I'll proceed with the 12AX7 removal and get the amp to work for the moment. The addition of an EM84 down the road should not be too complicated. And they don't seem to be too expensive either. Thanks for the tip!
IG
Last edited:
+1 on the EM84.
You want functionality, check this out: Since these can be used as voltage indicators, I see no reason why a circuit couldn't be designed to measure cathode voltage (above the cathode resistor) and scaled so that it can be used to display the power tube bias. IIRC the EM84 has two bars that meet in the middle which suggests the possibility you could even design the display for both power tubes. You might want to use an op amp to isolate the circuit and scale the voltage. I can promise you that you would absolutely be the first on your block!
You want functionality, check this out: Since these can be used as voltage indicators, I see no reason why a circuit couldn't be designed to measure cathode voltage (above the cathode resistor) and scaled so that it can be used to display the power tube bias. IIRC the EM84 has two bars that meet in the middle which suggests the possibility you could even design the display for both power tubes. You might want to use an op amp to isolate the circuit and scale the voltage. I can promise you that you would absolutely be the first on your block!
Last edited:
I did that. But it is a little bit more complicated than seems to be: magic eye tubes are all different so need calibration. That means, instead of calibrating output tubes only the owner has to calibrate magic eyes first. And more often, because they wear down faster than output tubes.
But did you do that on his block?
Of course your point is completely valid but no reason to quit. The calibration issue is the reason to go with the op amp. A trimmer resistor would make calibration no more complicated than bias adjustment and I think I'd bias first and then set the op amp trimmers next. Is it one more thing to fool with? Of course; that's really IS the point. Is it useful? Marginally. Is it cool? I think so.
One might simply label the trim pots "Left Hubris" and "Right Hubris" and fess up to the ruse, but I'd prefer to own the silliness and use "Left Magic" and "Right Magic".
Last edited:
But did you do that on his block?
Neither I did in mine.
Actually, I did, but was not satisfied. Then I added an opamp and a high voltage transistor in parallel with triode in the magic eye tube, to control voltage on deflecting electrode more precise. Then discovered that I need to stabilize in such case both anode and filament voltages. Then made narrow voltage window in which the eye goes from open to closed.
Then replaced it by 3 LEDs of different colors, to indicate "Low", "Normal" and "High". Finally, gave up the idea, since it is always possible to connect a DMM to 10 Ohm resistors in cathodes, changing tubes, and bias each time according to current taste.
Magic eyes were used always to observe dynamics, no more... And even in record level indicators they always had trimpots that needed to be adjusted each time during service.
But I agree with you, it looks cool!
Allright, finally an update...
I'm not one to ask for help/advice and do nothing about it. I did not get a chance to work on my amp much last week and some parts were being delivered to me, so things did not move forward much. But I got my stuff, and implemented the solutions you guys have found for my problem and it pretty much solved it!
Here's what was done. First off, I removed the 12AX7 stage, as it was way overkill on gain. I replaced the humongus 1mfd coupling caps between the triode and pentode sections of the 6GW8 by 0.01mfd units, as per the original design. Lastly, I installed a 47mfd capacitor as the input cap of the smoothing filter, followed by a 180ohm resistor into the original filter.
I checked my wiring a bunch of times, fired up the amp with a series bulb, no oscillation, YES! Tried with no series bulb, fired up and it works, music comes through! B++++ is at about 265V, after 10 minutes on. I'll let it on for a while and see if I need to revise the 180ohm resistor down a notch in the power supply to get closer to 273V. Still not bad for a shot in the dark! I get a slight buzz/hum, but very low level, what I'd think is acceptable for an SE amp. This is probably due to my spaghetti-ish wiring down there.
I'll check all voltages, especially tube bias and go over everything once more before I close it up and start working on the aesthetics. I'll be sure to post some pics!
A big thanks to all who helped! I learned stuff and will be better prepared to attack the next project, perhaps a build from scratch this time!
Best regards,
IG
I'm not one to ask for help/advice and do nothing about it. I did not get a chance to work on my amp much last week and some parts were being delivered to me, so things did not move forward much. But I got my stuff, and implemented the solutions you guys have found for my problem and it pretty much solved it!
Here's what was done. First off, I removed the 12AX7 stage, as it was way overkill on gain. I replaced the humongus 1mfd coupling caps between the triode and pentode sections of the 6GW8 by 0.01mfd units, as per the original design. Lastly, I installed a 47mfd capacitor as the input cap of the smoothing filter, followed by a 180ohm resistor into the original filter.
I checked my wiring a bunch of times, fired up the amp with a series bulb, no oscillation, YES! Tried with no series bulb, fired up and it works, music comes through! B++++ is at about 265V, after 10 minutes on. I'll let it on for a while and see if I need to revise the 180ohm resistor down a notch in the power supply to get closer to 273V. Still not bad for a shot in the dark! I get a slight buzz/hum, but very low level, what I'd think is acceptable for an SE amp. This is probably due to my spaghetti-ish wiring down there.
I'll check all voltages, especially tube bias and go over everything once more before I close it up and start working on the aesthetics. I'll be sure to post some pics!
A big thanks to all who helped! I learned stuff and will be better prepared to attack the next project, perhaps a build from scratch this time!
Best regards,
IG
Allright, finally an update...
I'm not one to ask for help/advice and do nothing about it. I did not get a chance to work on my amp much last week and some parts were being delivered to me, so things did not move forward much. But I got my stuff, and implemented the solutions you guys have found for my problem and it pretty much solved it!
<snip>
Best regards,
IG
That's great news.
FWIW SE amplifiers can and should be pretty quiet. (true of all amplifiers IMHO) I have a bunch of them here, and all of them have less than 1mVrms of output noise, and one of the very quietest ones is also very simple and compact. I think it would be worth your while to snap some really good pictures of the internal layout of your amp - while your feelings might or might not get bruised depending on your layout and build skills - in the process I bet you will probably end up with an even quieter amp..
You originally presented a schematic with a tube rectifier, however if you are not using a tube rectifier but if you have gone to SS rectification and are using 1N4007s or such, go to UF4007 or other ultrafast rectifiers and it might help.
I found it to be a problem in my first amp when I used 1N4007s. UF4007s fixed the problem.
I found it to be a problem in my first amp when I used 1N4007s. UF4007s fixed the problem.
I stayed with the EZ81 for rectification.
As for my noise/hum, I'll try to take some measurement to have an actual figure and not just my subjective impression, but it is very faint and I was about two feet from the test speakers.
The amp was behaving a bit strange at first yesterday, when under initial test. B++++ had slowly drifed down to ~235V, while only one channel got weak and distorted. The bias for this channel's pentode output had risen significantly to ~12-13V instead of 7. I then swaped the tubes to try to isolate the problem and the amp gradually started to behave better. B++++ was now stable at 265V after a good while on, and output bias was 7V and 8V. I guess either the tubes needed some exercise because they might not have worked for a long time, or something was not making very good contact in the sockets and swaping fixed it. I'll run some more tests on that to be sure everything is good and stable.
I might have to change the voltage drop resistor for the 220V triode supply, as I'm getting ~250V. The rest seems fine. I'll be able to try it with some better speakers tonight perhpas and see what it can do.
IG
Last edited:
Here are some pictures of the completed amplifier.
I changed a 10k resistor in the power supply to a 47.5k to lower the 250V B++ to the triode to a proper 220V.
I still have to find a correct value for the fuse. What type should I choose, slow or fast? And I'm not to certain how to size it up either.
I'll work on aesthetics a bit, perhaps a wood plinth, perhaps add an EM84, and that'll be it!
Thanks agian for the help everyone!
IG
I changed a 10k resistor in the power supply to a 47.5k to lower the 250V B++ to the triode to a proper 220V.
I still have to find a correct value for the fuse. What type should I choose, slow or fast? And I'm not to certain how to size it up either.
I'll work on aesthetics a bit, perhaps a wood plinth, perhaps add an EM84, and that'll be it!
Thanks agian for the help everyone!
IG
Whether you use a slow blow or fast blow is a matter of choice. I prefer a fast blow in the event of tube runnig away. That said, fast vs slow is a relative matter and there is not a lot of difference in a catastrophic failure.
You have two 6GW8s, a 12AX7 and a EZ81.
filaments :
6GW8 - 6.3 *.66*2 = 8.316W
12AX7 - 6.3 * .30 = 1.89W
EZ81 - 6.3 * 1 = 6.3W
Total filament power = 16.326W
Pentodes - 55mA each
Triodes – 0.5mA eacg
12AX7 – 0.75mA
Total current per channel 56.25mA
Total current 112.5mA
B+ Power = 225*1.414*.1125= 35.79W
Total power dissipation = 52.11W
52.11W/125V= 0.417A
Theoretically you could go with a 0.5A fuse, but I'd expect it might blow once a year under normal operating conditions because fuses are just not that accurate.
0.75A would be a good bet if you can find one.
A 3AG 0.75A 250V fast blow fuse is probably about right, although I'd go with the half amp fuse and expect it to blow occasionally (false failure).
Littlefuse has a 0.75A one that would be good.
http://www.littelfuse.com/data/en/Data_Sheets/312_318.pdf
Look at the fusing data. Note that at 100% it is supposed to blow in 4 hours minimum. At 135% the fuse is supposed to blow in 1 hr max. 200% 5 seconds max. You can get a lot of dammage in 5 seconds.
Fuses are lousey devices. Unfortunatly they are probably the best option you have unless you want to build a crowbar circuit.
Just noticed they have a 0.6A fuse in that series. 0.6A would probably be the best choice for a trade off between protection and not having occasional blown fuses.
You have two 6GW8s, a 12AX7 and a EZ81.
filaments :
6GW8 - 6.3 *.66*2 = 8.316W
12AX7 - 6.3 * .30 = 1.89W
EZ81 - 6.3 * 1 = 6.3W
Total filament power = 16.326W
Pentodes - 55mA each
Triodes – 0.5mA eacg
12AX7 – 0.75mA
Total current per channel 56.25mA
Total current 112.5mA
B+ Power = 225*1.414*.1125= 35.79W
Total power dissipation = 52.11W
52.11W/125V= 0.417A
Theoretically you could go with a 0.5A fuse, but I'd expect it might blow once a year under normal operating conditions because fuses are just not that accurate.
0.75A would be a good bet if you can find one.
A 3AG 0.75A 250V fast blow fuse is probably about right, although I'd go with the half amp fuse and expect it to blow occasionally (false failure).
Littlefuse has a 0.75A one that would be good.
http://www.littelfuse.com/data/en/Data_Sheets/312_318.pdf
Look at the fusing data. Note that at 100% it is supposed to blow in 4 hours minimum. At 135% the fuse is supposed to blow in 1 hr max. 200% 5 seconds max. You can get a lot of dammage in 5 seconds.
Fuses are lousey devices. Unfortunatly they are probably the best option you have unless you want to build a crowbar circuit.
Just noticed they have a 0.6A fuse in that series. 0.6A would probably be the best choice for a trade off between protection and not having occasional blown fuses.
Last edited:
Thanks for the advice!
It was where to use the 1.4142 multiplication factor I was unsure about.The 12AX7 was taken out, so I might do a bit better than your estimate on a 0.5A fuse. Also, the pentode runs on B++++ at ~270V and the triode on B++, 220V, so I think I should revise the wattage per branch taking this into account. The idle currents differ a bit also.
IG
Last edited:
The voltage to the tubes does not matter. The voltage at the output of the transformer is what is used along with current through each circuit to calculate dissipation (and this ignores transformer losses). Otherwise, if you used the voltage to the tubes you would ignore the dissipation in the rectlifier tube and dropping resistors that are used in the tube supplies.
We are after input power so we need to include everything.
Thinking about transformer losses, I bet there are several watts there as well.
We are after input power so we need to include everything.
Thinking about transformer losses, I bet there are several watts there as well.
I see, makes sense. I can disregard my previous calculation and correct with my different idle currents (P=0.39mA each, T=0.44mA each) and different B+ at the transfo, which actually puts out 250VAC.
Heaters total power: 14.62W
Pentodes: 2*39mA
Triodes 2*0.44mA
Total current = 78.88mA
250V*1.4142*78.88mA=27.89W
27.89W+14.62W = 42.51W
42.51W / 120V = 354mA
How's that sound?
Don't know what close enough value I can find though.
IG
Last edited:
Fuses in most cases provide limited protection to the components in the circuit, they are there in order to reduce the fire hazard in the event of a major component failure such as a shorted winding in the power transformer, a shorted cap, or an output tube that goes into catastrophic melt down due to a manufacturing defect, air leak or component failure.
Given the above it is enough to assure that the fuse blows under the above fault conditions, anything less severe is unlikely to result in the fuse blowing. (Without frequent nuisance blowing)
You need to allow something for inductive inrush in the transformer, plus cold filaments. The capacitive inrush is limited by the warm up characteristics of your EZ81 and the winding resistances. I think a 0.5A - 0.6A slow or medium blow would be just fine, sounds suspiciously familiar..
FWIW I have a different attitude towards nuisance fuse blowing, I want a fuse to blow when a real fault is occurs and under no other circumstance. (Other symptoms will usually tell me there is a problem before the fuse can blow anyway.)
Given the above it is enough to assure that the fuse blows under the above fault conditions, anything less severe is unlikely to result in the fuse blowing. (Without frequent nuisance blowing)
You need to allow something for inductive inrush in the transformer, plus cold filaments. The capacitive inrush is limited by the warm up characteristics of your EZ81 and the winding resistances. I think a 0.5A - 0.6A slow or medium blow would be just fine, sounds suspiciously familiar..
FWIW I have a different attitude towards nuisance fuse blowing, I want a fuse to blow when a real fault is occurs and under no other circumstance. (Other symptoms will usually tell me there is a problem before the fuse can blow anyway.)
Last edited:
IG Yes, that sounds correct on the power calculation.
If you use thermistors on the input power, you can avoid using slow blow fuses. I also hate nusence fuse blows, but view it as a trade off between how good the protection is and how anoying a blown fuse is.
As Kevinkr states, use slow blow to avoid blowing them with inrush current if you don't use inrush limiters. I use inrush limiters on all my designs, so I automatically switched to fast blow fuses and don't think much about it.
If you use thermistors on the input power, you can avoid using slow blow fuses. I also hate nusence fuse blows, but view it as a trade off between how good the protection is and how anoying a blown fuse is.
As Kevinkr states, use slow blow to avoid blowing them with inrush current if you don't use inrush limiters. I use inrush limiters on all my designs, so I automatically switched to fast blow fuses and don't think much about it.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.