voltage gave a better approximation for screen dissipation?

[nickb]

[QUOTE= screens 560V, 7.2V over the 470 ohm limiter meant 15.3mA x 560 = 8.578 W quite in border as seen on the scope, and this is the point where the amp start to clip[/QUOTE]

Is your meter true RMS? If you measure the RMS screen current and RMS screen voltage then you can simply multiply them to get the screen dissipation.

What you say 'lower value', what was it you lowered the value of?

[catalin gramada]

Lower value did not help at all. Changed to 470 ohm and drive to the border of clipping. Plate 567V, screens 560V, 7.2V over the 470 ohm limiter meant 15.3mA x 560 = 8.578 W quite in border as seen on the scope, and this is the point where the amp start to clip

[nickb]

... think I get it. It is the same if power tubes are drived with sine or almost square signal ,please ? You.ll never know what signal have in front...

In this context something vaguely musical is to be expected. The amp been in use 50 years and it hasn't been a problem in practice. I think that is significant.

There is nothing wrong with designing it to be bombproof but it can make then very heavy and expensive. I suspect that with the sort of overload you're giving it the power transformer will toast before too long. Seems better to limit the drive or reduce the screen supply. I had to take the drive limit approach for one customer.

Not having KT88's with a good screen current model to hand for sim, I used 6L6's. With a 540V supply, 2.2K load at hard clipping the screen dissipation was 7.4W with a 270 ohms and 6.8W with 1K ohms. Power out dropped by 4%. Less than I expected.

The supply drop from 630V to 530V is a lot. That's something like an effective resistance of 100 ohms. I suggest such a large drop is another indication that it's being pushed too hard,

[catalin gramada]

... think I get it. It is the same if power tubes are drived with sine or almost square signal ,please ? You.ll never know what signal have in front...

[Helmholtz]

I.m not talking what is happen up to clipping, but over instead. My note comes from direct dc measure in my circuit.

As a rough approximation you could maybe use averaged DC values, not sure with that high degree of voltage modulation. But I would expect the time averaged screen voltage with clipping to be considerably lower than 550V. And I have some doubts your meter could do correct time averaging without some additional filtering.
Maybe nickb knows better.

I am not saying that increasing screen resistors is not a good idea if the amp is going to be driven into clipping. But by the principle of UL operation excessive screen dissipation should be less of an issue than with standard pentode mode, where screen voltage is always high.

[catalin gramada]

I.m not talking what is happen up to clipping, but over instead. My note comes from direct dc measure in my circuit. I found also for moderate level of overdrive the dc current over 1k limiter goes to 22-25 mA arround. The dc screen voltages drop to 530v.
I will not take the risk to go over 12w limit. I already toasted two tubes with my hard tests...

[Helmholtz]

Maybe this thought will help: As fast as it start clipping the screens start to draw 18 mA through 1k limiters at 550v ; that meant 10w dissipation by the screens.

The simple formula: power equals voltage times current, only works for pure DC or rms AC across a resistor.
Both screen current and screen voltage consist of DC+AC and the AC parts are out-of-phase, so the formula must give a wrong result. I expect actual screen dissipation to be considerably lower than 10W.
UL operation at a B+ of 560V is a design example in the KT88 datasheet and I expect the screen dissipation limit of 6W not to be exceeded up to clipping.
The datasheet also recommends 270R screen resistors to avoid parasitic oscillation - not to limit dissipation.

[catalin gramada]

Maybe this thought will help: As fast as it start clipping the screens start to draw 18 mA through 1k limiters at 550v ; that meant 10w dissipation by the screens.
1k screen grid resistor will keep the screen dissipation at 12w arround for moderate level of overdrive. Anything over 2w will fit.

[boroman]

Thanks a lot. That hand-drawn schematic shows only a "approx" what's inside.. Hard to believe it to be 100% accurate if you know what I mean . I tested my KT88s and they measure quite "as new" (95% emssion) and no problems with shorting - they were bought few years ago by previous owner and amp was rarely played. If I loose some headroom by putting more than 220 ohm, I'd try to push the amp harder so it kind of does not make any sense I suppose? Were the originals 3 or 5 watters? Maybe just raise a bit to 330 ohm and use 4W?

EDIT: Andy Nelson from Vintage Orange sent me better schematics (below). Seems the screens are 270R / 1W. Maybe they just stressed too much because of lower wattage?

[nickb]

A schematic was previously given to you along with the advice to compare it to what you have. If you can see that it differs then you can note the changes can post a definitive schematic .You have the amp, we don't, so we can't do it for you. The switching in the preamp especially is not clear. That would be a truly worthwhile endeavor.

You can use dual 100uF caps if you want. Suggest to add insulation to the can of the upper one at it could be sitting 300V or so.

1000uf filter caps? That seems excessive especially as we have a choke input filter here. That has a 40dB/decade ripple roll-off instead of the 20dB/decade for a cap only filter. Also it must change the dynamics significantly and you said don't want to change the sound.

The screen grid resistors usually die due to bad tube(s) or higher than recommended load impedance. But yes, really pushing it could do it too. The fact the the fuse blew seems like confirmation that this was the reason. There are two schools of thought. One is to use high wattage screen grid resistors so they never die. The other is keep them small so they act as fuses and so protect the tubes. If you opt for the latter best to chose flameproof / fusible types. Yes, you can get some extra protection for the screens by using higher resistances at the expense of a little lower output. If i see blown screen grids I never really trust the power tubes again for fear than the screen grids have been compromised. I test then for gm and if OK I take it the screens are intact.

[catalin gramada]

The amps you mentioned used different supply configurations.The tubes are pentode connected.with screens supplied at 400v arround, not 600v. Remember you OR is a ultralinear...
I suggest a 1k/5w but don't drive it too hard.

[boroman]

1. No 470 ohm or 1k either will help you screens absolute safe, but will give a little bit more clearance. It well depends how hard is pushed.
2. I ended with 1000 uF in my amp, original suited with 100uF
3. A ppimv to limit the signal in front of power tubes should be a useful toy.

I wouldn't add master volume on it... Just want to keep it original.
What wattage of the screen resistors should I get? I see that Orange AD200B uses 1K for screeens, vintage Hiwatt 200 (4xKT88) uses 1K too (3W), but SVT-Classic uses 220R (old SVT even 22R as far as I remember). Should I go for 470 / 3W then?

I also see that one of my 1K resistor on the output transformer is toasted - maybe not complately because it read 1K, but time to replace.

How about biasing (4th question)?

[catalin gramada]

1. No 470 ohm or 1k either will help you screens absolute safe, but will give a little bit more clearance. It well depends how hard is pushed.
2. I ended with 1000 uF in my amp, original suited with 100uF
3. A ppimv to limit the signal in front of power tubes should be a useful toy.

The amp is a bomb. You have to manage to not explode...In my Simms the designer used a whimpy 12ax7 driver to limit the drive capabilities...I changed with a hefty 12bh7 and said bye bye to my tubes during the tests...I have to reconsider how to use it...
SoulFetish have one on the bench. Maybe he comes at different conclusions...