|
Vintage Television and Video Vintage television and video equipment, programmes, VCRs etc. |
|
Thread Tools |
7th Apr 2018, 4:44 pm | #61 |
Dekatron
Join Date: Mar 2005
Location: Gateshead, Tyne and Wear, UK.
Posts: 7,444
|
Re: DC Restoration questions
I'm certain without the cathode follower the DC restorer will not operate correctly. The reasoning for this is because the varying beam CRT current of the CRT is influencing the charge on the coupling capacitor. Compared with the small CRT employed in the Admiral TV set, the 17" CRT in the Ultra will be drawing a much higher beam current, say >200uA.
In a sense we can say the steady potential produced by the DC restorer is the result of pump charging by the sync pulse tips. So it follows the coupling capacitor functions on a short charge time (9uS 405 lines) and must maintain it's charge until the next sync pulse 80uS later. The varying CRT beam current will mess things up. I'm thinking about sync cancelled vision AGC for the Ultra V17-53. Only a few extra components are required. The Philips 1458U employed one half of an EB91 double-diode, for the Ultra any general purpose semiconductor diode can be used. DFWB. |
8th Apr 2018, 12:25 am | #62 |
Octode
Join Date: Apr 2016
Location: Ventnor, Isle of Wight, & Great Dunmow, Essex, UK.
Posts: 1,377
|
Re: DC Restoration questions
Hi Hugo,
Here is the circuit arrangement I used. It's a simplified version of your Admiral one. With a normal picture displayed, there is no clamping action at all. Only once the brightness is reduced to cut off does it start to clamp. I tried it with two 1N4148's in series, but the results were the same. I presume, as David said, the beam current is upsetting things? This set has a massive 21" tube!! David, I knocked up your circuit (including the cathode follower) with the bits just in free space hanging off the back of the IF board. I have to say it works very well. When the contrast control is varied, the sync tips stay rock steady. I must find a way of changing the video content without altering the overall signal level. It would make these tests more meaningful. The 0.1 is certainly essential to decouple the grid. You get big vertical hum bars without it!! As we will only be using these sets with a modulator, presumably there will be no need for AGC as the signal level will be constant? I agree though, it would be nice to incorporate just for the satisfaction of doing so! All the best Nick |
8th Apr 2018, 12:58 am | #63 |
No Longer a Member
Join Date: Oct 2016
Location: Maroochydore, Queensland, Australia.
Posts: 2,679
|
Re: DC Restoration questions
Ok, I see what is wrong; for this restorer to work it is important that the Thevenin resistance, or the source resistance that the restorer actually clamps to is low, or all that happens is that its reference level is pulled around with signal. The 250k pot/divider is far too high, so it won't work. I'm also just replying to Davids' post, when the source resistance there is about 15k, there is only about a 3V shift in level from min to max beam current with your CRT.
|
8th Apr 2018, 1:13 am | #64 | |
No Longer a Member
Join Date: Oct 2016
Location: Maroochydore, Queensland, Australia.
Posts: 2,679
|
Re: DC Restoration questions
Quote:
If there is a problem I don't think it is the CRT beam current influencing the coupling capacitor's charge between sync tips when the diode is not conducting because even if the full CRT current charged the capacitor between sync tips it could only absorb about 18nC charge over that time and for a 0.47uF the voltage would only shift by 30mV, or say 60mV for a 0.22uF cap between the very approx 90uS syncs. If the voltage to which the diode clamps to has too high a source resistance, the clamp won't work, as the voltage will be pulled. This is why Nick's circuit didn't work probably. Less than 20k is ok. Then it only pulls the reference voltage around by about 3 volts from near zero to max beam current. I have attached the circuit. One interesting thing, is that from the AC perspective, over the normal range of video frequencies, the cathode impedance is set by the Anode load resistor, which is about the same value you have chosen for your cathode follower too. In any case, I only know for sure that this circuit works in the set with obviously a much lower beam current. So for Nick's set whatever works best is best. |
|
8th Apr 2018, 6:33 am | #65 |
No Longer a Member
Join Date: Oct 2016
Location: Maroochydore, Queensland, Australia.
Posts: 2,679
|
Re: DC Restoration questions
Of course the impedance of a cathode follower output is much lower than the cathode resistor, typically a few hundred ohms, which does help.
|
8th Apr 2018, 9:58 am | #66 |
Octode
Join Date: Apr 2016
Location: Ventnor, Isle of Wight, & Great Dunmow, Essex, UK.
Posts: 1,377
|
Re: DC Restoration questions
Ah, that makes sense! I'd never thought about the source resistance of the reference voltage. I see your Admiral values are 52K in total for the potential divider, rather than my 250K!
Looks like another fun evening of experimentation ahead I'll try a much lower resistance potential divider and see what happens. Many thanks Nick |
8th Apr 2018, 11:26 am | #67 |
No Longer a Member
Join Date: Oct 2016
Location: Maroochydore, Queensland, Australia.
Posts: 2,679
|
Re: DC Restoration questions
Nick,
If you have a resistive divider with two resistors in series across a power supply of some voltage V, and lets say both of the resistors for arguments sake are 20k and you want to know what the center of that divider behaves like when its loaded (you try to push current into it or extract current out of it) then the junction of those resistors acts exactly as though it is a single supply of V/2 with a series 10K resistor, in other words it acts as though the two 20k resistors are in parallel. It is the Thevenin Theorem. So if you look at the the divider chain for the brightness control in the Admiral, then you could work out that the effective resistance on the slider of the pot is roughly in the order of 15k. In your circuit with the 250k pot, the resistance is effectively 125k. Still if you don't have any luck with that circuit, its good that David's one is working fine. Hugo. |
8th Apr 2018, 11:53 am | #68 |
Dekatron
Join Date: Mar 2005
Location: Gateshead, Tyne and Wear, UK.
Posts: 7,444
|
Re: DC Restoration questions
Argus25 wrote:
"Ok, I see what is wrong; for this restorer to work it is important that the Thevenin resistance, or the source resistance that the restorer actually clamps to is low, or all that happens is that its reference level is pulled around with signal." Thevenic resistance defined: http://hyperphysics.phy-astr.gsu.edu.../thevenin.html DFWB. |
9th Apr 2018, 12:33 am | #69 | |
Octode
Join Date: Apr 2016
Location: Ventnor, Isle of Wight, & Great Dunmow, Essex, UK.
Posts: 1,377
|
Re: DC Restoration questions
Quote:
After thinking of various ways I could vary the video content without varying the sync, it suddenly hit me:- use an actual program Talk about missing the obvious! So I did some dynamic tests with an episode of 'The Good life' from DVD. The video at the CRT cathode was viewed on the scope. Firstly with the TV returned to it's original configuration, the sync pulse tips and black level varied wildly with scene content. A good 20V or more of variation which will take this tube from black through to light grey. The black lever changes on the screen were very obvious. Next, David's circuit was reconnected and despite scene changes, the sync tips remained clamped rock steady. But the amplitude of the sync pulse was changing significantly with scene content. Consequently, the black level was still changing although not nearly so much. The logical conclusion was that the AGC was affecting the amplitude of the video appearing at the detector. This was disabled by removing the 820K resistor from the video amp anode. -12V was then fed to the AGC line with a separate power supply. The result was a constant amplitude sync pulse with it's tip firmly clamped. As the amplitude was constant this meant that the black level was also constant. It was nice to watch the video content on the scope dancing about wildly whilst the sync pulse & black level stayed rock steady. There is now so much stuff hanging out the back of the TV, including the LOPT (and cooling fan) that I can't turn it round to have a proper look at the picture. Viewed in the mirror though it looks really good! During my tests, I noticed that the cathode follower seems to improve the picture so even on a static test card it looked better that the original set up. I still wish to do some further experiments with Hugo's circuit using a much lower resistance potential divider just to see if I can get it to work. All the best Nick |
|
9th Apr 2018, 6:32 am | #70 | |
No Longer a Member
Join Date: Oct 2016
Location: Maroochydore, Queensland, Australia.
Posts: 2,679
|
Re: DC Restoration questions
Quote:
It sounds like the circuit you have will be hard to beat, I don't think the one I suggested will work quite as well. If the picture is improved with the cathode follower it must be because it has reduced the capacitive loading at the video output valve's plate circuit, so then there are two reasons to keep the follower. Hugo. |
|
10th Apr 2018, 8:52 pm | #71 |
Dekatron
Join Date: Mar 2005
Location: Gateshead, Tyne and Wear, UK.
Posts: 7,444
|
Re: DC Restoration questions
1100Man wrote: "Next, David's circuit was reconnected and despite scene changes, the sync tips remained clamped rock steady. But the amplitude of the sync pulse was changing significantly with scene content. Consequently, the black level was still changing although not nearly so much."
Hi Nick, I've experienced a similar effect in the Ultra V17-53, sure enough the sync tips remained rock steady irrespective of picture content, but there is certainly slight black level shifts on scene changes. However, the fault is not caused by the performance of the black level clamp circuit. Connecting the scope to the slider of the high level contrast control reveals that the black level shifts are a fault in the post demodulator circuits. In the PT650 "Spacemaker" and Ultra V17-63 there is not a direct connection between the vision demodulator and the grid of the 30FL1 video amplifier valve. In both receivers a 1microfarad capacitor is interposed between the demodulator and video amplifier valve, a 68Kohm resistor is connected across the capacitor so in fact we have partial AC coupling which in my opinion causes more problems than a complete loss of video DC component. Nevertheless, the pictures displayed by the modified Ultra V17-63 are far better than having AC coupled video to the cathode of the CRT. DFWB. |
10th Apr 2018, 9:55 pm | #72 |
Octode
Join Date: Apr 2016
Location: Ventnor, Isle of Wight, & Great Dunmow, Essex, UK.
Posts: 1,377
|
Re: DC Restoration questions
Hi David,
Examining the video coming from the demodulator showed that the amplitude of the sync pulses (and the whole waveform) was changing with scene content. Not by a huge amount, but enough to cause the black level to change once the sync tips had been clamped. As the set uses the average video level to control it's AGC line, I presumed that it was the AGC that was causing the amplitude changes. As a test, I disabled the AGC and fed it from the bench PSU. This resulted in constant amplitude video and therefore the black level remained rock steady:- the clamp does work very well! I guess mean level AGC, for a locally generated signal of constant amplitude, is actually counter productive. As a further experiment, I used the brightness control as the reference voltage for the clamp and grounded the CRT grid. The cathode follower was still in circuit. This seemed to work just as well. I then removed the cathode follower to see what would happen. The clamp still worked pretty well actually, but was not quite as good so it's better with the follower. So I'm just going to knock up the final version of the circuit and get rid of the mass of stuff hanging off the back of the board used for testing. Cheers Nick |
10th Apr 2018, 11:57 pm | #73 |
Octode
Join Date: Apr 2016
Location: Ventnor, Isle of Wight, & Great Dunmow, Essex, UK.
Posts: 1,377
|
Re: DC Restoration questions
So the circuit is now in it's final configuration and works very well. One further advantage of having the cathode follower is that it has eliminated the bright spot on the screen at switch off.
All in all, this has been an extremely interesting and useful exercise and I have learnt a lot from it. I will certainly use these mods in other sets as well as I think it's a very worthwhile improvement. Strangely enough, in my experiments last night, I re tried Hugo's circuit which I had not been able to get to work originally. It worked fine- even without the cathode follower, so I'm not sure what was going on there. It was probably me:- it usually is!! Many thanks to David and Hugo for taking so much time and trouble to come up with designs and answer my questions! All the best Nick |
11th Apr 2018, 3:38 pm | #74 |
Dekatron
Join Date: Mar 2005
Location: Gateshead, Tyne and Wear, UK.
Posts: 7,444
|
Re: DC Restoration questions
Hi Nick,
before we sign off these jobs completely the 1uF coupling capacitor between the vision demodulator and grid of the video amplifier valve should be shorted out, it's C35 in the V17-53 and V17-70. Also, short circuit the 0.04uF capacitor connected between the slider of the interference limiter and chassis, it's marked as C48 in the V17-70 and C34 in the V17-53. These measures will ensure that there is a direct connection between the demodulator and the grid of the video amplifier. Along with the DC restorer circuit and the modifications discussed in this post the results should be a near perfect black level irrespective of picture content. DFWB. |
16th Apr 2018, 10:01 pm | #75 | |
Octode
Join Date: Apr 2016
Location: Ventnor, Isle of Wight, & Great Dunmow, Essex, UK.
Posts: 1,377
|
Re: DC Restoration questions
Quote:
Shorting out the 1mfd, causes the video amp to clip on peak whites unless the signal level is reduced to a lower level. Monitoring the black level on the scope doesn't show any advantage in doing this. As I've said before, the biggest gain has been to disable the AGC. That really does ensure that the signal level stays constant and therefore so does the black level. I'm still not sure what the best way to do this is as there are shortcomings with my simplistic method of powering the AGC line from the bench PSU. It needs -10.86V and even 0.5V causes a big change in signal level. I managed to turn the TV round and watched an hour's worth of programmes. The black level performance was outstanding! But during the hour as the set got warmer, the signal level gradually increased. Also the fine tuner has a big effect on signal level. It does need some sort of AGC and I guess that would be the benefit of 'gated AGC'. The picture content would be ignored, only the sync pulse level would be monitored. You did suggest that earlier in this thread! So maybe, after all, that would be another good modification? All the best Nick |
|
17th Apr 2018, 12:30 pm | #76 |
Dekatron
Join Date: Mar 2005
Location: Gateshead, Tyne and Wear, UK.
Posts: 7,444
|
Re: DC Restoration questions
1100man wrote: "Shorting out the 1mfd, causes the video amp to clip on peak whites unless the signal level is reduced to a lower level. Monitoring the black level on the scope doesn't show any advantage in doing this."
Hi Nick, Along with C35 (1mfd) did you also short out C48 (0.04mfd)? Alternatively, you can adjust the interference limiter control so that the slider is set to chassis. DFWB. Last edited by AC/HL; 24th Apr 2018 at 7:18 pm. Reason: Thread split |
24th Apr 2018, 12:45 am | #77 | |
Hexode
Join Date: Oct 2017
Location: Manchester, UK.
Posts: 361
|
Re: DC Restoration questions
Quote:
|
|
24th Apr 2018, 3:35 am | #78 |
No Longer a Member
Join Date: Oct 2016
Location: Maroochydore, Queensland, Australia.
Posts: 2,679
|
Re: DC Restoration questions
Radio graham,
Unfortunately the polarity of a video signal drive to the CRT, either to the grid, or its cathode is opposite to what is required for DC restoration by the CRT gun itself. For example, to get the CRT grid to draw current and alter the charge on a coupling capacitor, it has to be taken positive with respect to the cathode, but the video drive at the grid has a negative going sync tip. High peak white levels might cause some grid-cathode conduction with high contrast signals and the brightness pot in some position, but that is not DC restoration, and probably not healthy for the CRT either. Also the CRT gun is a fairly resistive diode, unlike a good DC restorer diode like a silicon diode or a 6AL5. The grid of a CRT should at all times be run zero to negative with respect to the cathode anyway. I had read that myth about the CRT doing DC restoration too, but I never believed it. |
24th Apr 2018, 10:30 am | #79 |
Dekatron
Join Date: Mar 2005
Location: Gateshead, Tyne and Wear, UK.
Posts: 7,444
|
Re: DC Restoration questions
One of the concerns about the addition of the cathode follower circuit is the possibility of failure of the valve or even a bad connection somewhere in the circuit. These conditions will cause excessive CRT beam current.
The attached circuit diagram shows a circuit which will address this problem. Under normal conditions the diode will be conducting and it can be considered there is a direct connection between the CF load resistor and the cathode of the CRT. The 470Kohm resistor will have no influence on the operation of the cathode follower. In the event of the voltage at the anode of the diode going low, no current will pass through the diode and the CRT beam current will flow through the 470K resistor. A voltage will be developed across the resistor and this will reduce the beam current. If the CRT beam current is, say 100microamperes then the voltage across the resistor will be 47V, still too low for the diode to conduct. If however, the beam current rose to >200microamperes the diode will cease conducting and the CRT beam current will flow through the resistor. DFWB. |
24th Apr 2018, 10:36 am | #80 |
Nonode
Join Date: Jan 2004
Location: 1966-1976 Coverack in Cornwall and Helston Cornwall. 1976-present Bristol/Bath area.
Posts: 2,965
|
Re: DC Restoration questions
It might be worth taking a look at the circuitry used by black and white single standard 625 hybrid ITT/KB VC200 chassis.
It has a good DC black level, uses Line gated AGC and is the only telly that I know of to use CRT grid modulation. It works well and the CRT's live long healthy lives.
__________________
Simon BVWS member |