Electronics > Repair
General Radio 1396B Tone Burst Generator - Understanding power supply
(1/4) > >>
barbaroja:
Hi. In the process of recapping a GenRad 1396B. The old caps (around50+ years old) seem to be ok but for one or two that don't.

Leaky caps are

(1) Mallory marked 400 400 25, 4450-5621 85C. Present in main PSU. Has two 400uF caps really.
(1) Sprague 4450-6020, 500MFD-20VDC D73710. Present in scaler PSU.

Those caps are not available anymore, and if found, replacements are very old as to guarantee trouble free operation. Analyzing the schematics, there are some caps before dropping resistors and some more caps of different value.

So here come the questions:

1. The schematics suggest that the circuit sees the whole value of the dual Mallory PSU caps (parallel connection). Wondering why they took that design decision (using dual caps. Maybe was avaibable at the time?), and if it would change performance of the device to use different (bigger) values. Standarized caps today go 100uF, 150uF, 220uF, 330uF, 470uF, 680uF.

2. Could they be using different size caps as to reduce inductive reactance? (Dave has a video on that). I find particularly interesting that both the main PSU side and the scaler PSU side use different values after the dropping resistor.

Circuit diagrams here: https://manuals.repeater-builder.com/te-files/GENRAD/GENRAD%201396B%20Instruction.pdf

There's so much to learn from vintage tech.

amyk:
Those are just regular power supply decoupling. 680 and 820 would be fine for replacement. Keep in mind that the originals would've had 20% tolerance anyway.
barbaroja:

--- Quote from: amyk on November 29, 2024, 02:33:37 am ---Those are just regular power supply decoupling. 680 and 820 would be fine for replacement. Keep in mind that the originals would've had 20% tolerance anyway.

--- End quote ---

Yeh I can imagine that. Any reason to use different values in the original design? (I mean why not use 820 for everything for example)
Just trying to understand it a lil better.
coppercone2:
usually its

1) price
2) size
3) hearsay

The #1 thing you will notice is the change in capacitor film technology, aka improvement. They can make thinner films more reliably that do the job so the caps are smaller, and better electrolyte.

With old machinery you can guarantee the film is durable when its above a certain thickness. With newer processes, the film can be made thinner better. Possibly it has more dielectric resistance and stuff like that too. Or it might be the same stuff and just tested better to give you the 'reliable spec'.

Lets say you have a small instrument that uses a particular cap. If the big ones can use it, use it, then you can get more volume discount. They have to balance between quality and cost. It still came out really good. Its just like having a internal E-series they tell the designers to use, based on their manufacturing stockpile.

When they are too big it might stress things too, during turn on, absurd values require special management circuitry.


To me that one is kinda a loose spec. Like a ships anchor. It kind of depends on what the equipment is plugged into. But there is a minimum size that you can determine looking at ripple, theoretically, if you consider just a clean 60Hz sine wave with no distortion with some stable load plugged into the output of the equipment with some particular settings, at a standardized line impedance for the generic spec (automotive or residential/office like). But when you start having generators an heavy equipment nearby things can get very complicated I think. 


And parallel caps.. that has to do with
1) better form factor, its how you get a 'rectangle shape' instead of a 'tank shape'
2) lower ESR, its parallel after all
3) possibly more economical if you can pick from your companies "series". They could also pick from values they noticed statistically are in stock more or easier to source from a alternative manufacturer
4) because of #2, it probobly also filters RF better, since the caps are more ideal. That means maybe they can get rid of some other component, like a lowish value capacitor.. so really its possible that you end up with the same amount of capacitors because it turns out you need another one if you go with a big one that has higher parasitic values




Find the best lowest ESR capacitor you can find at some value, and compare it to the ESR you can get with parallel capacitors of equal quality but smaller size.  I am thinking the first one can be alot bigger, the second one, after the resistor, you want that one similar
coppercone2:
and in this case, it does look like a RC filter, that is normally not seen so much, because its wasteful. I think with the RC filter design (it looks like a CRC, Pi filter) they are definately thinking about bode plots when they build it that way, aka they want the frequency response to go out further, meaning capacitor parasitics might be being considered

I am kind of thinking if you combine the 600 and the 800, you might still want to put a 50uF or something in place of the 600, to get in on whatever they were trying to do with what looks like a RC filter


And that resistor, if you just have it right after a big cap, it might choke out the circuit? And if you put the resistor first, it might surge the circuit (the capacitor is too big)

I think, espeically if you put new diodes, the first cap can actually be made quite a bit bigger, and the second cap after the resistor.. you might wanna stick to the value they have. Like a 680uF or 820uF one would be the common choice
Navigation
Message Index
Next page
There was an error while thanking
Thanking...

Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod