• 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.

Morgan Jones HPA ...strange startup behaviour

Morgan-Jones HPA ...strange startup behaviour

I recently started to convert my MJ headphone amp (3x6922, see this thread at the bottom "MJ") to led bias in the first stage. Before it has been running for over a year without problems and with no hum issues.

Since the first stage runs at -3V bias (~4mA) I thought an led with "matching" Vf might be a good idea and chose a green led (3.0V).

Upon startup and after the rectifier (EZ81) warms up there is a relatively loud buzz for a few seconds followed by a not very headphone friendly popping noise (5-6 times) and then...silence!
After this initial "sequence" the amp performs flawlessly, only bias now is a tiny little bit below 2V.

Can anyone clue me in why bias it lower than it "should" be?
And does anyone have an idea about the noise symptoms?

As always, thanks in advance...
 

7N7

Member
2003-01-18 10:43 pm
England
Re: Morgan-Jones HPA ...strange startup behaviour

I recall that MJ wrote that a green led has a highish slope resistance. You would probably be better off using two cheap red leds in series although this would reduce the current somewhat.

As for the bias change, device variation in "sand" components is colossal

7N7
 

SY

diyAudio Moderator Emeritus
2002-10-24 10:19 pm
Chicagoland
www.SYclotron.com
All of the green LEDs I've tested were 2V, not 3V, so it's not surprising. You might try a red and an IR in series to get 3V.

edit: @7N7- the forward voltage drop across a semiconductor junction does not vary enormously from device to device; it is extremely predictable and consistent. It does however vary from material to material. Slope resistance for green LEDs tends to be higher than for red, but not hugely so, a couple of ohms.
 
Sounds like you have a form of motorboating.

Strange...there was no such behaviour with conventional bias by resistor and bypass cap.

Today i took the green led's out and put in a string of two orange ones (spec'ed at 10mA, bias now at roughly 3.5V)...now the popping sound (and the flashing) won't stop at all.

Any hint as to where I could look further...?


I also found this...

To get them up to the currrents required for proper operation connect the top of the LED stack to B+ with a two watt min (five would be better) resistor in the 65000 ohm range. That should bias the LEDs a ~ 4ma...

Could not enough current thru the led be the reason for the aforementioned symptoms?
 

EC8010

diyAudio Moderator Emeritus
2003-01-18 7:57 am
Near London. UK
Stixx said:
Strange...there was no such behaviour with conventional bias by resistor and bypass cap.

But as frequency went down,the bypass capacitor became an open circuit leaving the cathode resistor to produce negative feedback. With the LED, you have the same gain all the way down to DC, so any power supply problems show up. Leave the orange LEDs in and try looking at your HT smoothing. A simple regulator would probably be a good idea.
 
EC8010,

I'm adressing you since you seem the only one bearing with me...

I read up a bit on motorboating but am not able to find a reason for it in my psu...it is CLCRCRC using a choke with about 300ohms resistance and 100uF capacitors in the last three positions.

The flashing/popping stops when I disconnect the load (cheap "test" cans with 70Ohms impedance).

I am really short on experience in this matter...any further help?
 

SY

diyAudio Moderator Emeritus
2002-10-24 10:19 pm
Chicagoland
www.SYclotron.com
With all those sections, have you checked it for stability? What does PSUD predict for current bounces?

(Sorry, I'm not EC8010, but he might suggest the same thing)

I don't know this circuit- is it a single stage? If not, how is it coupled and how do the stages connect to the supply?

edit: It appears from your photo that there's two stages which are directly coupled, with B+ feed separated by an RC. Is that MJ's original circuit?
 
SY: your help is appreciated very much as well!

To make things a little easier...here is the schematic how I built the Morgan.Jones amp.

schem

It is the modified version over from headwize with a slightly different psu and one more RC section than shown. I have made it a habit to redraw every schematic I'm interested in myself (on Adobe Illustrator). It's time consuming but helps me understand things better while I am drawing away...
 
I've built a number of similar headphone amplifiers using 5687 and 5842 and I believe you now have LF instability due to feedback through your power supply decoupling to the first stage. Try any or all of the following slightly increase the value of that 1K resistor, increase the value of that 10uF, increase the value of that last 100uF in the main supply, or even convert to a current source load on the first 6922. A good VR regulator would also solve this issue.

You can also pick off the supply to the first stage early in the existing psu (right after the choke) and apply several stages of rc filtration to reach the same supply voltage - this will greatly increase the series impedance through the supply between the two stages. (Two separate supply branches.)

SY is right, when you converted to led bias you increased the available loop gain right down to DC.

With the load connected large currents flow in the output circuit and these modulate your supply which is probably not very stiff at low frequencies.

Once you get this sorted out, making the lower triode in the white CF fixed bias will net big performance gains sonically by getting rid of that huge electrolytic in the cathode circuit. I recommend that you make the bias individually adjustable and measure the current across 1 ohm cathode resistors. (one per channel) A series diode string referenced to the negative end of the power supply can be used to derive the voltage required.
 

EC8010

diyAudio Moderator Emeritus
2003-01-18 7:57 am
Near London. UK
'ere, who put that 1k resistor and 10uF capacitor in the HT? That's where your motorboating is coming from; the original design didn't have them. Chop 'em out.

Test the amplifier with even lower impedance headphones (32 Ohm?); the higher current draw will reveal any remaining problems.
 
Chop 'em out

So I did, but no change...

As a next step I simplified the psu by removing the last R and simply increasing the last C to 200uF. Of course I also upped the R before to 1K. B+ now is at a very exact 220V, but the thing is still pulsating...

The rythmic noise over my test cans is a very short electronic buzz followed by a "tock" rather than a softer putt-putt or whatever other people have come up as description.

I then carefully started to touch things (with an isolated screwdriver) and wiggle to find any loose connections...and, err, the next strange symptom occurred.
When I touch ONE of the wires that form the feedback (again not shown in my schematic but on headwize --100k from output into grid of V1) the noise stops...when I touch the OTHER wire the frequency of the noise approx. doubles...???


When there was a headscratching smiley I'd use it...
:xeye:
 

EC8010

diyAudio Moderator Emeritus
2003-01-18 7:57 am
Near London. UK
Stixx said:
When I touch ONE of the wires that form the feedback (again not shown in my schematic but on headwize --100k from output into grid of V1) the noise stops...when I touch the OTHER wire the frequency of the noise approx. doubles...???

OK, that sounds like RF oscillation causing motorboating. Time for an oscilloscope. Unless, of course, you've got something wired up wrongly. How about a photograph of the internals?