I thought you might update the spreadsheet calculations or the signature png with 0.55V or 0.6V for the diodes since 0.7V is provably too high and arbitrarily using that in circuit design is a beginner level mistake. Instead you want to defend against any changes.
I didn't originally pay attention to the 100nF / 0.1uF replacement caps on the famous diagram spread round the web. Pulling up jungle chip datasheets, I noticed the LA7615 discussed here recommends 1uF. Earlier discussion of JVC C-13011's TA1242N shows 0.47uF even though the Japanese TA1242N datasheet has 0.1uF.
Would be helpful to note not to change the existing capacitor value where it is known, unless Voultar's page 1 suggestion is all the proof needed. A series capacitor is a high pass filter and the higher the value, the higher the time constant but less distortion of the RGB square waves.
Second thing I overlooked was the inevitable horizontal shift that 6tanks on Reddit
explained nicely. So the shift is not necessarily fully correctable and I don't suppose all CRTs with an OSD have an option for horizontal shift. My 10 second thought is using an all pass filter that can be adjustable with a potentiometer to add enough group delay to undo the shift.
MarkOZLAD wrote:
The mux spreadsheet also tells you an exact resistor calculation, the "standard size" calculation is only there for convenience. You choose whatever resistors are available to you. In my experience the ones that will be available will be the ones in the list...
An electronics novice is going to use the values your png says to use, or what the spreadsheet says to use. The recommendation function took more Excel skills than I have, impressive that it's there and helps people. I don't understand leaving out context that tighter resistor values may exist but that's okay.
Now that I think about it, I suppose rounding up versus down or to the nearest value is a mistake. Always rounding up to a higher value that increases the output voltage above the rated voltage via voltage division would seem worse.
MarkOZLAD wrote:
NewSchoolBoxer wrote:
Exceed the rated 0.7V by 7-10%, I think that's really pushing it but I never tried overvolting a CRT.
You're not overvolting the CRT, all you are doing is feeding the jungle chip RGB over 0.7V. If you look at jungle datasheets you'll find they handle way over 0.7V on their RGB inputs. In fact many OSD's voltages are way over the 0.7V which actually calls into question the need to match the 0.7V.
5% Resistors are also what you will find in the sets from the factory....
I see the pin diagrams on most jungle chips have internal diode clamping so that's super nice and not what I anticipated:
If every jungle chip in a moddable CRT has diode clamping on every pin, that doesn't prove exceeding the rated voltage by any amount is safe for any/all chips. Diodes don't switch instantaneously and may not be working after 10k hours of use either. But sure, if 0.7V is expected and the datasheet shows 0.5-1.0V is within spec then you're right for that chip.
First datasheet I found showing resistor tolerances without forcing me to look up the part numbers is the PVM 20M7MDE:
Most but not all values are indeed 5% and is a luxury device made for video professionals. It's also a device with 500 or so resistors where the cost increase for 1% would take justification. As would using ceramics for 1uF and 4.7uF but I see only electrolytics. The L2 does have a few 1uF ceramics.
On the other extreme, I saw Nintendo used 5% resistors on their JP-21 RGB cable for 75 ohm termination. I think a premium SCART cable doing that today would face real criticism.
I guess I'm saying what I said before, the spreadsheet should recommend 1% but say 5% is acceptable. Diode voltage drop is wrong, horizontal shift is a design flaw but my suggestions are opinions and debatable.
