Inductor size verification power board

[Tim88]

Howdy, I am working on an old Gould 1604 scope. Ya, it's a piece of junk but it's been a great learning project for me. It's basically dead tho the power light comes on. Took the power board out and immediately noticed L1 (labeled on the schematic as 100µH, 2.5A) had one of its solder joints burned. It is enclosed in some sort of sealed epoxy case and the terminal is burned up into the casing. I tested it with an inductor tester I have and it was 224µH. Long story short, there are indeed errors on this schematic. So I am wondering if the inductance changed from overheating it or the true value should be in the 220µH range instead of 100µH. I found a good PDF on the L296 regulator and used the calculator they had listed to figger a possible inductance. I got:

Vi = 22V
ΔIL = 150 mA  <- clue if I should use this value
ΔV = 3mV <- also no clue if I should use this value
Vo = 5V
f = 100 kHz <- ditto on the no clue but should be fast enough?

L= 17/330000 = 51.52µH (.0000515 x 1000000)

So, I guess rounding up would indeed be a 100µH inductor? I found a 100µH, 6A one on digikey https://www.digikey.com/products/en?keywords=7447070 as a possible replacement. Higher amp rating so it won't crisp maybe? Look forward to comments, been a lurker here for a while.

[PKTKS]

I got  confused  where you post L2 but actually I have notice as L1..

Assuming Li as typo I would note a tricky topology which IMHO I would
**NEVER** use that myself:

1) the circuit uses a single rectifier
2) that DC is split into a virtual floating ground (aka the 2 caps)
3) TYPICAL topology for half-bridge - not for direct buck
4) that "arrange" they seem to did just to economy  cents having another isolated LM317 branch
5) it means that all your reference (including the wild L1 or L2 spikes) are floating on that ground

BAD THING.

I would start checking those CAPS ASAP if they still can split properly the floating ground
and  try that rough guess for a new inductor  **WITH A DUMMY LOAD**

Remember that those inductors can drag the rails with negative surges making
the "common" which should be balanced by those caps at center - float... a lot

if possible i would ditch that crap and put a secondary isolated LM317 on a real common ground
(even at expense of a secondary transformer - should be easy to do that)

If the BUCK still holds the proper voltage - should be fine
I would not worry about negative floats anymore

Paul

[Tim88]

Paul,
Sorry for the typo, fixed it in the original post, it is L1 . The original post I cropped out the other parts of the board to make it easier to read the circuit I was interested in. Had to look up the floating ground, as I said learning experience , and I see what you mean, I think. The transformer is a toroid and is grounded in the full schematic, correct? But, the output negative rail should go to ground as well? It had some weak caps yes, got those replaced. As for the load, I included a diagram from the L296 PDF, 5V output would be a 4.7k and a 3k resistor for about 7.7k total resistance? Assuming this is right, how big a wattage resistor should a guy use?  Really appreciate your help!

[PKTKS]

The complete schema looks better.

The problem with that "arrange" is that the 2 branches are not SYMMETRICAL

Usually you do that with two 317 (pos and negative cousins) or two 78xx
sharing a single rectifier and ground. You hope that being symmetric they will
balance each other - and usually they do.

That is not the case there:
- They have MIXED a BUCK and a LINEAR (negative 317) branch
- With a single rectifier
- Where high and asymmetric spikes will occur (mainly risky on power on/off)

Very bad arrange.
The correct way to do so they did on the other transformer windings
Each winding w/own rectifier  isolated from each other

But to put things scary it "seems" they all share that "floated split" ground on the CAPS
Very BAD. That rail may be dragged and float

I would even be cautions about where to put the negative of the instruments there
But.. VERY SIMPLE PARTS.

Easy even to build another similar yourself  with INDEPENDENT SMPS and Linear branches
free of the spikes and high freq problems

that is quite enough to start thinking a new arrangement  there
separating those 2 branches - putting a common ground  on them all

You need to consider **WHY** they have done that weird thing...
Worst case scenario *IF* they really want that branch to float ...
I have never seen that odd mix.

A last note on that is the SECOND LINEAR regulator  on the upper branch
to put the mix even more weird... a different class from the 317 below...

go figure

Paul

[PKTKS]

2 cents of "thinking"

- being 22 V DC input on the SMPS buck
- anyone can easily try it "OUT OF THE BOARD"  using regular cheap 24/ (12/12) input
- any decent BUCK which handles 22 will handle 24 V as well

That I would try first to make dead sure the Buck works.

That said I would try figure WHY mix that crap all together  that way...
Paul

[Tim88]

Paul,
From what I have read, Gould scopes were pretty notorious for being unreliable, go figger  . The main reason I am screwing with this is to learn about power supplies and design. I have repaired quite a few power supplies which were mainly bad electrolytic caps. Working on a seemingly poor design power supply really gets a guy into the nuts and bolts. I ever ordered a book on them, Switching Power Supply Design by Pressman, which I think I read the suggestion on this forum. I have a 20V DC power supply but now I see the use for a good variable power supply. Maybe Santa will bring me one  . When I get some inductors, just one question when I get to testing it. Is the load resistor values from my prior post adequate and what wattage value to be safe? Thanks!

[PKTKS]

Sleeping on that..

As far as any reasonable reason to use that crappy BUCK there..
I could not figure ...

For a 2AMP +5Volt I would put a boosted TL431 even w/oldskool 2N3055 there
damn simple

That  "virtual split ground" thing is used in POWER HALF-BRIDGES
w/MOSFETS  BJT or controlled SCRs

weird thing to put there where a simple LINEAR 2A/5V would do it

That would require a set of small regulators made home...
I have dozens myself..

Very handy
Not that splitted CAP thing w/center TAP grounded
Paul