Bottlehead Forum
Bottlehead Kits => Crack => Topic started by: Loon on May 05, 2014, 04:53:55 PM
-
Hi there
I would like to see if it would be possible to use a 6528 or 6528A tube. It has a pin out the same as 6080 with a much, much higher transconductance of about 37,000. http://frank.pocnet.net/sheets/077/6/6528A.pdf
The only potential problem I can see is that it draws about 5A of current, much greater than the 2.5A of the 6080.
To overcome this I am thinking of using a separate transformer to feed the heaters of the power tube. I have found a Hammond transformer with two secondaries that output 9V @ 6.67A. I would hook this up to a couple of regulated DC filament supplies, one feeding the input tube and the other the power tube. These would output the correct voltage to each tube as a DC supply.
Do you think my plan would work? Expensive, I know, but an exercise in learning.
Thanks.
Naïfurne
-
If you rectify a transformer output, the rms current in the winding is roughly double the current into a resistor. So you need a 10-amp current rating.
Few sockets can deliver 5 amps to a heater for long. Google around for the sockets and precautions common with the Russian 6C33, which also has high heater current and transconductance. It's a serious undertaking - good luck!
-
Thanks Paul.
You guys are great! I told you it would be an exercise in learning and I have learnt something within a couple of hours of coming up with the idea.
Thank you.
-
OK. Wow. The 6C33 is one serious tube.
A quick read has suggested the following to me:
1. A tube with heat output that can melt solder needs some serious cooling
2. Active cooling might be necessary, perhaps in the form of a very quiet computer fan (thinking low noise Noctua, which can be silent for all intents and purposes, but will they be electrically silent)
3. I will need to leave some space between the tube and other components. Was going to custom build a case anyway. May just have to make it a bit bigger.
4. The tube socket will need to be very robust. I have ordered a teflon octal, but may need to go ceramic with very firm pin contact.
Does this sound right yo you? Any feedback? Need to find a suitable transformer in the meantime.
-
Fans aren't really necessary for something this big, just give it some breathing room.
I don't know about octal sockets, but when I went searching for 6C33 sockets, I used Russian military surplus sockets designed for that tube in particular. Teflon would be one of my last choices.
Rectifying the heater supply will add enormous cost, weight, heat, etc. to your project, and it isn't necessary.
-
Thanks Paul. For some reason I was under the impression that rectification would lower the noise floor. Also had hum issues with my last Crack build when using a 12AU7- 6SN7 converter in the input position and thought rectification of heater supply would solve this problem. Thought I may as we'll rectify both to see what happens.
You think a DC current to the heater of the CF won't make any impact on sound?
Thanks!
-
I have done a bit more reading to try to work out any benefit of rectification in the output tube.
Please check what I have summarised below to ensure it is right.
1. AC current feeding the heater of a tube can cause hum because the heater radiates an electromagnetic field
2. By implication, DC current is less likely to cause hum (will not cause hum?)
3. The power tube (eg 6528 in my case, 6080 normally) is a cathode follower. Cathode followers always have a gain (slightly) less than 1. Therefore, any hum arising at this point in the circuit will not be amplified.
4. In the event that hum does result from the use of AC power feeding this tube, it can be reduced by use of a transformer centre tap, an artificial centre tap or DC elevation.
Source:
http://www.valvewizard.co.uk/heater.html
http://www.valvewizard.co.uk/accf.html
In view of this, Paul's objection makes perfect sense as the transformer specs to supply the heater of the power tube can be greatly reduced by not converting AC to DC (the process of rectification)
This is what I think I should do:
1. Rectify the input tube (12AU7 et al) using this from Peter Millett http://www.pmillett.com/LV_reg.html
2. Feed this rectifier circuit with a Triad VPS16-5000 transformer. When the secondaries are wired in parallel it will output 8.0V @ 10A (I calculate that the rectifier requires at least 6.326V RMS @ 0.54A so, yes, I have over specified but I think it will work and be cooler as well)
3. Run the 6528 heater directly from the secondary windings (in series) of a Hammond 266PA6, which would supply 6.3 V @ 6A (the 6528 heater requires 6.3V @ 5A)
However......
I have already ordered two of Peter Millett's rectifier PCB and two nice heat sinks that will dissipate 10-15W heat.
So this is what I think I will actually do.....
1. Feed the 12AU7 heater 6.3VDC as noted above
2. Provide the 6528 heater with 6.3VDC via Peter Millett's rectifier circuit
3. Feed the rectifier from a second Triad VPS16-5000 outputting 8.0V @ 10A. To produce the required output voltage of 6.3 V @ 5A, the rectifier requires an input of at least 7.953V @ 9A (according to my calculations). The Triad VPS16-5000 seems to come just within spec.
To have some voltage headroom, I might reduce the heater supply to 6.0 V. This should also have the bonus of slightly extending tube life. Will the resultant reduced emissions degrade sonic performance at all?
The voltage loss across the rectifier is only approx 1.65V, which my heat sink should be able to deal with (up to 15W).
The Triad transformer I have chosen is about $36 AUD per unit.
Do you think my plan will work.
Please give me feedback about the conclusions I have reached. Am I on the right track or barking up the wrong tree?
-
I don't know about octal sockets, but when I went searching for 6C33 sockets, I used Russian military surplus sockets designed for that tube in particular. Teflon would be one of my last choices.
Paul. Teflon starts to deteriorate talk about 260 degrees Celcius. What temperatures would you expect the pins to reach when a 6528 is in operation? Is heat the only reason you wouldn't use teflon?
-
Thanks Paul. For some reason I was under the impression that rectification would lower the noise floor. Also had hum issues with my last Crack build when using a 12AU7- 6SN7 converter in the input position and thought rectification of heater supply would solve this problem. Thought I may as we'll rectify both to see what happens.
Providing a 6V supply for that tube may help if you plan to use adapters. There's a big difference between making 6V for a tube that wants 600mA, and making 6V for a tube that wants 5A.
-
1. Rectify the input tube (12AU7 et al) using this from Peter Millett http://www.pmillett.com/LV_reg.html
2. Feed this rectifier circuit with a Triad VPS16-5000 transformer. When the secondaries are wired in parallel it will output 8.0V @ 10A (I calculate that the rectifier requires at least 6.326V RMS @ 0.54A so, yes, I have over specified but I think it will work and be cooler as well)
I would use a 9V at 1.5A transformer. An 8V/10A transformer is well beyond overkill, and the transformer regulation will be poor, and the chassis space that such a transformer uses will be wasted.
3. Run the 6528 heater directly from the secondary windings (in series) of a Hammond 266PA6, which would supply 6.3 V @ 6A (the 6528 heater requires 6.3V @ 5A)
That should work.
I have already ordered two of Peter Millett's rectifier PCB and two nice heat sinks that will dissipate 10-15W heat.
That's not quite how this works, other aspects have to be taken into consideration.
2. Provide the 6528 heater with 6.3VDC via Peter Millett's rectifier circuit
I would expect failure of the regulator in this situation. I would check with Pete first. Part of the issue is that you want ~8V at your lowest line voltage, so if your line voltage creeps up, you may be cooking off 15 Watts on that heatsink rather quickly.
-
Thanks Paul
I will find a transformer for the 12AU7 heater matching your specs. Have just contacted Pete to see if the Landfall heat sink will do the job.
-
Your summary of heater noise relative to cathode followers also mirrors my own experience. I would focus more on the real circuit adjustments that will need to be made to run that tube rather than going after a noise source that very likely will not exist.
-
Thanks Paul. Good advice. Sounds like my build will be much simpler if I supplied the Cathode follower heater with AC current and left it at that.
-
I took a class in design of experiments a couple decades ago, and learned the best, most useful advice: Budget 25% of your money (/time/effort/...) for the first experiment. The only thing you learn from the first experiment is how you should have done it. You need enough left over to do it right, and doing it right is always more work than you thought.
-
Wise words. If at first you don't succeed.... Should be "when at first you don't succeed". Particularly in my case. I am a life member of losers anonymous.
-
Hah! Those who persevere will inherit the earth!
-
Hmm...
Having trouble trying to find a suitable transformer with secondary of 9V @ 1.5 A for the rectifier feeding the 12AU7.
Have three reasonably close contenders:
1. Triad FS16-800-C2 [email protected] - 25% voltage regulation
2. Triad F-3152XP 7.5V @ 1.0A - voltage regulation unknown
3. Triad VPT18-1390 [email protected] - 13% voltage regulation
Is voltage regulation a factor I should be concerned about? Any other variables apart from size I should be looking at?
I am tempted to go with option 3, which is a well-stocked toroidal ( although is $26.57, about double the price of other two options). Is it worth exploring the possibility of learning to wind my own transformers?
Thanks again.
-
You'd have to talk to Pete about his voltage regulator, and how much voltage you need to shove into the regulator to avoid dropout.
-
Another issue is the operating point of the cathode follower.
I would like to experience the sonic effect of varying output impedences determined by the mutual conductance of the cathode follower. Therefore, I would like my Crack to be able to accept the following tube types:
6AS7G
5998
7236
7802WB
6528(A)
It's one thing to be able to plug these tubes in and have sound come out. It is quite another to plug these tubes in and have a great sound come out.
The Vp / Ip curves for these tubes are all quite different. The effect of this is that the operating point will move depending upon the cathode load and the grid voltage.
I would like to determine the optimal, most linear operating point for each of these tubes. I have a couple of questions in this regard:
1. Can I use the Vp / Ip curves to determine the best cathode load and grid voltage? Would the Cathode curve be the same a the anode curve, except with reversed polarity?
2. How do I determine resistor values for the speed ball to achieve a good cathode operating point?
Implementation
I assume that changing cathode load would be easier than changing grid voltage, as the latter is determined by anode voltage of input tube and playing with this could mess up the 12AU7's operating point.
When I went under the user name "w0lfd0g" I floated the idea of putting plate load switches on the input tube on the underside of the chassis. I plan on having variable cathode resistance to set the operating point of the cathode follower through a similar implementation, but I will probably need a switch with more throws.
Could someone more knowledgeable than me (most people!) please chip in to answer my questions and provide feedback on my plan?
-
You'd have to talk to Pete about his voltage regulator, and how much voltage you need to shove into the regulator to avoid dropout.
I used the formulae Pete provided with his schematic to determine minimum voltage and current input. This is determined by input AC frequency, peak voltage, value of capacitance and the requirements of the diodes used for rectification. All three transformers listed exceed the minimum requirements.
I was mostly concerned about the rated voltage regulation of the transformers and grossly exceeding the minimum input current requirements.
-
This project really calls for more information than can fit in a forum post.
I'm sure you can learn a lot about power supply design on the web, including the meaning and effect of transformer regulation.
Less easy is to determine optimum operating points. To over-simplify, those tubes that have a higher transconductance will drive lower impedance phones. But transconductance is itself a function of the operating point. More cathode current will increase the transconductance but creates more heat and may require a new design to replace the Speedball current source. Incidentally "transconductance" is another way of saying "tendency to oscillate" - that's why I referred you to 6C33 design issues.
...
I once took a drawing class, thinking it would be cool to be able to sketch people and places I've seen. I learned two things - it takes a lot of practice and many mistakes, and a drawing takes more than 10 times as long as I had thought. But I still work on it from time to time because it's still rewarding. Same goes for electronic design - it's more work than you think, but it's also more rewarding than you think.
-
Thanks for the encouragement Paul. This is a hobby that I would like to continue for the rest of my life and that I am only just starting. It is really helping with my depression issues atm as a kind of diversion therapy.
You have gone above and beyond the call of duty in answering my questions.
Warm regards and thanks
-
I'd recommend buying a piece of plywood, some octal relay sockets, a beefy variable power supply, and Morgan Jones's book on designing tube amps. There are high voltage supplies around that will give you 100-300V and a few hundred mA if needed. Some even have a beefy 6.3V supply just to keep you out of trouble.
Of course, you'll need to sketch circuits and buy the appropriate extra bits to get them running (a few caps and some very large resistors)
You can use those tools to try a whole ton of different things without committing to buying a whole bunch of parts. If you try 20-30 different designs, odds are that there will be one or two configurations that make enough of an impression on you to move forward.
-
Awesome idea. I imagine there would be cheap thrills and spills to be had with this set up! The perfect balance of theory and practice.
-
That's how I started. Only the original gangstas will remember my frightening breadboards in the basement listening room back on the 90s. As long as you are willing to take full responsibility for the potential hazards of a bunch of clip leads carrying high voltage strung all about it can be kind of a fun way to quickly sketch out a circuit.