600 ohm primary has no business connected to a headphone. Those are preamp output transformers.
A great suggestion was sent to me by Lauscher to use 6A3's or equivalent in filament bias with the coleman regs for the output tubes.
I've been considering this and here are the pro's and con's:
Pros:
Gets rid of the big bypass cap.
Less heatsinking required for coleman regs
NOS tubes are much cheaper and easier to come by
Cons:
The Ra raises over 1k ohms which saps power and power delivery.
Less flexibility with output transformer due to the higher Ra.
Higher Filament voltage may mean more noise, of course with a DHT headphone amp noise is the greatest challenge.
I think I am sticking with 2A3 for now, almost think PSE 4p1L's would be a better choice if you must have filament bias than the 6A3. But for ultimate power PSE 6A3's in filament bias may be the ticket over a 300B.
Thoughts?
I've been considering this and here are the pro's and con's:
Pros:
Gets rid of the big bypass cap.
Less heatsinking required for coleman regs
NOS tubes are much cheaper and easier to come by
Cons:
The Ra raises over 1k ohms which saps power and power delivery.
Less flexibility with output transformer due to the higher Ra.
Higher Filament voltage may mean more noise, of course with a DHT headphone amp noise is the greatest challenge.
I think I am sticking with 2A3 for now, almost think PSE 4p1L's would be a better choice if you must have filament bias than the 6A3. But for ultimate power PSE 6A3's in filament bias may be the ticket over a 300B.
Thoughts?
Hi,
to get a good sound from a DHT tube with DC filament, you have to decouple the music signal from the DC supply.
- conventional CLC supply and then a 0,2H to 0,5H choke before the filament. A CRC supply will generate a lot of heat with 1A - 2,5A.
- the reg. from Rod Coleman are no usual voltage or curent regulators. They sound so good because they separate the signal from the DC supply.
On my 300B amp there is no noise on my 101db speakers. Headphones are much more sensitive. I suggest send a pm to R. Coleman. He is verry helpfull and has a lot of theoretic and practical experience. His kits are not expensive. You can not use them with a simple 10,000µF cap in front. With a CLC you get them silent.
The input voltage must be 6V-7V higher than ouput voltage. It dissapates more heat if you use 2,5A instead of 1A.
The Ra does generaly rise when you use filament bias, because you remove the cathode capacitor.
Ra total is=Ra+(mu+1)xRcathode
Your OPT has 5KOhm, so 1K Ra is no problem. Usual formula for OPT is 3 to 4 times the Ra. For the Hifiman HE-6 you need a lot of damping. PSE with 4P1L is one solution. Plus you have a lot less heat dissipation in filament bias than with 6A3/6B4G/6C4C.
If you want low Ra, then use 6C33.
Johann
to get a good sound from a DHT tube with DC filament, you have to decouple the music signal from the DC supply.
- conventional CLC supply and then a 0,2H to 0,5H choke before the filament. A CRC supply will generate a lot of heat with 1A - 2,5A.
- the reg. from Rod Coleman are no usual voltage or curent regulators. They sound so good because they separate the signal from the DC supply.
On my 300B amp there is no noise on my 101db speakers. Headphones are much more sensitive. I suggest send a pm to R. Coleman. He is verry helpfull and has a lot of theoretic and practical experience. His kits are not expensive. You can not use them with a simple 10,000µF cap in front. With a CLC you get them silent.
The input voltage must be 6V-7V higher than ouput voltage. It dissapates more heat if you use 2,5A instead of 1A.
The Ra does generaly rise when you use filament bias, because you remove the cathode capacitor.
Ra total is=Ra+(mu+1)xRcathode
Your OPT has 5KOhm, so 1K Ra is no problem. Usual formula for OPT is 3 to 4 times the Ra. For the Hifiman HE-6 you need a lot of damping. PSE with 4P1L is one solution. Plus you have a lot less heat dissipation in filament bias than with 6A3/6B4G/6C4C.
If you want low Ra, then use 6C33.
Johann
Member
Joined 2009
Paid Member
I had similar thoughts several years back - has anybody built it ?
http://www.diyaudio.com/forums/tube...g-triode-fet-complementary-feedback-pair.html
Hello Guys,
I have work on a DHT amp with old american tubes since a week. And I just discover this thread ! Hope you can give me some advises
I was first focus on my need and the output stage.
So for the need : my headphone are HE-500 and a custom based on trp50 mkII. And if I can drive, in the futur, an LCD2c I will please.
At the end of my research I have found this option : LL1689/25 mA with type 46 old tube in triode strapped. 46 biased around 240 V and 23 mA.
The primary inductance will be around 50 L so the cutoff frequency will be at ~6 Hz.
With the HE-500, 38 Ohms, the load will be 12300 Ohms at the output. The load line seems ok and in the linear region. But the maximum output power will be 400 mW .... Did you think it could be ok ?
If not, I can see two other possibility :
- 4P1L and LL2765
- 2A3 familly with LL2765/50mA. With this one, at least, I'm sure about output power ... maybe too much power by the way .... Don't know the primary inductance of the LL2765 in 50 mA configuraiton.
Please can you help me or give me some advice ?
Thank you in advance !
ps : for the story, I have start to think about a SPUD with LL1689/D3A ... but well building an amp take time ... so I will taje more pleasure to build a DHT with triode
I have work on a DHT amp with old american tubes since a week. And I just discover this thread ! Hope you can give me some advises
I was first focus on my need and the output stage.
So for the need : my headphone are HE-500 and a custom based on trp50 mkII. And if I can drive, in the futur, an LCD2c I will please.
At the end of my research I have found this option : LL1689/25 mA with type 46 old tube in triode strapped. 46 biased around 240 V and 23 mA.
The primary inductance will be around 50 L so the cutoff frequency will be at ~6 Hz.
With the HE-500, 38 Ohms, the load will be 12300 Ohms at the output. The load line seems ok and in the linear region. But the maximum output power will be 400 mW .... Did you think it could be ok ?
If not, I can see two other possibility :
- 4P1L and LL2765
- 2A3 familly with LL2765/50mA. With this one, at least, I'm sure about output power ... maybe too much power by the way .... Don't know the primary inductance of the LL2765 in 50 mA configuraiton.
Please can you help me or give me some advice ?
Thank you in advance !
ps : for the story, I have start to think about a SPUD with LL1689/D3A ... but well building an amp take time ... so I will taje more pleasure to build a DHT with triode
Hello,
Sorry, forget to mention this information, and I don't see the edit button
- LL1689/25 : ratio 18 or 9. The 9 ratio could give a load of 3079 with the a 38 Ohms headphones. But with a little lest impedance headphone it could be just.
- LL2765/50 : ratio 12 or 6.
Any opinion will be appreciate. Well any experience with an HE-500 and something under 500 mW ?
thx !
Sorry, forget to mention this information, and I don't see the edit button
- LL1689/25 : ratio 18 or 9. The 9 ratio could give a load of 3079 with the a 38 Ohms headphones. But with a little lest impedance headphone it could be just.
- LL2765/50 : ratio 12 or 6.
Any opinion will be appreciate. Well any experience with an HE-500 and something under 500 mW ?
thx !
Hum ... Bad News .... Really hard to find a matched pair of nos 46 tubes ....
I have sent mail to multiple retailers since a week now, europe and nord america, and most of them answer me negatively to my request ....
And I don't found a clear answer about the amp power needed for HE-500 : 300 mW of real undistorted power is enough or not ?
I'm searching in the thread ... but it seems the search engine doesn't want to work with "HE-500" key word ...
I have sent mail to multiple retailers since a week now, europe and nord america, and most of them answer me negatively to my request ....
And I don't found a clear answer about the amp power needed for HE-500 : 300 mW of real undistorted power is enough or not ?
I'm searching in the thread ... but it seems the search engine doesn't want to work with "HE-500" key word ...
Somewhere in this long thread, I'm pretty sure regal made some power calculations.
300 mW seems low to me.
I have a ming da mc66, modified for different output tubes.
It is supposed to put out 1.5 watts
mc66
FWIW, it has plenty of power to drive my he500's.
In this old review, they recommend 1 watt of power
he500 review
Here is a power calculator
hp power calc
Based on the power calc, 300mW seems marginal. But if 300mW is your max power, usable power is less.
Randy
300 mW seems low to me.
I have a ming da mc66, modified for different output tubes.
It is supposed to put out 1.5 watts
mc66
FWIW, it has plenty of power to drive my he500's.
In this old review, they recommend 1 watt of power
he500 review
Here is a power calculator
hp power calc
Based on the power calc, 300mW seems marginal. But if 300mW is your max power, usable power is less.
Randy
Hello,
Thank you for reply. Really good timing, because I have made some mesure that's says I need less than 5 Vpp after the opt.
I have done two test :
- First one : Archimago's Musings: MUSINGS: How much amplifier power do you really "need"? Why not test for yourself?
With the HE-500 I have read 0,7 Vrms. But with the Zmf Omni I have read 1 Vrms.
- Second one, playing Metallica S&M at a loud level, and do some mesure with my oscillo. With the Zmf Omni I have read, at the max, 2,9 Vpp.
Regards,
Thank you for reply. Really good timing, because I have made some mesure that's says I need less than 5 Vpp after the opt.
I have done two test :
- First one : Archimago's Musings: MUSINGS: How much amplifier power do you really "need"? Why not test for yourself?
With the HE-500 I have read 0,7 Vrms. But with the Zmf Omni I have read 1 Vrms.
- Second one, playing Metallica S&M at a loud level, and do some mesure with my oscillo. With the Zmf Omni I have read, at the max, 2,9 Vpp.
Regards,
In case someone reads merlin el mango's post and subsequent posts on the previous page,
those are interesting transformers. According to Edcor's web site, they are good for 100mA primary current, and they are 15K primary with UL tap.
EDCOR - GXSE5-600-5K and GXSE5-600-15K max DC current
The model number on the top is misleading unless you read and see that the primary is 15K and the secondary is 600R.
They have since then (2011) changed the model number to be more in line with their standard convention. The transformer is now listed as the GXSE5-15K, and one has to select the output impedance in a drop down menu.
Last edited:
