I want to build a stereo headset valve amplifier

[Herald1360]

Murphy 310 has a valid point......

However, I simply took the design as it stood (and as it will probably be built by a novice) and pointed out a rather glaring error in the PSU design as presented. Given that the layout may be less than ideal and solid state bits are cheap and cheerful, DC heater supply can do no harm. I've used it successfully myself (admittedly as a quick fix before selling it on) on a rather high gain (for mike input) small self contained PA box I lashed up years ago. It killed off the residual hum from the first stage stone dead.

45 years ago a solid state regulated DC heater supply wasn't really an option- certainly not commercially- so one had to do it "properly", maybe a humdinger would have featured somewhere, though. Nowadays a suitable wirewound pot might well cost more than the regulator bits

The zener diode idea would not work as described. To make it function in this application would need high powered (ie expensive) zeners and would be nowhere near as effective as the 7806 fed from a suitable unregulated DC supply.

[Top Cap]

I have to agree with Murphy 310 considering the low gain involved here. A grounded centre tapped heater transformer may help or in the case of a single winding, a 100 Ohm resistor taken to ground from both sides of the winding. However, good construction may be the only thing to consider here, using twisted heater wiring kept close to the chassis, formed into corners if need be but kept away from the input wiring. I successfully built a three valve phono preamp where because of the high gain (input only a few mV) I used DC heaters with a simple series transistor (see attachment) but I needed to use a 0-9V transformer. The wire wound resistors helped to keep the dissipation low as possible in the Darlington pass transistor. To add belt to braces I lifted the heater supply to +50V using a potential divider from the HT line. Not the slightest hint of hum. I suggest you should pursue the construction with just the twisted wiring, then try the amp. You can always go back to modifying the heater supply if you find you need to. That to me would be your best approach.
Les

[Don_Zalmrol]

Well I am just following the schematic I've got from that website it seems that the hum in the previous design had issues with a hum, so Kurt Strain changed the heater supply from AC to DC.

12/8/98: Revised power supply (figure 2) for improved ripple rejection and channel isolation.

5/3/99: V2 changed from 6922 to 6CG7/6FQ7 for improved stability. The 6CG7/6FQ7 is pin-for-pin compatible with the 6922 and no other circuit changes are required. The author advises that this substitution be made for existing builds because the 6CG7/6FQ7 tube is better able to handle the power dissipation of the circuit and will last longer.

Also upgraded filament power supply (figure 2) from AC to DC for reduced hum and greater stability.

5/21/99: Added pictures of OTL amplifier built by Jason Portman.

5/25/99: Ole W. Saastad built this version of the OTL amplifier:

So I am following the currently designed Ole W. Saastad design that improves on some views of the Kurt Strain design.
Unless you guys really find it to be "extreme" to use a DC heater system instead of AC. So if I would implement AC, I just need to wire wound the 6,3V directly into the heaters.
This would make the schematic much more easier instead of the DC design! But why did they (the designers of the circuit) implemented a DC system?

[Don_Zalmrol]

To late to edit, I've answered my own question about why they changed the AC to DC.

From what I can see is that they did not wire wound (used twisted pair) the heater supply cables! (see the attachment)
This could result in the hum like Trevor says.

I will build this amp without the DC heater supply, if I notice that the hum would be excessive, I can always implement the DC design to cancel out the hum.

Do I need to order a special kind of wiring for the heater supply or would a normal 1mm/ 1,5mm cable suffice?

This will make the circuit easier to build. But I think I will order a different transformer that supplies 2x 6,3V so I can split the valves more equally.

The audio input comes from my TV (tele), not my computer.
The computer is connected to the TV trough HDMI and the TV would be connected trough shielded analog cable to the headphones amp.

Is there anything else I should be concerned/ improve about the circuit?

[murphyv310]

Hi Don.
The above picture really says it all. Without being blunt or unkind the whole thing needs to be ripped apart and properly laid out.
If you look at quality valve amps from years ago the wiring would be neat, properly formed, components would be mounted in order. This was done for two reasons firstly to have short cable runs, reduce the possibility of instability and hum, and of course pride in the finished article.
I am not totally against DC fed heaters but in all honesty in a one or two valve headphone amp (per channel) that has only a few DB gain it is quite pointless.

[Don_Zalmrol]

Alright then, I will not use the DC heater supply. And yes, when you look at the picture the circuit layout is quite horrible.

I think I have everything I need to order the parts, do I need a transformer with 3 windings (1x high voltage supply, 2x 6,3V heater supply) or is the Hammond 369HX enough?

A side question: What would the sound quality output be? Excellent, good, average or bad?

[glowinganode]

Wow, that's a good little transformer, with a bias winding as well. This looks about the right size for a stereo p-p amp with around 20mA standing current per valve.
What rectifier do you intend to use? It is usual to have a seperate winding for the rectifier valve.
There is a 5V winding provided for this purpose, to power for example a 5Y3 rectifier.
One 6.3V winding is fine, 6A will run a lot of valves.
As suggested elsewhere, create a centre tap with a pir of 100 ohm resistor, and conect the CT to chassis.
DC heater supplies are only strictly necessary with directly heated valves (in this context).
I wholeheartedly agree with Trevor's comments on fads and construction techniques.
I also think the 6V6 makes a cracking audio valve, I used a pair of them in the last amp I built.
Whilst the bench is clear, I'm currently toying with a 12AU7 p-p amp.
Results so far are good, with around 1 Watt of output.
It's based on the Mullard 25W audio amp, less the two EL34's.
Have you decided what output valve, and hence what output transformer you intend to use?
Rob.

[Don_Zalmrol]

The rectifier would be a bridge rectifier 500V 5A, so not a valve rectifier.

This amp would be used to only power my headset (Sennheiser HD558). And would use 2x 12AU7 (pre amp stage) valves and 2x 6CG7 (output stage) valves in parallel.

The power output is 25dB (316mW) so this will suffice for powering headphones (I think).

Now I'm starting to think about it, do I need a volume control?
I use the Sennheiser volume controller which is next to my headphones.

The TV is in front of my bed, so to stand up and lower or raise the volume each time is ridicules…
And when I'm going to a LAN or a friend I would use this amp to power my headset since it's small design…

So the sound quality would good or not?

[glowinganode]

Hi Don,
firstly the rectifier should have a piv rating of at least 800V, and preferably 1000V.
With a capacitor input filter each didoe will see at least twice the peak voltage across it when reverse biased. The peak voltage is 1.414 x the rms voltage. With a 225Vrms winding, this works out at around 320V.
This is also approximately the no load HT voltage you can expect.
With a centre tap winding, you will only need to use two of the diodes (three connections on the rectifier).
How will it sound?
That's impossible to predict.
The amplifier will be as good as the design, manufacture and components allow.
By home brewing, you can at least know that the ironware will be good, Hammond transformers are very good, as are VVT made here in the UK.
These are probably the most important and expensive parts of the amplifier, so it pays to use good quality components.
Saying that, I have built amplifiers using all sorts of transformers of unknown heritage, with still very good results.
The one thing that may catch you out is parasitic oscillations, especially around the phase splitter. Construction techniques and layout play a big part in elimanating these. You may need to add additional screening around the valve bases, and obviously you need to be able to check for their presence.
An oscilloscope is really the only way to do this, you will need access to one if you intend to take this up seriously.
Rob.

[Don_Zalmrol]

Thx for the reply Rob,

The rectifier (B500C5000-3300A) can withstand up to 1000V (piv), as you can see here in the data sheet.

The components are all good/ decent quality, the design hold some errors on the PSU which Chris, Trevor and Les ironed out. Now for the amplification side they could not say, so it would be trial and error.

As I see the creator this amp, still uses it. The last update/ revision was in 2008 and he changed his circuit to the one from the bloke from Sweden (with DC heaters).

I don't have a scoop, but might be able to use one if necessary…
or buy one, the are becoming quite cheap +- € 100 (1 channel from Velleman).

But can't I cancel most of those parasitic oscillations out by grounding the chassis with the ground connection of my house?

- Laurens

[jjl]

Separate 6.3V AC heater windings for the 6CG7 and 12AU7 may well be a good idea with this circuit.

The reason being that the 12AU7 cathode will typically sit a long way above 0V in this circuit, whereas the 6CG7 cathode will be much closer to 0V. Sharing a heater supply may stress the 12AU7 heater to cathode insulation. Indirectly heated valves have limits on the heater to cathode voltage that they can withstand, if the limit is approached or exceeded, all sorts of hum problems or worse can be encountered.

Not wishing to be unkind, but my hunch is that the originators of this circuit encountered hum problems and did not understand the cause, hence they resorted to the use of DC heater supplies.

John

[Diabolical Artificer]

Saw this,thought of you. http://www.dogstar.dantimax.dk/tubestuf/miniblok.htm. Cant find anything as simple as that. Should be very easy to pop a headpnone socket on. Andy

[murphyv310]

Hi.
The ECC8* range has been used in series fed heaters in many TV sets and are quite happy with 100v p-p to cathode voltage with no ill effect, the 10-20v H-k volts strapping the heaters with the o/p valve will make no odds.

[Don_Zalmrol]

First of all happy new year!
Secondly Thx for the link Magnetic!

So I don't need an other transformer for the heater supply for the 4 valves?

If needed I will order the Hammond 370CX for JR.

Also is it a problem that the transformer delivers 320V instead 330V, or this doesn't make a difference even if the transformer would deliver 389V?

[Don_Zalmrol]

If I don't need separate windings for this amp, I will order the 370AX which is a more lightweight transformer:

Total power 44 VA

Secondary: 240-0-240 Volt
DC current : 50 mA
Bias tap: 50 Volt
Filament 1: 6.3V - 2.5A

So to copy from Rob's post the peak voltage would be:

The peak voltage is 1.414 x the rms voltage. With a 240V winding, this works out at around 340V. Which is only 10V more than the original circuit.

[Don_Zalmrol]

I've redrawn the PSU circuit to work on AC heaters and not DC.
Is this good?

[Herald1360]

Connectivity wise it's fine, but the layout of the connections needs to be considered. If you think of each line on the circuit as a small resistor (as they are, even if only milliohms), you need to avoid having ripple current (the AC current through the smoothing capacitors) or heater current flowing in the same wires as supposedly smooth HT DC.

The redrawn circuit below illustrates this separation if you compare it with the original.

[glowinganode]

Originally Posted by Don_Zalmrol

Secondary: 240-0-240 Volt

Hi Laurens, as this transformer has a centre tap HT winding, you need a full wave (bi-phase) rectifier, not a bridge rectifier. See here.
I'd also add a resistor in series with each rectifier, say 47ohm to limit the surge current at switch on.
Rob.

[Don_Zalmrol]

I've revised the PSU with usage of the bi-phase rectifier and centre tap from Rob's schematic and implemented Chris's corrections.

I've also added the 47 Ohm resistors as Rob suggested. I will use 1W resistors like the other 220 Ohms resistors.
If all this is correct I will place the order today. And with luck I will get my parts by the end of the week or some time after that…

The rectifier diodes I've placed in the basket are 2x fast recovery diodes with a max input of 1000V peak and 1000V inverse input state peak, which can channel 5A of current.

The resistors are 2x 47 Ohms 1W with 1% tolerance, metal film types.

Isn't it better that I change all the rest of the 0,6W resistors to 1W resistors?

[glowinganode]

Up to you Laurens, the 1W resistors will probably look a bit more "authentic".
Don't forget to create a centre tap on the heater winding with a pair of 100 ohm resistors in series.
Rob.