• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Advice sought on thermionic headphone amp

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I wonder if any forum members could offer practical advice on the construction of this circuit:

Class_A_amplifier_headphone_amplifier

I think it might be more demanding in some respects than the equipment I usually work on, which is musical instrument electronics.

I'd like to avoid any errors or oversights that might spoil my enjoyment of this headphone amp when I use it in its intended environment. Perhaps some members might have useful practical experience they could share? Serving and ex-military comms guys might have the most to offer.

 
Personally, I prefer solid state rectification and regulation, it's more efficient. Sourcing the transformers may present the biggest challenge, but otherwise this is a fairly textbook two stage single ended amp. I personally prefer using a triode rather than a pentode in the first stage and you might be able to replace the cathode resistor/capacitor with a LED string that drops the appropriate voltage, which many people prefer.
 
Hi Simon, Nice little headphone amp, with the right enclosure would look nice. Any idea of the parts costs.
For construction, I would reccomend a 2.4-3.2mm thick double sided PCB with quite heazy copper (2oz min, if you use 2oz on a PTH board, it will be about about 2.5 ounce the plating process has finished). Though the design will probably go on one layer with the odd jumper, going for a two layer solution will give more options. Then assembly is just like any PCB. This way avoids any wiring errors, if you go from (checked) schematic direct to PCB. You have to observe the creepage and clearance spacings on a PCB, but easily achived. Also you can pick your enclosure and match the PCB to it, making assembly easier. The thicker boards are much more rigid than standard 1.6mm boards, when you are putting valves in their sockets they feel solid and give a impression of quality, plus they make a rigid assembly less affected by vibration and shock, not to much of a problem in a houshold envoirenment, but in more taxing envoirenments it is critical.
 
Thinking a little more, depending on your tastes...
You could do a retro design with curvy tracks with teardrops on the pads, brown bakerlite sockets, iron cases for the transformers and a plain varnished PCB (no coloured solder resist). Or a modern sleek design with MU metal shields for the transformers (mounted on the PCB) with modern routing, nice sockets and a nice black resist to contrast the metalwork.
 
I would search around for a better headphone amp to build up.

these "designs" are littered with bugs because they are very generic.

try to stay away from odd tubes. the more standard the tube types used, the ease of swapping and availability of replacement is a great advantage.

I take an "old school" approach to building - Point to point.
I know a lot of people don't have a friend that is a machinist.
precut chassis are available. There are electronics merchants on the web that cater to DIY audiophile and guitar amp building.

tube vs solid state rectification, I recommend using solid state but put the high voltage DC on a separate standby switch. Tube rectification is an added cost in tubes and power transformer requirements.

I posted on this fourm techniques for heater supply design as well as others. The simple approach seems to work for me.

When I build point to point, I keep all of my high voltage away from signal lines. My heater wires are assembled in twisted pair. and routed physically as much as possible in 90 degree angles from the signal path.

Placement of tubes in a staright line helps construction as different sections are identified easier. a good rule of spacing is 1 to 1 1/2 the socket distance as this will help the tubes and the parts ventilation.

try to place power transformer and choke (if used) as far away from the output transformer. in stereo output, place one in 90 degrees orientation from the other channel this will help preserved .

this is a picture of point to point design on a prototyping chassis. I use polypropylene caps where ever possible. and 1 w resistors that are temperature stable ( less than 50 ppm ).

electrolytic caps should be low leakage or low impedance (low esr) type.
 

Attachments

  • tubeotl.jpg
    tubeotl.jpg
    940.5 KB · Views: 188
Thank you for your replies so far.

I'm thinking of using a vibrator-type power supply for the HT,
possibly a synchronous rectifier type, fed from a rugged sealed lead acid battery, though I haven't worked with this rather old technology before.

I'm not necessarily averse to using a pcb, though I need to research which board materials are least prone to absorbing moisture as this could obviously be a problem. Even point-to-point on turrets or eyelets might be susceptible to this, though I should think the range of materials that these can be mounted on is basically anything I can drill suitable holes in. Perhaps with attention to creepage and clearance spacings, and the right board material, a pcb would be suitable.

I need to source suitable tar-potted transformers and the valves used will be metal can pentodes.
 
Er wouldn't worry much about moisture content, best material is a decent Tg rated FR4. Unless you are going into outer space or somehwere, the moisture absorbancy of the PCB material wont be a problem, I know this for a fact as I have done boards with high voltages that work in -40 to +85 all around the world no problem. Even if you design to pollution class 3 (not likeley in your house) the creapage is 8mm for 600V peak.
All you can do with point to point wiring can be achieved on PCB's, thats why they came along, I have valve amps from both design types, so I have no view on the effect on sound, but a PCB based project is an easier build.
 
Thank you for your replies so far.

I'm thinking of using a vibrator-type power supply for the HT,
possibly a synchronous rectifier type, fed from a rugged sealed lead acid battery, though I haven't worked with this rather old technology before.


wait... this means that you are going to have all the trouble of using a battery to supply clean DC, only to connect something ( vibrator) that introduces all sorts of frequencies? Better just use AC from a wall plug.
 
Just for interest,

I have never built this, however the real thing sounds quite nice when I had a chance to listen..Has anyone ever buit it?

HeadWize - Project: The Morgan Jones Mini Tube Headphone Amplifier by Chu Moy

Yes I did... quite a while ago.
It is a nice sounding amplifier, and I improved with with a somewhat better CLCRC supply and LED bias in the first stage. There was a thread about that on the Headwize forum...dunno whether it still can be found...

My later builds though using transformers (SE or Parafeed) do sound a good bit better ;)
 

Attachments

  • MJ-headamp-1.jpg
    MJ-headamp-1.jpg
    63.8 KB · Views: 98
which headphones?


headphone impedance, sensitivity vary by orders of magnitude - sensitive multidriver IEM need very high step down ratio, low output Z given their lumpy impedance curves

some dynamic headphones require 20 Vrms, 400 mA, others 200mV rms, 10 mA for 120 dB SPL peak capbility

amp topology and output xfmr design need to targeted for a much smaller range for decent performance - not just "I'll add another xfmr output tap"
 
Last edited:
Disabled Account
Joined 2010
Yes I did... quite a while ago.
It is a nice sounding amplifier, and I improved with with a somewhat better CLCRC supply and LED bias in the first stage. There was a thread about that on the Headwize forum...dunno whether it still can be found...

My later builds though using transformers (SE or Parafeed) do sound a good bit better ;)

I quite like the idea of "small" HP amp..LOL

Can you remember component values or do you have a circuit for your mods for the Ear max?

Regards
M. Gregg
 
which headphones?

All Sennheiser... HD600, 650, 800 high impedance 'phones.

Can you remember component values or do you have a circuit for your mods for the Ear max?

Sure I can... my built was based upon the latest schematic and then converted to LED bias. I even had help from the master himself (MJ) sort of tutoring me through the upgrade. PM me with an email adress and I'll send you my schematic :) (not to lead this thread too much OT).
 
the 600-800 Senns nedd ~ 12 Vpk, 40 mApk (round #, includes some "dynamic peak headroom" without clipping rare ~120 dB SPL peaks)

that is in range of OTL design with bigger tubes if you don't insist on the "flavoring" of a output xmfr - could be cheaper since quality output Iron will run US$50+ per channel
 
My headphones are high impedance (400ohm) DT100s

I appreciate that members have favourite amps that they like to discuss, but my original post was only about the amp I gave the url of:

Class A headphone amp

It's only this amp I'm concerned with, so discussing other amps isn't really that helpful to me.

Mains power would be quite infeasible for me, hence my interest in the vibrator type supply, an interesting old technology, and from the heyday of valves too - they were used in many applications including car radios (think "Motor-ola") and, according to a friend who's an ex Royal Marine, some of the radio sets he lugged through swamps and jungles.

FR4 apparently has near zero absorbency so hopefully that won't be a problem.

Project documentation in DjVu is also an excellent idea. It's great that LizardTech have chosen to use open standards, so that anyone who wants to can read the small print and fully understand what's involved, thanks for putting the url up DF96.

In the same vein, if I need to post a graphic, I'll use .png format (always makes me think of penguins somehow) - though it's a shame that the forum rules allow much smaller .pngs than .jpegs.
 
Disabled Account
Joined 2010
Ahh,

So you are looking for a portable tube amp....With O ring /gasket seals and water resistant circuit board...A vibrator circuit and Lead acid PSU..

Sealed environment...It makes sense now.. :)
I think you need to rethink the circuit...Even with button trodes "9 Pin" you need vibration mountings..and a way to remove the heat..
Interesting Idea though..
The problem was that the reed switches in a self energised coil did have a habit of wearing out..Cast aluminium case with Oring seals..and rubber mounted clamps for the tubes..Heater temperature can be an issue.. Point to point works well...Circuits boards are less durable..Coat with varnish or lacquer..The circuit needs to be in a seperate compartment IE sealed box within a box. with tubes in the outer box access to connections via tube sockets held in place with the push and twist Mu metal screening cans..these can be drilled with holes to give better heat transfer they can be clamped so they heat sink to the outer box..depends on how sealed the environment has to be..
HT flash over can be an issue again keep distance between voltage levels...Gold type sockets don't work well in this evironment..plating strips and corrodes (Silver is better) but the old style are as good as anything (Not the new cheap things)..Fins are better any ventilation needs to be carefully considered..

Hope this helps..

Regards
M. Gregg
 
Last edited:
the 400 Ohm DT100 have higher Z, lower sensitivity than the Senns, for similar headroom ~18 Vpk, 45 mA drive capability would be my idea of adequate

the OP schematic is a 4/8 Ohm loudspeaker amp with a R divider on the output for a headphone - lots going to waste there - especially for a battery powered project

for some perspective the 18 Vpk I'd like an amp to supply for the DT100 would correspond to 20 Wrms into a 8 Ohm load


the shown SE output amps is Class A – dissipates more power continuously than it delivers to the load – you’d need several kg of lead-acid battery per hour of run time (not counting DC-DC loss)



mechanical vibrator supplies are considered antiquated for good reasons - today the EMI from the mechanical contact opening is unacceptable; size, noise and efficiency can all be improved hugely with modern DC-DC SS designs with integrated chip controllers
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.