AVS Forum banner

Status
Not open for further replies.
1 - 14 of 14 Posts

·
Registered
Joined
·
4,815 Posts
Discussion Starter · #1 ·
Well at least I hope this turns into that anyway. I gleened every scrap of information about the Marquee LVPS that I could from this board; Read through the torturous Marq. Maint. thread, copied and pasted all the posts and pics related to the LVPS, printed them out and sat down to discuss them with an Electronics engineer. My friend Dana is very sharp and and he designs and builds custom circuits for scientific instruments for a living.

The conclusions he came to kinda surprised me. The common approach to the Marquee P14 voltage drift seems to be either to

1) check and set P14 using the factory trim pot and then re-check once a year. (this one is very popular because it dosen't cost you anything assuming you already own a good voltage meter)

2) substitute a larger trim-pot (this assumes that the P14 drift is due entirely to the pot "leaking" , but not a conclusion I am convinced is the problem here)

3) insert a resistor before the trim pot to limit the range of adjustment to between 6.15 and 6.35 volts. This is a more complicated version of no.2 and also assumes it's the pot that is at fault here. This also makes the pot almost entirely meaningless since you are in effect locking-down the voltage with resistors and should anything leak downstream from the pot you would not be able to correct it very much.

the conclusion that Dana came to after looking at all this was that both 2 and 3 above were approaching the problem from the wrong side of the circuit. According to him, filamnet voltage drift on ANY vaccum tube is entirely normal. Being a collector of very old tube-radios he explained that they used to come with a knob on the front that the user could set the filamnet voltage every time you listened to the radio. In other words the drift isn't happening because the trim-pot is leaking but because the voltage is changing after the pot due primarily the aging of the tube filaments and resistors downstream. basically, according to him neither of the above accepted mods will garuntee your P14 voltage will stay put after you mod the circuit with the methods that are generally accpepted here.

This assertion is supported by the fact that the newer Marquee Ultras actually have a complete new circuit-board piggy-backed onto the LVPS that will adjust P14 up OR down depending on what kind of changes are happening to the circuit downstream from the trimmer, kinda like a brain monitoring the P14. Is this the only sure-fire way to really fix this issue. If I employed his help in mimicking this approach would it be possible to bring old Marquee LVPS's up to the standard of the newer Ultra's, maybe sell the mod as a kit?

I have to believe that if fixing this was as simple as just sticking a bigger pot in the current manufacturer would not have to chosen to drastically change the enire LVPS?

I'm really curious to hear from all the experts, Tim, MP, Newman, etc. what's the real story behind this, how do other brands line Sony or NEC handle this problem, it must be a common issue for any CRT machine, right?:confused:
 

·
Registered
Joined
·
384 Posts
So what is the new circuit board doing? Is it just a voltage regulator? (Is the LVPS an AC or DC supply?) Sounds unlikely, given the above description.


How about a current regulator? We do that sometimes for DC filament supplies in tube preamps and amps, but the current requirements are, I'm guessing, miniscule (maybe 2.5amps, at most) compared to CRT filaments.


Hope the thread continues - Pat
 

·
Registered
Joined
·
2,124 Posts
Dragman

I can't say that I know the root cause of the problem but the solution as described by "dana" for tube radios is easily acheivable with the mod I suggested using a larger pot adj from the front of the PJ. Just monitor (occasionally) P14 voltage and adj as neccessary.

Of the three PJs I have modded this way 2 have been going for over a year and have required no tweaking. The third I just did.

I am not going to argue that this is the best solution but IMO is the cheapest, easiest effective solution that I have seen. I think that is what most people are looking for and I'd bet that if you posted a schematic of the new piggyback board very few members if any would build or buy it.

Of course I'm not 100% sure about the solution I am using (which by the way was suggested to me by a VERY prominent reseller whom does the same mod himself) But again it has been over a year on 2 PJs which were drifting and now are not.

AGAIN. If I'm wrong and flirting with disaster let me know.!!!!

Please see attached 3 pics (1 per post since i don't know how to attach multiples) of the mod.
 

·
Registered
Joined
·
2,124 Posts
3rd pic this time of the front with the cover on. This image shows how important it is to ensure you have enough clearence between the pot and the cover.

The pot can be adj with the cover on if you are carefull yet is fairly safe from accidental adj where it is.
 

·
Registered
Joined
·
4,815 Posts
Discussion Starter · #6 ·
tubesguy, I think this is all DC, the specified value to the filaments is 6.35 but these machines are known to drift to 7 or higher which the experts agree will fry the filaments after a few hundred hours. I wish I knew more about how the new LVPS works but I only saw a pic of it on this board once. I was hoping MNewman would elaborate on this last part as he seems to know the most about the Ultras. Here is a quote from the Marquee Maintenance thread.
Quote:
I just looked at the heater circuit on the LVPS, it is just using a lm117 voltage regulator http://www.qsl.net/ok1duo/lm317.htm, cool webpage to do the math . The circuit in the LVPS has a 120 ohm as R1 and a 1k pot 0-1K for R2, that gives us 1.25v - 11.67V.
Energeezer, I do want to thank both you and Sly for the pics and the work you guys have put into this. The fact that your machines are stable after a year seems to be a testament to itself. My friend and I also agreed that being able to adjust P14 on the fly is a very cool solution with the panel mount trimmer. The question that's bugging me is why did the factory feel the need to re-design so much of this circuit if all that is needed is a bigger trim-pot? Is it just engineering overkill?
 

·
Registered
Joined
·
2,124 Posts
Quote:
The question that's bugging me is why did the factory feel the need to re-design so much of this circuit if all that is needed is a bigger trim-pot? Is it just engineering overkill?
Thanks

I'd be interested to know myself.
 

·
Registered
Joined
·
344 Posts
I just set mine and tapped on the pot a bunch of times while watching the voltage to see if it would stay put. It did


Update: While that is what I did, I think the better thing to do is to modify the circuit to limit the excessive range of the pot. There is another thread here, not mine, with instructions how to do it. In that thread it is suggested to add a 475 ohm resistor and a 20 ohm pot. I have not tried it but it makes sence to me. Use your own best judgement as to wheather you can do this yourself. I'm a newby at this projection stuff, but I'm not a newby at electronics.

There is no technical merit to the idea that the filaments age and cause the voltage drift.
 

·
Registered
Joined
·
1,056 Posts
There really isn't any mystery here. The output of the LVPS (internally) is designed to be constant (regulated). The filaments of the CRT tubes are designed to operate at a specific voltage and will draw a specific current which does not change. So the filaments themselves do not cause what is being referred to as 'drift'.


If the only thing between the constant supply voltage and the constant current load of the filaments is the adjustment pot, the only thing that can vary the voltage at the filaments is that pot. This is simple Ohm's law.


It's well known that with age some of these LVPS pots change resistance and therefore change the voltage at the filaments. The filaments themselves do not cause the problem. Only the series resistance (the pot) is at issue.


Sly is correct that it's a dangerous situation to have a single, wide-range resistance pot involved. Not so much because of drift, but because of the potential of being able to fault (lower resistance) and allow too much voltage to be applied to thousands of dollars worth of tubes. His solution limits the amount of range the pot has, thus saving the tubes should the pot fail. Assuming the resistor in series with the pot is stable, his circuit limits the amount of voltage range which the pot can control. It's a simple but elegant method of protection and since the filaments don't drift in and of themselves it works just fine.


But a more accurate way to maintain an absolute voltage is a regulator circuit. This does not compensate for downstream drift, it produces a fixed unchangeable output voltage that maintains its integrity regardless of what is going on downstream (within its design limits). A good quality regulator in production quantities also costs far less than a pot so it's a good choice for multiple reasons.


In practice, as a repair for an existing non-regulated circuit, a better quality pot is a perfectly acceptable solution. Even though the potential (however slim) for a future failure causing damage to the tubes still exists, it's highly unlikely. Also possible but unlikely, a regulator could fail and short causing the same problem.


--Bill
 

·
Registered
Joined
·
4,815 Posts
Discussion Starter · #10 ·
arent the filaments shedding off electrons? Don't they consume themselves as they live out their rated lifespans and actually get thinner the longer they burn? therefore their resitance value changes with age, right?

at least this was how it was explained to me.

My friend did agree that Sly's tweak is an old and accepted method way of limiting voltage. The overly large range of adjustment in this circuit was probably a manufacturing oversight. But dosen't locking it in to such a narrow range make the trim pot serve almost no purpose at all. It assumes that both the resistor and everything downstream of the trimmer will stay rock-stable.

BTW, do you really beleive that the additional circuit added to the new Marquee Ultras is cheaper than a $3. trim pot? Seems rather unlikely to a layman like me but?
 

·
Registered
Joined
·
1,056 Posts
It's probably less than $1. Generally, it's much cheaper and less error prone to use a semiconductor, even with a few outboard components, than it is to use a mechanical device.


No, the filaments are not shedding off electrons in any significant quantity. The filament is just heating the cathode which is actually the source of emission for beam electrons. The cathode does degrade with age and heat and can cause various degrees of beam anomolies depending on its condition.


The spec'd filament operating voltage of these tubes (6.35 volts DC) does not change with age.


There were some tubes in early radios (pre-1940) and others in the 50s and 60s that were battery powered and did have the filaments that were treated to also act as cathodes. Some rectifier tubes (low and high voltage) had the cathode integral to the filament (as a coating) so there is some basic truth to your friend's statement in the right context. But it doesn't pertain to CRT Projection tubes.


--Bill
 

·
Registered
Joined
·
827 Posts
the ultra supplies have small daughterboard instead of the pot

to regulate the 6,35 V. It is attached to some reference voltage

inside the LVPS so I assume that is truly fixing the problem.

I personally use a very expensive 1 W pot in exchange of the factory

blue one on the old LVPS and works fine.
 

·
Registered
Joined
·
264 Posts
This is very simple the factory set up is too sensitive way to sensitive.

The pot can be in spec (spec drift over time and temperature) and still let the voltage go to a dangerous setting.

I have not seen the PCB fix but my guess is they just fixed the regulator like I did and put it on a small PCB. That is the cheapest way to fix the problem with out spinning the mane PCB of the power-supply.


Their is some need for some small adjustment so a pot is still needed. There is some voltage drop from the v-reg to the filament and the small adjustment is needed to correct for this. Their should never be a problem after the mod unless something falls. But before the voltage could get way out of spec with nothing failing. It was just wrong before. Some one messed up on that part and it was not noticed until it was too late and the power-supply was in production.



Let me clarify some things.

There seems to be some confusion about what the circuit is.


The circuit IS a REGULATOR and always has been..


I have seen over and over where some one says it should be replaced with a regulated voltage. That is what it is to begin with.. The sensitive pot is by design(Bad design)


If you car has a bent rim changing to a better tire might help but the rim is still bent.

Changing the pot might help but the circuit is still wrong. Fix the circuit and there is no problem.


Ask anyone that has done the fix to take a reading. And I bet it will be right where they set it.


There is no magic here the LM317 is a very common regulator and there is nothing wrong with it. It was just configured wrong that is all. Just the fact that is works as well as it does with a very wrong config show how good the LM317 is.

Some one way way back had a slip-up and it it shipped it happens. It is such a simple circuit no one thought to review it is my guess. They might have drew it on a late Friday and was forgot about on Monday so it never got finished or something like that.



So let have some feedback who did the mod (the 475 ohm resistor and a 20 ohm pot mod)

has it drifted at all? and how long has it been?
 
1 - 14 of 14 Posts
Status
Not open for further replies.
Top