I'm working on an HH Scott 299D with the 7591s. It clips very severely at about 12 o clock on the volume (EDIT: with a 0.65 volt sine wave to the line stage input). The top 10% of the wave forms actually inverts.
Just as the clipping begins, the voltages drop globally. After much head scratching and tracing, I see that the screen grid voltages plummets from 410 to below 300 at the onset of clipping and it would go even lower if you were to keep the volume up. The plate voltages fall too but not so drastically.
I believe the voltages fall as a result of the extra current going to the screen grids. As a matter of fact, the two power resistors in the CRC filters before the screen taps were burned up when I got it and the resistors further down the chain are fine.
So, I now suspect that the clipping is a result of the dropping plate voltages on the 12AX7s in the first stages. In other words, I think the clipping may be caused by the falling plate voltages in the AF amp and tone amp sections; a symptom rather than the cause of the problem.
I also note that the bias actually goes more negative when the clipping/voltage drop begins. From -22 to perhaps -32. The voltage is relatively unaffected close to the bridge rectifier but closer to the grid (after the bias controls) the swing is greatest. I don't understand that.
BTW, I have swapped power tubes for a new quad and the problem stays with the amp. (I have 10 ohm resistors on the cathodes and they are biased at about 70%).
So, whats happening here? What would cause the screen current to shoot up so drastically? I'm stumped.
Thanks for the help.
Just as the clipping begins, the voltages drop globally. After much head scratching and tracing, I see that the screen grid voltages plummets from 410 to below 300 at the onset of clipping and it would go even lower if you were to keep the volume up. The plate voltages fall too but not so drastically.
I believe the voltages fall as a result of the extra current going to the screen grids. As a matter of fact, the two power resistors in the CRC filters before the screen taps were burned up when I got it and the resistors further down the chain are fine.
So, I now suspect that the clipping is a result of the dropping plate voltages on the 12AX7s in the first stages. In other words, I think the clipping may be caused by the falling plate voltages in the AF amp and tone amp sections; a symptom rather than the cause of the problem.
I also note that the bias actually goes more negative when the clipping/voltage drop begins. From -22 to perhaps -32. The voltage is relatively unaffected close to the bridge rectifier but closer to the grid (after the bias controls) the swing is greatest. I don't understand that.
BTW, I have swapped power tubes for a new quad and the problem stays with the amp. (I have 10 ohm resistors on the cathodes and they are biased at about 70%).
So, whats happening here? What would cause the screen current to shoot up so drastically? I'm stumped.
Thanks for the help.
Last edited:
I should add a couple of things. The cathode current, as measured across the 10R resistor shoots up to about 250% rated power when this happens. Naturally, I don't allow it to go on for more that an second or so (which makes troubleshooting a bit challenging).
Design plate voltages and screen voltages are 420/410 on one drawing and 456/425 on another.
I considered the possibility that the transformer might be the culprit so pulled the rectifier and hooked up the 500ma lambda power supply. The same thing happened. B+ current went from 100ma at idle to 200ma during the clipping episode and cathode current doubled. The transformer is not the culprit.
Here is a Schematic
Design plate voltages and screen voltages are 420/410 on one drawing and 456/425 on another.
I considered the possibility that the transformer might be the culprit so pulled the rectifier and hooked up the 500ma lambda power supply. The same thing happened. B+ current went from 100ma at idle to 200ma during the clipping episode and cathode current doubled. The transformer is not the culprit.
Here is a Schematic
Last edited:
Screen current increases when the anode voltage drops below the screen. This is normal pentode/tetrode behaviour. You appear to be driving the ouput valves into hard clipping - perhaps aided by a negative feedback loop?
I don't know why the grid bias supply is changing. There must be some interaction in the power supply.
I don't know why the grid bias supply is changing. There must be some interaction in the power supply.
You might also have a bad (or going bad) filter capacitor in the power supply. As old as that Scott is, it most definitely needs to be recapped. I own the LK-72 and it rocks the house out! It starts to clip at about 3/4 volume, but by then, its too loud!
What was the power output level at the onset of clipping? These amps aren't as powerful as the manufacturer's ratings would seem to imply because of the way they specified watts. The D is good for perhaps around 34W rms per channel IIRC.
These amplifiers also have relatively high gains from the aux and tuner inputs, most modern sources will easily drive them to clipping. All it really takes is a few hundred mVrms of input. You were hitting yours very hard IMHO..
Not sure whether the behavior you were seeing was due to anything other than massive overdriving..
The supply caps most certainly need to be replaced at this point.
These amplifiers also have relatively high gains from the aux and tuner inputs, most modern sources will easily drive them to clipping. All it really takes is a few hundred mVrms of input. You were hitting yours very hard IMHO..
Not sure whether the behavior you were seeing was due to anything other than massive overdriving..
The supply caps most certainly need to be replaced at this point.
Thanks for you comments. I replaced the caps this am and it didn't really make any difference in the specific problem I cited but of course you are correct about replacing them. I believe you are both correct about overdriving the circuit. I scoped the circuit again today and I now have evidence.
The circuit has
1) scratch and rumble POC followed but a triode. The grid voltage is 1 v peak to peak from my function generator with a 1.25 volt cathode. The plate is about 47 pp.
2) Then tone circuit and another triode. Now it's about 2.5 with a 1.25 cathode. Very close here. Then 40 pp from the plate.
3) now the volume pot and a voltage divider of 150K/1meg followed by the grid of the 6BL8. At full volume thats going to be 6 volts which is way over the top for the 1.3 volt cathode.
So one must wonder why so much signal? The Scratch and Rumble circuit is a POC and there should be a 270K/2.2 meg divider there. As a matter of fact the drawing shows the first stage to be -13db so my 1 volt should be something like .22 volts going into the first grid. That looks to be the problem right there.
OK, pause for some more scope work. I just set my function generator up for .22 volts at the first grid and guess what? The amp acts properly. I clip at about 90% travel on the volume pot and the output is about 15 pp at the 8 ohm tap. It looks like the problem is in the POC. I just might replace the POC with a voltage divider and be done with it. What is the value of scratch and rumble features anyway?
Thanks for the ideas. I was high centered on bad ideas last night.
The circuit has
1) scratch and rumble POC followed but a triode. The grid voltage is 1 v peak to peak from my function generator with a 1.25 volt cathode. The plate is about 47 pp.
2) Then tone circuit and another triode. Now it's about 2.5 with a 1.25 cathode. Very close here. Then 40 pp from the plate.
3) now the volume pot and a voltage divider of 150K/1meg followed by the grid of the 6BL8. At full volume thats going to be 6 volts which is way over the top for the 1.3 volt cathode.
So one must wonder why so much signal? The Scratch and Rumble circuit is a POC and there should be a 270K/2.2 meg divider there. As a matter of fact the drawing shows the first stage to be -13db so my 1 volt should be something like .22 volts going into the first grid. That looks to be the problem right there.
OK, pause for some more scope work. I just set my function generator up for .22 volts at the first grid and guess what? The amp acts properly. I clip at about 90% travel on the volume pot and the output is about 15 pp at the 8 ohm tap. It looks like the problem is in the POC. I just might replace the POC with a voltage divider and be done with it. What is the value of scratch and rumble features anyway?
Thanks for the ideas. I was high centered on bad ideas last night.
Last edited:
I wrote POC - I mean PEC.
I discovered that someone had soldered a wire between PEC pins that shorted that entire -13 db stage. That little fix also excluded the negative feedback loop. I couldn't see it until I clipped it out. So I was overdriving it by a factor of 5 or therebouts.
I've removed and discarded the PECs and am wiring in a new circuit. The PEC includes an RC filter with a -3 frequency of 58hz (1200pf and 2.2meg). I think I'll leave that out unless someone can explain why . There is another just after the PEC on the first grid that is .0015uf and 2.2 meg that should do the same; I can't imagine a reason that one would want two in series. A larger cap would probably be a good idea too.
I discovered that someone had soldered a wire between PEC pins that shorted that entire -13 db stage. That little fix also excluded the negative feedback loop. I couldn't see it until I clipped it out. So I was overdriving it by a factor of 5 or therebouts.
I've removed and discarded the PECs and am wiring in a new circuit. The PEC includes an RC filter with a -3 frequency of 58hz (1200pf and 2.2meg). I think I'll leave that out unless someone can explain why . There is another just after the PEC on the first grid that is .0015uf and 2.2 meg that should do the same; I can't imagine a reason that one would want two in series. A larger cap would probably be a good idea too.
- Status
- Not open for further replies.