Return to the CurtPalme.com main site CurtPalme.com Home Theater Forum
A forum with a sense of fun and community for Home Theater enthusiasts!
Products for Sale ] [ FAQ: Hooking it all up ] [ CRT Primer/FAQ ] [ Best/Worst CRT Projectors List ] [ Setup Tips & Manuals ] [ Advanced Procedures ] [ Newsletters ]

 
Forum FAQForum FAQ   SearchSearch   MemberlistMemberlist  Photo AlbumsPhoto Albums  RegisterRegister 
 MembershipClub Membership   ProfileProfile   Private MessagesPrivate Messages   Log inLog in 
Blu-ray disc release list and must-have titles. Buy the latest and best Blu-ray titles to show off in your home theater!

Getting rid of raster ringing on Marquees
Goto page 1, 2  Next
 
Post new topic   Reply to topic   Printer-friendly view    CurtPalme.com Forum Index -> CRT Projectors
View previous topic :: View next topic  
Author Message
gjaky




Joined: 05 Jun 2010
Posts: 2789
Location: Budapest, Hungary


PostLink    Posted: Sat Sep 16, 2017 7:33 pm    Post subject: Getting rid of raster ringing on Marquees Reply with quote


        Register to remove this ad. It's free!
A few weeks ago someone was interested in my plans in raster ringing topic, and I've made a summary for him. I repost this information here with an intent that it could be useful for others as well, and this could be a base of an open source brainstorming, if you will.

I've spent a lot of time with studying basically all mainstream CRT projector's deflection circuits and tried to gather as many information about them as I could, even though I haven't seen every of these in action. Amongst all CRT projectors the Marquee have the worst reputation in raster ringing. While Sony G90 is thought one of the best and also there are also little complain about the Barco 909. The most accepted reason for ringing is the high interwinding capacitance in the deflection coil. Others discovered that the G90 has a significantly lower inductance deflection coils (~85uH / winding, which are always in parallel) than the Marquee (~270uH / winding, which are in parallel above 60kHz). Lower inductance would suggest lower interwinding capacitance, clear case... BUT, as I've mentioned there are little complaints about the Barco 909's raster ringing, and the 909 have an 1,8us standard retrace time, which is about the same as with the Marquee in short retrace above 60kHz. The funny thing is that the 909 have a (forum) reported coil inductance of 1,5mH (not switched) that is 10x of the Marquee coil, so one must think about its interwinding capacitance as well...

When my interest started in deflection circuits I only had a NEC XG135LC and a NEC 9PG xtra. The deflection circuit of the XG is very similar to the Marquee's, ie. there are different retrace time ranges depending on the scan frequency, but with the NEC the retrace time is fully automatically chosen, and unfortunately the lowest retrace time (1,4 us) which would be useful for the standard HD timings, is only selected in the 120-135kHz scan range, which of no use in HT. Anyway with a PC I tried a lot of exotic resolutions on the NEC to try out the different ranges. All-in-all I found the XG's deflection system having also remarkably low ringing. (A side note, I've made a hack to the XG to force the 1,4 us retrace above 60kHz scanrate, so standard HD timings could be also used, and there was no sign of raster ringing at all). I measured the inductance of the XG's deflection coils, it turned out to be 80uH/winding (that is either in series or parallel depending on scanrate, just like with Marquee) this is very similar to the G90's windings, but the NEC used the so called thin-neck tubes so the coils are no use in any 9" machines. I've also tested my NEC 9 PG xtra, that was never meant for short retrace applications as it is rated at 2,5us retrace time, but it also has a low inductance deflection coil (which is not switched over the range) of 190uH, this is very close to the XG's value, when we consider them in series (2x80uH vs. 190uH). I've made a simple setup to determine the resonance frequency of the NEC PG xtra's deflection coil. The result was 1,5MHz, I could not repeat this experiment on the XG for various reasons, but I managed to perform on the Marquee deflection coils and I got a 1,3MHz result, which is lower than the PG's but not significantly different. An 1,8us retrace pulse would equal to 277kHz frequency sine signal that is still more than 2 octave away from the measured resonance point), so I believe it is NOT the interwinding capacitance that has the most effect on raster ringing. Also this is supported by the fact that the 909 uses extremely high inductance coils, where the resonant frequency surely can't be significantly higher than say the Marquee's coil (which has 10 times smaller inductance on the other hand).
So this is where I strated to focus on deflection circuits.

The retrace time is simply given by the time constant of the deflection coils and the installed resonant capacitors. The minimum retrace time is limited by the blocking voltage of the drive transistors on the first hand, as the retrace time gets shorter the voltage peak gets larger for the same inductance. Generally for a given time constant the lower inductance you have the lower the voltage peak will be. This was a crucial point in the G90 deflection system as it is using bipolar transistors which only rated at 1500V, on the other hand the Barco 909 deflection circuit was designed to stand 3000V, the Marquee and the NEC XG was designed at 1800V (in Marquee Ultra pushed up to 2000V). the first idea was to use more up to date MOSFETS in the HDM so it would allow higher blocking voltage and decrease the retrace time with the hope that the ringing will settle down still in the blanking time not affecting the screen content. For this an old modification was dusted off , Dennis et. al. experimented with CREE SiC-MOSFETS to decrease heat dissipation. They were only focusing on the heat dissipation, but in the meantime the blocking voltage was also increased to 2400V with the FETs they used, but they forgot to match the resonant capacitors when they switched to two FETs instead of three, so they actually ran higher retrace times than the standard. With a friend we made the experiment focusing on the lowest possible retrace time, but it was a failure basically because the ringing only got worse.

Most CRT maker used MOSFETS in the deflection outputs which have built in protector diodes against the negative pulse that would occur during retrace. The Marquee uses 3 lower voltage FETs in series, and only relying on the built in damper diodes (ie. 3 diodes in series), regardless what the datasheets say the FET's built in diodes are usualy not very high quality, rather slow to turn on, high forward voltage. For some reason all the other Manufacturers added exernal diodes for suppression (which are by chance always faster than the built in diodes of the FETs), and also usualy they used much more FETs (NEC XG: 2x5 FET in series, Barco 3x3 FET in series, consider their built in diodes as well!), yet in the Barco there are I think 5 additional damper diodes. The G90 deflection circuit strongly relies on the good old 1250/70/90 series, only small modifications were made to it, it uses two parallel bipolar transistors for the Horizontal output, and there are a total of 7 (!) damper diodes added of two kinds: one high current, and other six lower current (but even faster) in 2x3 series-parallel configuration.

Now I'm currently in the progress of continuing the HDM experiments on my own machine. Very recently I also did the CREE MOSFET modification but I've selected the MOSFETS to have indeed the lowest disspiation (guranteed by their low on resistance) and the blocking voltage is not higher than the original 1800V, and I also kept the retrace times at their original value by modifying the original resonant caps. But I also added external (very fast) diodes of different types total of 6. Simulations showed to me that the inductance (ESL) of the capacitor's (legs) are playing an important role in ringing. Needless to say that Electrohome was the only manufacturer who intentionally put ferrite beads on the legs of the resonant capacitors. So I removed these as well. The result so far is that I have a very cool running HDM which actually produce a noticeably more ringing than the original HDM. But, hey I've made difference to the ringing performance of the Marquee without touching the deflection coils, I just need to go to the other direction Smile

Seriously, my resonant capacitor setup is still not optimal for lowest inductance at the moment, so I certainly want to play with that further, also one clue could be the new FETs' lower channel resistance, causing worse Q factor in the HDM perhaps? (however the channel resistance difference seem insignificant compared to the 12 Ohm series resistor in the -B supply line).

_________________
projectors in the past : NEC 6-9PG xtra, Electrohome Marquee 6-7500, NEC XG 1351 LC ( with super modified Electrohome VNB neckboard !!!)
current: VDC Marquee 9500LC
The MOD: VNB-DB, VIM-DB
Back to top
View user's photo album (1 photos)
AnalogRocks
Forum Moderator



Joined: 08 Mar 2006
Posts: 26690
Location: Toronto, Ontario, Canada

TV/Projector: Sony 1252Q, AMPRO 4000G


PostLink    Posted: Sat Sep 16, 2017 7:45 pm    Post subject: Reply with quote

Nice reverse engineering there Thumbs Up
_________________
Tech support for nothing

CRT.

HD done right!
Back to top
View user's photo album (27 photos)
gjaky




Joined: 05 Jun 2010
Posts: 2789
Location: Budapest, Hungary


PostLink    Posted: Sat Sep 16, 2017 7:59 pm    Post subject: Reply with quote

Thanks! Smile
Here I share a comparisonal table (backed up with calculations) for retrace times as well.
Two things are interesting in that: one the calculations for the Marquee does not match with the advertised data (but I know the advertised data is correct -verified), that could be because of wrong measured yoke inductance most likely? Also the Barco 909 coil inductance does not seem to match so I would be happy to see others' readings on coil inductances as well...



General-H-DEF.xlsx
 Description:

Download
 Filename:  General-H-DEF.xlsx
 Filesize:  18.33 KB
 Downloaded:  360 Time(s)


_________________
projectors in the past : NEC 6-9PG xtra, Electrohome Marquee 6-7500, NEC XG 1351 LC ( with super modified Electrohome VNB neckboard !!!)
current: VDC Marquee 9500LC
The MOD: VNB-DB, VIM-DB
Back to top
View user's photo album (1 photos)
redfox001




Joined: 16 Mar 2009
Posts: 2251
Location: The Netherlands


PostLink    Posted: Wed Sep 20, 2017 12:48 pm    Post subject: Reply with quote

Can we measure these coils with a simple peakatlas inductance meter? My 909 has some ringing too but I chose to get used to it Smile However it would be nice if we could use all phosphor and have little ringing to blank.
_________________
701s->runco933->8500ultra->hd1->hd350->vw100->cinemax+919sp+3x919+9500mp->cinemax+919sp(modded)+kuro600a
Back to top
gjaky




Joined: 05 Jun 2010
Posts: 2789
Location: Budapest, Hungary


PostLink    Posted: Wed Sep 20, 2017 6:33 pm    Post subject: Reply with quote

redfox001 wrote:
Can we measure these coils with a simple peakatlas inductance meter? My 909 has some ringing too but I chose to get used to it Smile However it would be nice if we could use all phosphor and have little ringing to blank.


Yes that meter would be appropriate to measure the coil inductance. I'd be glad if you'd take the effort and measure the Horizontal winding's inductance.
You have/had Marquee and 909 at the same time does the ringing appears to be the same on both machine?

_________________
projectors in the past : NEC 6-9PG xtra, Electrohome Marquee 6-7500, NEC XG 1351 LC ( with super modified Electrohome VNB neckboard !!!)
current: VDC Marquee 9500LC
The MOD: VNB-DB, VIM-DB
Back to top
View user's photo album (1 photos)
cmjohnson




Joined: 03 Apr 2006
Posts: 5180
Location: Buried under G90s


PostLink    Posted: Wed Sep 20, 2017 8:39 pm    Post subject: Reply with quote

I made up a test yoke with a pair of variable inductors in series with each coil and adjusted them to see if there was any change of
the ringing that was visible in my test pattern.

I did not find that it made any visible difference.
Back to top
gjaky




Joined: 05 Jun 2010
Posts: 2789
Location: Budapest, Hungary


PostLink    Posted: Wed Sep 20, 2017 9:05 pm    Post subject: Reply with quote

cmjohnson wrote:
I made up a test yoke with a pair of variable inductors in series with each coil and adjusted them to see if there was any change of
the ringing that was visible in my test pattern.

I did not find that it made any visible difference.


I guess it depends on how much you alter the inductance. The retrace time clearly have an impact on the ringing as changing between short and long retrace in the Marquee show quite dramatic difference.
Also as you tune the inductance only, the retrace time changes accordingly.
The retrace time is given by: tr=pi*sqrt(L*C), where L is the total inductance of all coils, and C is the resonant capacitance.

A more precise experiment would be to have considerably different coil inductances, and tune the resonant capacitance so that all system would have the same retrace time (assuming we have no limitation for the flyback pulse magnitude) and only then compare the ringing behaviour.

But as I said earlier simulations point out that lead inductance of the resonant capacitors are also affect ringing, and the less capacitor you use the more their the inductance will be, but this is exactly how the shorter retrace time is achieved in our projectors.

_________________
projectors in the past : NEC 6-9PG xtra, Electrohome Marquee 6-7500, NEC XG 1351 LC ( with super modified Electrohome VNB neckboard !!!)
current: VDC Marquee 9500LC
The MOD: VNB-DB, VIM-DB
Back to top
View user's photo album (1 photos)
Curt Palme
CRT Tech



Joined: 08 Mar 2006
Posts: 24296
Location: Langley, BC

TV/Projector: All of them!


PostLink    Posted: Wed Sep 20, 2017 9:38 pm    Post subject: Reply with quote

I really don't think the issue is yoke related. After all, 1000s of Marquees were made, with no issues. I'd swap the CLM or deflection sub board.
Back to top
redfox001




Joined: 16 Mar 2009
Posts: 2251
Location: The Netherlands


PostLink    Posted: Thu Sep 21, 2017 11:59 am    Post subject: Reply with quote

gjaky wrote:
redfox001 wrote:
Can we measure these coils with a simple peakatlas inductance meter? My 909 has some ringing too but I chose to get used to it Smile However it would be nice if we could use all phosphor and have little ringing to blank.


Yes that meter would be appropriate to measure the coil inductance. I'd be glad if you'd take the effort and measure the Horizontal winding's inductance.
You have/had Marquee and 909 at the same time does the ringing appears to be the same on both machine?


On my present setup with the 909 I have standard 1080p@60 timings from my oppo and see a small band from say 10cm with some small waves in it (220cm total screen width). On my 9500 that I used to have I remember a larger area with somewhat bigger waves. But this is from memory and I am not saying the 909 solved the problem.

I try to do some measurements. The girlfriend takes a lot of my time lol Smile

_________________
701s->runco933->8500ultra->hd1->hd350->vw100->cinemax+919sp+3x919+9500mp->cinemax+919sp(modded)+kuro600a
Back to top
redfox001




Joined: 16 Mar 2009
Posts: 2251
Location: The Netherlands


PostLink    Posted: Thu Sep 21, 2017 12:17 pm    Post subject: Reply with quote

I measured horizontal deflection 130,7uH +- 0,1uH
Barco 909.

Hmmm seems to be at least 10 times smaller than advertised.

_________________
701s->runco933->8500ultra->hd1->hd350->vw100->cinemax+919sp+3x919+9500mp->cinemax+919sp(modded)+kuro600a
Back to top
gjaky




Joined: 05 Jun 2010
Posts: 2789
Location: Budapest, Hungary


PostLink    Posted: Thu Sep 21, 2017 7:26 pm    Post subject: Reply with quote

redfox001 wrote:
I measured horizontal deflection 130,7uH +- 0,1uH
Barco 909.

Hmmm seems to be at least 10 times smaller than advertised.


Indeed it seems to be a bit low. Entering this value into the xls sheet gives 1.1us retrace time, while the advertised is 1.8us. BTW calculating backwards based on the retrace time the coil inductance should be around 400uH, so this thing is fishy either way Smile

_________________
projectors in the past : NEC 6-9PG xtra, Electrohome Marquee 6-7500, NEC XG 1351 LC ( with super modified Electrohome VNB neckboard !!!)
current: VDC Marquee 9500LC
The MOD: VNB-DB, VIM-DB
Back to top
View user's photo album (1 photos)
mp20748




Joined: 12 Sep 2006
Posts: 5681
Location: Maryland

TV/Projector: 9500LC Ultra / Super 02 and 03 VIM


PostLink    Posted: Sat Sep 23, 2017 10:21 am    Post subject: Reply with quote

I think I've figured out how to tackle this, but would want to try it out before mentioning any details.

After talking the Craig, and giving thought to what happens when you change the timings, I no longer see the timings changing as the best way to get rid of the ringing. I instead now think there is some negatives in changing the timings for this. I'll need to get on the test bed unit to confirm my theory, but I'm sure I can tackle this.
Back to top
gjaky




Joined: 05 Jun 2010
Posts: 2789
Location: Budapest, Hungary


PostLink    Posted: Sat Sep 23, 2017 11:33 am    Post subject: Reply with quote

mp20748 wrote:
I think I've figured out how to tackle this, but would want to try it out before mentioning any details.

After talking the Craig, and giving thought to what happens when you change the timings, I no longer see the timings changing as the best way to get rid of the ringing. I instead now think there is some negatives in changing the timings for this. I'll need to get on the test bed unit to confirm my theory, but I'm sure I can tackle this.


Sounds very mysterious, perhaps you can tell a bit more details, even if it is not tested yet..

_________________
projectors in the past : NEC 6-9PG xtra, Electrohome Marquee 6-7500, NEC XG 1351 LC ( with super modified Electrohome VNB neckboard !!!)
current: VDC Marquee 9500LC
The MOD: VNB-DB, VIM-DB
Back to top
View user's photo album (1 photos)
mp20748




Joined: 12 Sep 2006
Posts: 5681
Location: Maryland

TV/Projector: 9500LC Ultra / Super 02 and 03 VIM


PostLink    Posted: Sat Sep 23, 2017 12:00 pm    Post subject: Reply with quote

gjaky wrote:


Sounds very mysterious, perhaps you can tell a bit more details, even if it is not tested yet..


It should... and I will go into some detail, but later. There are a few things I need to look at first, but for now, it's very important to first isolate where it's really coming from on this particular projector model. I have been thinking about and looking at some of what Scott had said about it, and that leads me to pick up kinda where he left off.

But I promise I will share what I find out, as there are some other things I will also be sharing with you in due time..Wink
Back to top
lydmann




Joined: 12 Aug 2006
Posts: 56
Location: Norway


PostLink    Posted: Thu Nov 16, 2017 7:07 pm    Post subject: Reply with quote

Any progress?
Back to top
gjaky




Joined: 05 Jun 2010
Posts: 2789
Location: Budapest, Hungary


PostLink    Posted: Thu Nov 16, 2017 10:01 pm    Post subject: Reply with quote

lydmann wrote:
Any progress?

From my side, little. Until recently I only had one deflection board on wich I worked and used, now I have a backup board as well, and that could serve as a reference too.
BTW I fired up the modified HDM last week, and it wasn't that bad actually as I remembered, so a direct comparison for sure will be handy.
I also got some parts for improving the experimental board, I just did not have time to play with that.

_________________
projectors in the past : NEC 6-9PG xtra, Electrohome Marquee 6-7500, NEC XG 1351 LC ( with super modified Electrohome VNB neckboard !!!)
current: VDC Marquee 9500LC
The MOD: VNB-DB, VIM-DB
Back to top
View user's photo album (1 photos)
mp20748




Joined: 12 Sep 2006
Posts: 5681
Location: Maryland

TV/Projector: 9500LC Ultra / Super 02 and 03 VIM


PostLink    Posted: Fri Nov 17, 2017 11:05 pm    Post subject: Reply with quote

I'll also be getting back to this dreadful problem on the Marquee, and it should be soon.

I have my shop better setup for this, juts need to move the test unit there and make this happen.
Back to top
dvh99




Joined: 25 Dec 2009
Posts: 2158
Location: nederland


PostLink    Posted: Sat Nov 18, 2017 3:47 am    Post subject: Re: Getting rid of raster ringing on Marquees Reply with quote

gjaky wrote:
A few weeks ago someone was interested in my plans in raster ringing topic, and I've made a summary for him. I repost this information here with an intent that it could be useful for others as well, and this could be a base of an open source brainstorming, if you will.

I've spent a lot of time with studying basically all mainstream CRT projector's deflection circuits and tried to gather as many information about them as I could, even though I haven't seen every of these in action. Amongst all CRT projectors the Marquee have the worst reputation in raster ringing. While Sony G90 is thought one of the best and also there are also little complain about the Barco 909. The most accepted reason for ringing is the high interwinding capacitance in the deflection coil. Others discovered that the G90 has a significantly lower inductance deflection coils (~85uH / winding, which are always in parallel) than the Marquee (~270uH / winding, which are in parallel above 60kHz). Lower inductance would suggest lower interwinding capacitance, clear case... BUT, as I've mentioned there are little complaints about the Barco 909's raster ringing, and the 909 have an 1,8us standard retrace time, which is about the same as with the Marquee in short retrace above 60kHz. The funny thing is that the 909 have a (forum) reported coil inductance of 1,5mH (not switched) that is 10x of the Marquee coil, so one must think about its interwinding capacitance as well...

When my interest started in deflection circuits I only had a NEC XG135LC and a NEC 9PG xtra. The deflection circuit of the XG is very similar to the Marquee's, ie. there are different retrace time ranges depending on the scan frequency, but with the NEC the retrace time is fully automatically chosen, and unfortunately the lowest retrace time (1,4 us) which would be useful for the standard HD timings, is only selected in the 120-135kHz scan range, which of no use in HT. Anyway with a PC I tried a lot of exotic resolutions on the NEC to try out the different ranges. All-in-all I found the XG's deflection system having also remarkably low ringing. (A side note, I've made a hack to the XG to force the 1,4 us retrace above 60kHz scanrate, so standard HD timings could be also used, and there was no sign of raster ringing at all). I measured the inductance of the XG's deflection coils, it turned out to be 80uH/winding (that is either in series or parallel depending on scanrate, just like with Marquee) this is very similar to the G90's windings, but the NEC used the so called thin-neck tubes so the coils are no use in any 9" machines. I've also tested my NEC 9 PG xtra, that was never meant for short retrace applications as it is rated at 2,5us retrace time, but it also has a low inductance deflection coil (which is not switched over the range) of 190uH, this is very close to the XG's value, when we consider them in series (2x80uH vs. 190uH). I've made a simple setup to determine the resonance frequency of the NEC PG xtra's deflection coil. The result was 1,5MHz, I could not repeat this experiment on the XG for various reasons, but I managed to perform on the Marquee deflection coils and I got a 1,3MHz result, which is lower than the PG's but not significantly different. An 1,8us retrace pulse would equal to 277kHz frequency sine signal that is still more than 2 octave away from the measured resonance point), so I believe it is NOT the interwinding capacitance that has the most effect on raster ringing. Also this is supported by the fact that the 909 uses extremely high inductance coils, where the resonant frequency surely can't be significantly higher than say the Marquee's coil (which has 10 times smaller inductance on the other hand).
So this is where I strated to focus on deflection circuits.

The retrace time is simply given by the time constant of the deflection coils and the installed resonant capacitors. The minimum retrace time is limited by the blocking voltage of the drive transistors on the first hand, as the retrace time gets shorter the voltage peak gets larger for the same inductance. Generally for a given time constant the lower inductance you have the lower the voltage peak will be. This was a crucial point in the G90 deflection system as it is using bipolar transistors which only rated at 1500V, on the other hand the Barco 909 deflection circuit was designed to stand 3000V, the Marquee and the NEC XG was designed at 1800V (in Marquee Ultra pushed up to 2000V). the first idea was to use more up to date MOSFETS in the HDM so it would allow higher blocking voltage and decrease the retrace time with the hope that the ringing will settle down still in the blanking time not affecting the screen content. For this an old modification was dusted off , Dennis et. al. experimented with CREE SiC-MOSFETS to decrease heat dissipation. They were only focusing on the heat dissipation, but in the meantime the blocking voltage was also increased to 2400V with the FETs they used, but they forgot to match the resonant capacitors when they switched to two FETs instead of three, so they actually ran higher retrace times than the standard. With a friend we made the experiment focusing on the lowest possible retrace time, but it was a failure basically because the ringing only got worse.

Most CRT maker used MOSFETS in the deflection outputs which have built in protector diodes against the negative pulse that would occur during retrace. The Marquee uses 3 lower voltage FETs in series, and only relying on the built in damper diodes (ie. 3 diodes in series), regardless what the datasheets say the FET's built in diodes are usualy not very high quality, rather slow to turn on, high forward voltage. For some reason all the other Manufacturers added exernal diodes for suppression (which are by chance always faster than the built in diodes of the FETs), and also usualy they used much more FETs (NEC XG: 2x5 FET in series, Barco 3x3 FET in series, consider their built in diodes as well!), yet in the Barco there are I think 5 additional damper diodes. The G90 deflection circuit strongly relies on the good old 1250/70/90 series, only small modifications were made to it, it uses two parallel bipolar transistors for the Horizontal output, and there are a total of 7 (!) damper diodes added of two kinds: one high current, and other six lower current (but even faster) in 2x3 series-parallel configuration.

Now I'm currently in the progress of continuing the HDM experiments on my own machine. Very recently I also did the CREE MOSFET modification but I've selected the MOSFETS to have indeed the lowest disspiation (guranteed by their low on resistance) and the blocking voltage is not higher than the original 1800V, and I also kept the retrace times at their original value by modifying the original resonant caps. But I also added external (very fast) diodes of different types total of 6. Simulations showed to me that the inductance (ESL) of the capacitor's (legs) are playing an important role in ringing. Needless to say that Electrohome was the only manufacturer who intentionally put ferrite beads on the legs of the resonant capacitors. So I removed these as well. The result so far is that I have a very cool running HDM which actually produce a noticeably more ringing than the original HDM. But, hey I've made difference to the ringing performance of the Marquee without touching the deflection coils, I just need to go to the other direction Smile

Seriously, my resonant capacitor setup is still not optimal for lowest inductance at the moment, so I certainly want to play with that further, also one clue could be the new FETs' lower channel resistance, causing worse Q factor in the HDM perhaps? (however the channel resistance difference seem insignificant compared to the 12 Ohm series resistor in the -B supply line).

_________________
1 answer always poses multiple questions.
marquee 9500ultra HD10L moome hdmi1.3 v3+ some mods.
Back to top
dvh99




Joined: 25 Dec 2009
Posts: 2158
Location: nederland


PostLink    Posted: Sat Nov 18, 2017 3:50 am    Post subject: Reply with quote

scott mentioned retrace timins so i put the caps back.
there is a formula for calculating retrace timings on page 39 or 40 from memory from the same scott aka tse.

the cree fets have been replaced by wolfspeed fets on the farnell site at much lower cost.

_________________
1 answer always poses multiple questions.
marquee 9500ultra HD10L moome hdmi1.3 v3+ some mods.
Back to top
dvh99




Joined: 25 Dec 2009
Posts: 2158
Location: nederland


PostLink    Posted: Sat Nov 18, 2017 3:53 am    Post subject: Reply with quote

scott also said the fets resistance also should improve hor linearity.
_________________
1 answer always poses multiple questions.
marquee 9500ultra HD10L moome hdmi1.3 v3+ some mods.
Back to top
Display posts from previous:   
Post new topic   Reply to topic   Printer-friendly view    CurtPalme.com Forum Index -> CRT Projectors All times are GMT
Goto page 1, 2  Next
Page 1 of 2
Jump to:  
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
You cannot attach files in this forum
You can download files in this forum