Bias Supply

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Jump to Latest
T

tyu

Member
Joined 2009
bolserst.........you have my trust...you have raised my understanding of all things ESL..or any other subject ..that I have read your info, input on....your memory is over the top!
but I have to tell you ...how I see it...when I am in a room were speakers are playing ....I am in the signal path.....my ESL bias point

For years now I have been buying older Martin logans an reworing the panels when I can.....over 10 pr.... in all the pic I can find....of the ML bias boards
look at this pic...of my latest ML finds... bias board....the one with the brown caps came out of a $10k pr of ML Prodigy...an is in all others ML I have found...back to 1995...not the cls..clx...
the bias board with the blue cap came out of a pr ML Ascent floor models I pic up for a song...... these blue caps give a sound...I have never heard out of any ML....not just one mans opine..
So from what Iam hearing..an others...night an day diff ...only thing diff in the boards is the caps... what added too the bias setup... CAN make the sound of a ESL....
I know the a diff cap can change the voltes....but that still changes the sound....right
Your a physic guy...I get it..an try to live there my self.....but I can hear big diff in the sound of any thing I do to the ESL bias.... with all do respect!...thanks for just being there..an yes I well speak for others...gladly
 

Attachments

  • cap 003.JPG
    cap 003.JPG
    736.1 KB · Views: 88
B

bengel

Member
Joined 2010
Sorry, still pondering this :).

Say the light is on and thus my diaphragm charges up to 4000V...then the light goes off. How much of a voltage drop must the diaphragm experience before the light turns on again and charges it back up to 4000V? Presuming the bulb trigger voltage is 100 VDC, would it be 4000V-100V = 3900V ??? i.e. it "bounces" back and forth between 3900V - 4000V.
 
Calvin

Calvin

Member
Joined 2004
calvins-audio-page.jimdo.com
Hi,

yes, that is the way it works.
When the voltage difference between supply and membrane reaches the treshold voltage of the lamp, the lamp fires and recharges the membrane.
So the membrane voltage 'bounces' slowly up and down.
One last remark:
The lamp's firing treshold (say 94V) is typically specced in the DS.
If the lamp fires, it doesn't burn till the voltage over its pins drops to 0V, but it switches off somewhere between 50-60V.
Taking the above example the membrane voltage never reaches the 4kV of the supply but would 'bounce' between ~3.90kV and ~3.95kV.

jauu
Calvin
 
bolserst

bolserst

Member
Joined 2005
I agree...usually see voltage across neon lamps ramping between “strike” voltage level of about 95V and “extinction” voltage level of about 65V. If you are so inclined, LTSpice includes a neon lamp model in which you can define these voltage levels and the on-resistance. It will give you a feel for how the blink rate changes with component values and, unfortunately, bias voltage. If you increase bias voltage you will notice blink rate increases. This does not mean the resistance of the leakage paths have changed, just that the leakage current is increasing since it is proportional to applied voltage.

The attached LTSpice screen capture shows sim for the default neon lamp model parameters of 100V & 50V. Notice the 100K resistor in series with the neon lamp. You can use a resistance here to increase the “extinction” voltage if you want to try and minimize the amplitude of the voltage cycling. The downside is that the “ON” flash of the neon lamp gets shorter and fainter when resistance is added; blink rate also increases. If the resistor is too high in value, the lamp will dimly flicker at a high rate which is not useful for monitoring leakage currents.

@ bengel/tyu
Note that the panel mount neon lamps you guys posted links for include a resistor in series with the neon lamp inside the assembly. Depending on the value of this resistor, you may need to bypass it to get an easily seen blink indicator.
 

Attachments

  • LTSpice_Neon_Charge_Indicator.PNG
    LTSpice_Neon_Charge_Indicator.PNG
    88.3 KB · Views: 80
B

bengel

Member
Joined 2010
Wouldn't this "oscillation" be a bad thing? i.e. wouldn't the spl be, in effect, "modulated" by the changing diaphragm charge? If there was leakage (always is) and you just had it hooked up directly to the 10 MOhm EHT output, the voltage would stay the same (unless the leak exceeded the current thru the 10 MOhm resistor????).

Sure a 100V oscillation would have almost negligible effect on the SPL though - if I did the math.
 
bolserst

bolserst

Member
Joined 2005
The 30V variation in bias voltage at a frequency < 1 Hz would cause the sensitivity of the ESL to slowly ramp up and down by less than +/- 0.05dB; less variation for higher bias voltages. So, not really a concern, but easily bypassed with a jumper when not testing for leakage if it bothers you. Like tyu I find the flashing can be distracting if visible from the listening position.

If you have significant leakage in your panel, most ladder multiplier supplies will have supply ripple of similar or larger amplitude even without the neon lamp circuit...and this ripple is varying at audio frequencies so more of a concern from modulation standpoint.
 
Last edited:
bolserst

bolserst

Member
Joined 2005
tyu said:
…I know the a diff cap can change the voltes....but that still changes the sound....right
Click to expand...
Absolutely. If a different capacitor (type or value) results in different bias voltage, this would be noticeable as a change in the ESL sound. My experience is that if you adjust the input voltage for two different bias supplies such that the output voltage reaching the panel was the same, the type and/or value of capacitors made no noticeable difference in sound.

tyu said:
With my Acoustat 2+2s.....I can pull the ac on my bias .......an at med loudness...I get output for 3-5 min.... with out much loss in loudness output..
Click to expand...
Just pulling the AC plug allows the panels to keep using the energy stored in the HV supply multiplier capacitors.
How long do the panels keep playing if you disconnect the panels from the bias supply?
This would be the indication of how much leakage your panels have.
 
B

bengel

Member
Joined 2010
bolserst said:
If you have significant leakage in your panel, most ladder multiplier supplies will have supply ripple of similar or larger amplitude even without the neon lamp circuit...and this ripple is varying at audio frequencies so more of a concern from modulation standpoint.
Click to expand...

Just curious, could this be mitigated by using large caps in the ladder? (i.e. reduce ripple)
 
bolserst

bolserst

Member
Joined 2005
bengel said:
The only downside is 3kV+, .1uF caps are not cheap :)
Click to expand...
If all you are concerned about is ripply, you only need to upsize half of the capacitors.
http://www.diyaudio.com/forums/planars-exotics/190716-esl-power-supply-question-6.html#post3772446

Increasing the frequency of the voltage applied to the voltage multiplier and/or using fewer sections in the multiplier also reduces the ripple and allows use of much smaller capacitors. This is why I recommend using a CCFL HV oscillator to drive the multiplier rather than the more traditional 50 or 60hz wall powered supply.
http://www.diyaudio.com/forums/planars-exotics/190716-esl-power-supply-question-5.html#post3714088

Taking a step back though, this discussion of supply ripple is mostly just academic.
If your ESL is leaky enough to be putting ripple in the HV supply, it would probably be best to fix the leak.
But for those who wear belt and suspenders... ;)
 
B

bengel

Member
Joined 2010
Hmmm, interesting... Funny, I was just thinking along these lines last night thinking about using a 555 timer.

Forgive me, as my electronics knowledge is somewhat weak but you are saying take something like this:
CXA M10M L TDK DC to AC Inverter for Non Dimming Type Cold Cathode Lamp CXAM10ML | eBay

Feed it DC (like 12V or something), take that output (which I think is something like 1 kV) and run it through a ladder..... So if you need 4 kV you need 4 "steps"..... And you need >1 kV caps & diodes to handle it???
 
B

bengel

Member
Joined 2010
Last edited:
bolserst

bolserst

Member
Joined 2005
bengel said:
...2 adjustable bias supplies capable of running up to 5 kV ????
Click to expand...
Pretty much :), although if you only plan on using a 5 stage multiplier to get 5kV you will be running the parts right at their 2kV rating. Remember caps and diodes in multipliers see twice the peak input voltage during operation. I would recommend upping the voltage rating on the parts to 3kV or consider using a few more sections on the multiplier. Also, the 0.01uF cap size is a bit overkill when used with the HF oscillator. 2200pF will work just fine and are a bit cheaper at the higher voltage ratings.

BTW, this is essentially what the EMCO modules have inside them. It is just a bit cheaper when you construct them yourself.

ALso, your mentioning the 555 timer reminded me of this post:
http://www.diyaudio.com/forums/planars-exotics/190716-esl-power-supply-question-6.html#post3744943


(and I apologize for hijacking the OP's thread :) ).
Click to expand...
Perhaps move discussion of your HV supply to your ESL build thread?
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Top Bottom