CurtPalme.com Home Theater Forum

stridsvognen · Guest

Some time ago i was so unlucky, or stupid to let the smoke out of my Ultra convergence board, and was forced to repair it.

While waiting for parts, i was looking and checking the components over and over again, there is nothing i hate more than blowing stuff 2 times.

Suddently it made sence to me the way the convergence is drifting on the Marquee.. The 3 channels are not identical, Red channel having all its transistors on one side of the board, while blue and green was split on top and bottom of the board.

I have messed a lot with thermal cupling on power amps, and its the same thing that happens with DC drift when all transistors dont have the same thermal conditions.

I asked Curt for some defect convergence boards i can play with, and got them yesterday.. I repaired one of them right away, and tested it, its working.. It needs a bit more love, someone changed the caps and the 24V supply resistors, and dident do a very nice job, and a blown red chanel had been atempted repaired, with not to much luck.

Well the main idea is to get this board up running perfect, i have replaced all 1,2ohm resistors and 2ohm resistors with new 0,6w 50ppm resistors, and will do the same to all the resistors in the 24V suply, and the jumpers on the 24v is removed.

When i know what all these resistor changes do, ill continue moving ½ the red output transistors.. Q100 Q101 Q102 Q103 Q104 to the other side of the board, giving red convergence channel same thermal conditions as the others.

Anyone have some experience with this ?

I like to improve the drifting without adding active coling.

gjaky · Joined: 05 Jun 2010 Posts: 2802 Location: Budapest, Hungary

I am pretty sure electrohome also suggested at least 20minute warm up time for the projector, after the 20 min. the heatsink should evenly warm, so tthe thermal difference could've been ruled out. Of course there are a lot of possibilities. My 7500 was rock solid I think.

Nashou66 · Joined: 12 Jan 2007 Posts: 16171 Location: West Seneca NY

Its in my My Marquee thread. MP suggested many years ago to get larger Resistors to help with drift and to also clean the thermal
contact of the amps to the heat sink. the faster everything gets to operating temp the less drift there will be.

Nashou

stridsvognen · Guest

Sure 20 min do a lot to it, but it would be cool if its ready to fly 1080P after 2 min. And have less drifting.

Im sure everyone with a well setup Marquee running 1080P is tweaking convergence some times.

I do.. on all 3 Marquee i have.. And they all need more adjustment to red, i think red is actualy the one most stable, blue and green shifts right under heatup, so the idea was to make red do the same.

And if changing some resistors to low ppm helps, then they it should also be less sensive to temperature changes.

stridsvognen · Guest

gjaky · Joined: 05 Jun 2010 Posts: 2802 Location: Budapest, Hungary

You should notice that there are three smaller and two bigger transistors in each amp, the small transistor tied between the two big is responsible for the thermal stability of the amplifier, you should try to improve their thermal coupling further. Anyway, how serious is the convergence drift during warm up (expressed in test pattern thickness)?

gjaky · Joined: 05 Jun 2010 Posts: 2802 Location: Budapest, Hungary

Actually, what I would find an interesting experiment is to replace the analog (class AB) amplifiers with class D amplifiers. This would mean reduced heat consumption (increased efficiency), and maybe less heat dependence...

stridsvognen · Guest

I dont think ill want to reconstruct the board, thats a compleetly different thing.

The thing im focused on here is the obvious non symetrical layout of the Marquee concergence board, and how easy it is to try change it.

There is many ways to attack this problem.. Design a watercooled convergence board .. its possible.. but not sure anyone like to pay for it..

Keep the budget low.. Wink

jbmeyer13 · Joined: 03 Dec 2010 Posts: 1135

stridsvognen · Guest

jbmeyer13 · Joined: 03 Dec 2010 Posts: 1135

Never checked the temp. Prior to changing out the caps it would take 15-20 minutes for my convergence to stabilize and there was slight drift over time.

One other issue to consider is mechanical set up. The better that is I believe the less strain on the CVA. I wish I could provide you with a more academic answer on the CVA/caps but it's simply what I have found as a result. I also replaced caps on Astig and VDM but I don't believe that made any difference to the image.

mp20748 · Posted: Tue May 28, 2013 9:56 pm Post subject:

mp20748 · Posted: Tue May 28, 2013 10:00 pm Post subject:

jbmeyer13 · Joined: 03 Dec 2010 Posts: 1135

Well in my case I had removed the back plane boards multiple times for maintenance, trying different caps, etc. and never saw any change to conv stability until trying the wet tantalums. If not the caps then it sure is an interesting coincidence. What on the HDM/VDM do you feel could be causing the issue?

jbmeyer13 · Joined: 03 Dec 2010 Posts: 1135

mp20748 · Posted: Tue May 28, 2013 10:18 pm Post subject:

stridsvognen · Guest

jbmeyer13 · Joined: 03 Dec 2010 Posts: 1135

mp20748 · Posted: Tue May 28, 2013 10:49 pm Post subject:

gjaky · Joined: 05 Jun 2010 Posts: 2802 Location: Budapest, Hungary

Now I think I've found out the ultimate solution Very Happy

But certainly nobody would willing to do it...
During retrace the convergence amp probably stays idle, ie. 0V output, so, at that period one should take a sample of the output (and hold it until the next measurement is done), that would give the DC drift of the amplifier compared to ground, with thess a servo circuit would be able to set the lowest possible DC drift, and re-set in every retrace period.

As I understand the amplifier is a fully current feedback construction, so the impedance change of the yokes (due to the heat) would directly affect the gain factor, thus the DC drift.