sbelyo said:
It should. I have a pair of E-P 5K:32 OPTs gapped at about 35mA I think for headphones. I am using them in a goofy config right now with a single 7KY6 per side. Good stuff, indeed.
One bit of advice I can offer is to allow some distance between these and the power transformer. Induced hum is more of an issue than with speaker outputs.
Oh, also, if Jack didn't start winding, you might want to add an 8 ohm tap as well since you'll be making a powerful enough amp for speakers - I measured my Grados as needing about 5mW (0.005W) to power them to "plenty loud". You could also buy yourself some 32:8 transformers too, I suppose
I thought about the 8 ohm tap, I'm pretty sure I'll never use this amp with speakers. If this design works well I may build a 300B version that will have speaker taps.
I spec'd it to 65ma for 2A3's as well as 45's. That should give me enough power for my K340's and future K1000's.
I suspect this will take me several weeks to gather the rest of the parts.
How do you measure the wattage?
I spec'd it to 65ma for 2A3's as well as 45's. That should give me enough power for my K340's and future K1000's.
I suspect this will take me several weeks to gather the rest of the parts.
How do you measure the wattage?
sbelyo said:
The poor man's method: I downloaded a 60Hz tone and played it louder than comfortable. I then used a multimeter to measure the VAC across the 32 ohm headphones. I repeated this with other frequencies (and other meters, and an o-scope) and the numbers seemed pretty consistent. I don't really know if this works, but that's what I did. I'm guessing I'm within an order of magnitude.
Re: Re: Custom Headphone OPT ordered for the TubelabSE
My plan is to have the power tranny and choke on the left side of the chassis. Then put the board in the middle with the opt's on the right side.dsavitsk said:
I'll have to try that... Is there a cheap pc based analyzer that might do that. Like sub $200?dsavitsk said:
First let me answer some of your other questions:
The 2SK2700 mosfet is extinct, due to ROHS. Toshiba has an unleaded version listed on their website, but no one has them in stock. Many users have used the Fairchild FQP1N50. This works well in most amplifiers, but there have been 2 incidents of these oscillating at high voltages (over 350 volts) with other circumstances (long wires leading to the tube sockets). I have several Toshiba parts to evaluate for use as a better substitute for the 2SK2700. I had planned to evaluate these this weekend, but recent surgery has slowed me down a bit.
If your transformer has a center tap on the 5 volt winding, cut it off or tape it up so that it can not touch anything. It will have the full B+ voltage on it.
The motor run cap can be mounted anywhere you want, but you MUST ground the can. It is not connected to anything inside the cap, but leakage currents can cause the can to be electrically charged if it is not grounded. I have one mounted on the top of the chassis in my "industrial amp".
A few observations about a headphone amp:
I am not a headphone person (my head sweats too much) but I built a headphone amp for a very discerning audiophile using a TubelabSE. The user tried many combinations with several different phones. This amp was a major project that required some modifications to the PC board. Not recommended for a first time builder. The story is here:
http://www.tubelab.com/Headphone.htm
First off 50 to 100 mW can blow the headphones right off of your head. A 45 amp will have more than enough power. A resistive attenuator between the OPT and the headphones may help with the excess power and the hum.
Headphones can be really sensitive, equivalent to a 110 db speaker strapped to your head. Hum will be an issue. The amp above wound up (after the web page was written) with two chokes and a big motor run cap.
The user had me build the amp, and it was working great when it left here, but it was still mounted to a piece of plywood. The user was having a custom case made following the same general layout of the plywood base. The finished amp had some faint hum, so it came back. The hum was fixed by rotating the power transformer 90 degrees and cleaning up the wiring. We added the second choke and the supplemental cap at the user's request.
The 2SK2700 mosfet is extinct, due to ROHS. Toshiba has an unleaded version listed on their website, but no one has them in stock. Many users have used the Fairchild FQP1N50. This works well in most amplifiers, but there have been 2 incidents of these oscillating at high voltages (over 350 volts) with other circumstances (long wires leading to the tube sockets). I have several Toshiba parts to evaluate for use as a better substitute for the 2SK2700. I had planned to evaluate these this weekend, but recent surgery has slowed me down a bit.
If your transformer has a center tap on the 5 volt winding, cut it off or tape it up so that it can not touch anything. It will have the full B+ voltage on it.
The motor run cap can be mounted anywhere you want, but you MUST ground the can. It is not connected to anything inside the cap, but leakage currents can cause the can to be electrically charged if it is not grounded. I have one mounted on the top of the chassis in my "industrial amp".
A few observations about a headphone amp:
I am not a headphone person (my head sweats too much) but I built a headphone amp for a very discerning audiophile using a TubelabSE. The user tried many combinations with several different phones. This amp was a major project that required some modifications to the PC board. Not recommended for a first time builder. The story is here:
http://www.tubelab.com/Headphone.htm
First off 50 to 100 mW can blow the headphones right off of your head. A 45 amp will have more than enough power. A resistive attenuator between the OPT and the headphones may help with the excess power and the hum.
Headphones can be really sensitive, equivalent to a 110 db speaker strapped to your head. Hum will be an issue. The amp above wound up (after the web page was written) with two chokes and a big motor run cap.
The user had me build the amp, and it was working great when it left here, but it was still mounted to a piece of plywood. The user was having a custom case made following the same general layout of the plywood base. The finished amp had some faint hum, so it came back. The hum was fixed by rotating the power transformer 90 degrees and cleaning up the wiring. We added the second choke and the supplemental cap at the user's request.
Re: Re: Re: Custom Headphone OPT ordered for the TubelabSE
Just breadboard it before you drill. I had about 8" between mine (from center to center) and 90 degree turns, and still got hum. It was so low that it could only be heard at night when the city is quiet, but it was enough to drive me to rip the amp apart.
A $5 meter should work fine.
Thanks Mike, I appreciate that.
Part of the reason the power supply is separate, aside from the noise mentioned above, is that I have a number of these cases, and I am building a series of amps with the same tubes at more or less the same operating points, but with various topologies. There are so many anecdotes about which topologies sound best, but very few "controlled" experiments to really explore. So, since headphones allow very simple and low powered amps, I thought I'd find out what I really like. So far it's been fun and informative.
sbelyo said:
Just breadboard it before you drill. I had about 8" between mine (from center to center) and 90 degree turns, and still got hum. It was so low that it could only be heard at night when the city is quiet, but it was enough to drive me to rip the amp apart.
sbelyo said:
A $5 meter should work fine.
MQracing said:
Thanks Mike, I appreciate that.
Part of the reason the power supply is separate, aside from the noise mentioned above, is that I have a number of these cases, and I am building a series of amps with the same tubes at more or less the same operating points, but with various topologies. There are so many anecdotes about which topologies sound best, but very few "controlled" experiments to really explore. So, since headphones allow very simple and low powered amps, I thought I'd find out what I really like. So far it's been fun and informative.
Thanks for answering my questions....tubelab.com said:
I read that page on your site at least once a month. I pretty much said to myself that there was no way I'm going to buy a bunch of 01A's just to find 2 good ones. And those transformers are like gold dust. I've seen them pop up every now and then, but they're way too pricey.
My way of thinking was that the 45 should put out 1 Watt and I have 2 pairs of headphones that are pretty hard to drive. I have a class A soild state amp putting out 1 Watt and that just gets the headphones loud enough. I guess I really wanted more power than I needed.
I'm hoping that I can find a layout that will keep the hum in check. I know that I will need a choke for sure, but I may need some help in tweaking the power supply to reduce any hum.
It may take me a while to finish it, but I'll take pictures and keep track of what was done so that you can use it on your site.
sbelyo said:
You are fighting 3, or maybe 4, main hum issues. The first, and easiest, is ripple in the power supply. For that, as you say, a choke, or two, may be enough. Otherwise, you can increase the PS caps.
The second, which was my problem, is the magnetic coupling between the power transformer/chokes and the output transformer. This just requires distance -- how much will require some experimentation.
The third is from the filaments. AC will probably not work, but might. Again, it will take some trial and error.
Finally, grounding and shielding may be a problem. Tubelab's board should take care of the grounding and hopefully the tubes don't pick up too much noise. These aren't 26's after all.
The real difficulty is that once it is together, if it does hum, you won't necessarily know which of the above is the problem, or if it is only one. It should all be solvable, though, and should yield a really nice amp.
I plan to place everything out on a piece of wood the size of the chassis and find a good layout there. Now, should I start with the power tranny and coke 180 degrees from the opt's? And does the choke have to be in the same orientation as the power tranny?
The chassis (Aluminum) should be grounded to earth right?
We shall see. I'm definitely building it now...
sbelyo said:
There are 3 different orientations all at 90 degrees to each other (think of a cube.) The power transformer, choke, and OPTs should each get one of these.
sbelyo said:
Short answer, connect the chassis to safety ground and connect signal ground to safety ground in one place. Do a search, though, for lots of grounding threads.
One other thing w/r/t/ wattage. The problem is not the output tube here, the problem is really the input tube. The fact that the output tube is capable of 2 watts, or whatever, does not mean the amp is always putting out 2W. It depends how high the output signal is. Here, the driver is pretty high in gain which will translate to a high signal on the output. I don't know of anything pin compatible with the 417A with lower mu, so you are stuck with that.
One solution is to shunt the pot to cut gain a little. Another solution is a stepdown transformer on the input, something like a Jensen JT-6110K-B. However, what may be the best solution would be to up the primary impedance on the OPTs to 10K, or even higher. This essentially trades some power (that you don't need) for lower distortion and better damping, which is good.
One solution is to shunt the pot to cut gain a little. Another solution is a stepdown transformer on the input, something like a Jensen JT-6110K-B. However, what may be the best solution would be to up the primary impedance on the OPTs to 10K, or even higher. This essentially trades some power (that you don't need) for lower distortion and better damping, which is good.
No, the 5842 tube used in the TubelabSE has significant Miller capacitance. A high value volume pot can cause a loss of high frequency response.
Hum is one thing that the TubelabSE design doesn't have much of. The filaments are all DC. The output tubes filaments are also regulated DC. I have measured the hum on an unshielded PC board operating on the bench with the power transformer, choke and 100uF motor run cap about a foot to the left of the board and the OPT's about 18 inches to the right of the board to be 80db below 1 watt on a 45 amp (input shorted at the PC board).
The picture of the Lexan amp on my home page is an example of how NOT to do it. The 3 transformers are lined up in a row along the back of the amp. There was no choke or supplemental cap. That amp had a very faint hum with my ear next to the speaker on my speakers which are 87db at 1KHz and about 70 - 75db at 60 Hz. The first time that amp visited some real speakers wasn't pretty. I have since rotated the power transformer 90 degrees, killing the hum on some 96db speakers.
When that amp had an appointment with some 106db speakers containing active subs, I added a choke and a 20 uF polypropelyne cap. There was no audible hum with my ear to the speakers.
http://www.tubelab.com/Sound Checks.htm
Re: Re: Re: Re: Custom Headphone OPT ordered for the TubelabSE
I've done similar comparos in the past, mostly around the 12B4A as a SE and PP linestage. All sorts of configs were tried. Not gonna comment on which I preferred as it's buried in the old threads, and my notebooks aren't to hand, but it's something I've thought about repeating.
My next two tube based projects are a similar tapped autoformer, CCS loaded tube design for dynamic cans (will get Jack to quote on the iron), and a PP direct coupled one for my STAXes and Magnavox electrostatic cans. This will probably use resistively loaded trioded EL84's as outputs.
I'd be very interested in reading about that comparo, especially if there are some measurements with it to give a reference.dsavitsk said:
I've done similar comparos in the past, mostly around the 12B4A as a SE and PP linestage. All sorts of configs were tried. Not gonna comment on which I preferred as it's buried in the old threads, and my notebooks aren't to hand, but it's something I've thought about repeating.
My next two tube based projects are a similar tapped autoformer, CCS loaded tube design for dynamic cans (will get Jack to quote on the iron), and a PP direct coupled one for my STAXes and Magnavox electrostatic cans. This will probably use resistively loaded trioded EL84's as outputs.
Re: Re: Re: Re: Custom Headphone OPT ordered for the TubelabSE
Here's the parafeed version: http://www.ecp.cc/less-pressivo-plus.html
dsavitsk said:
Here's the parafeed version: http://www.ecp.cc/less-pressivo-plus.html
Hello
I'm revamping this thread because I'm planning to make a tubelabSE with 45 form my headphone too.
I'm revamping this thread because I'm planning to make a tubelabSE with 45 form my headphone too.
The story is here:
Headphone AmpQUOTE]
Here I read:
"A low impedance set of headphones (8 to 32 ohms) can usually be driven by a conventional output transformer made for driving a speaker.
...
The usual single ended transformers work as expected. They give good results with the 40 ohm phones (with a 24 ohm resistor in parallel)"
I have a 40 OHM phone then I think it's correct put one 24 Ohms resistor (R1) in this way:
An externally hosted image should be here but it was not working when we last tested it.
and soldering it to the active cable and to the ground cable of each 8Ohms loudspeakers output. I'm correct?
Resistors are different depending by the Ohms attenutation (in this case 24 ), and by the Watt: in this case I must use a lower Watt as possible? A 24 Ohms-1 W, for example, used in a 2,5 watt RMS output amplifier, attenuates the amplifier power to 1,5 Watt RMS?
With my 40 Ohms can, a 24 Ohms resistor is correct if I use it even with any other amplifier, a 20 or more Watt RMS too ?
...newby questions
Rob
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.