CurtPalme.com Home Theater Forum

[dvh99]

. pic was wrong
the link i got will do i have to learn to write those things down, cause nashou gave me those values a while back.

[draganm]

this shows the circuit

http://www.curtpalme.com/forum/viewtopic.php?p=178247#178247

[dvh99]

peaking is for 1 set frequency, but is peaking dependent on drive g2 brightness and contrast setting?

[Nashou66]

No.

[dvh99]

nash have you tried peaking with the el5166 on the 03p vim?

[tse]

[dvh99]

tx scott that is what happened exactly.
so how to prevent this on the 03p vim, add another trimcap on the opamp after the ad834 with a 10ohms in series parallel to the feedback resistor?

[tse]

There is only so much that can be done at the VIM level. Better performance ultimately calls for a better neck card. Good luck in finding high bandwidth, high voltage PNP transistors anymore. They have disappeared since CRT sales have dried up.

Scott

[Nashou66]

Scott what are the ones you used on the new board better than the Motorola's? I know a company that makes discrete power regulators, I wonder if they can make discrete high bandwidth, high voltage PNP's, doesn't the VNB also use NPN also?

Athanasios

[mp20748]

Nothing is/was better than the Motorola's for a discrete high bandwidth power transistor. The Motorola's were very unique mainly because they were high bandwidth RF power transistors that could be used in mirrored operation.

I used to have a data sheet on them.

There are still some NPN RF power transistors available, but the PNP's are no longer available. And that's the main part that prevents this design from still be manufactured.

[tse]

The Mot parts are definately the best for this application (the parts on the neck card have the same characteristics as the ones attached other than increased power dissipation because of the package that allows attachment to heatsink).

Sanyo made a line of video output type transistors that could be used for higher bandwidth amps but at lower voltage than needed for the projector CRTs (ok for monitors).

There are a couple of Sanyo parts with the voltage needed but required more overdriving at the limits of bandwidth to match the Mot parts.

I believe that technique could be applied to the Mot type neck cards to get increased bandwidth over standard cards. Peaking network like C68 and R101 could be added to a better opamp. L1, L2, L3, and L7 should be lower value (to resonate at higher frequency). Probably, R84 and R85 should go away. R16 and R24 values would have to be optimized. L4, L5, and L6 might have to be slightly different value.

If attempting this I would recommend first replacing the inductors with 0 ohm jumpers and setting the peaking network for quickest rise and fall times at amplifier output. Then place the inductor values to get the shortest rise and fall times.

When I did this for the new VDC neck card I used 400MHz scope, 500MHz probes (that were close to the CRT capacitance), and pulse generator with rise/fall time less than 1ns.

Scott

[dvh99]

if i understand correctly this peaking isnt dependent on resolution but only sets the amp for fastest rise and fall with the proper inductor values.

very helpful answer, thanks again.

http://www.farnell.com/datasheets/5092.pdf

dennis

[tse]

[Nashou66]

Well i adjusted the peaking circuit tonight to have it set for 96hz refresh rate. Now I have to redo my Calibrations on both PJ's as it screwed it all up!!

One thing I noticed Scott is the pattern changes color as you move out from the center of the screen. Probably because of the light drop off on the edges of the tube. I wonder if this would be a good way to do contrast modulation? Move the pattern to the edges and corners and then adjust the color zone contrast till the vertical and horizontal boxes match intensity? Would that work?

Athanasios

[dvh99]

as if you like the pj to be set up so you cannot play with it anymore

[Nashou66]

[dvh99]

getting the peaking right should be done first.
when doing peaking brightness is increased because the hor resolution isnt attenuated as it was before.
you will actually need less gamma because now i have much more shadow detail with a lower gamma setting.
g2 didnt have an effect from what i saw on the peaking its only the contrast setting so once set do not touch the contrast button.
i really want to play further with the neckboards to see if this horizontal overshoot can be reduced so i get an almost black white 1 on 1 off.

[Nashou66]

What happens when i move the pattern towards the edge using the phase you can see the pattern get a green tint to it.

Put up the pattern and shift it towards the edge of the tube in the center and see what happens. you can turn the trim pot while at the edge and fix it but wen it goes back to the center obviously it will be off there now. I think we need a zone peaking circuit lol

Athanasios

[dvh99]

i haven`t done anything to the neckboards yet (besides changing out the electrolytes, opamp and carbon resistors) but i still have some inductors and varicaps to play with.

the question is what values i should be thinking of for r16 and r24 (they are 390ohm).

scott said they should be slightly different, but what is slightly?

all help is welcome.

dennis

edit: almost forgot about the smb connectors , that has been done too.

[gjaky]

It's an odd question, and off topic too, but: is there any tweak to make better the old and creepy -03p VIM, or the only way is the replace?