Sorensen DCS Series - Modification and Repair Thread

Sweet. . .

I support decriminalization
I dont support impairment while driving, while working, etc.
.
.
Screenshot_2018-08-05-09-57-28.pngScreenshot_2018-08-05-09-57-12.png
.
.
Nothing more irritaring than stoned drivers
(Except stoned people on the job)

-methods
 
Ah....
Another flight thru the "friendly skies".

Well
I am feeling a tiny sense of loss over returning the 600V supply. There is a 420V battery that I have been meaning to charge for the last year or two.

So
For charging with the 600V supply of course any heavy duty HV dioide will work. As voltage rating goes up, Vf goes up, so its safe to assume 1V Vf.

1V * 2A = 2W

I would spec a diode good for at least double... So 1200V (not hard to find).
...

Back to the 60V supplies

Paralleling diodes is generally a "no no".
The reason is the thermal profile.
Diodes become MORE consuctive as temperature rises
If you have 10 in parallel (to handle 20A instead of 2A...)
...
Well... First you are dropping 20W of heat (not insignificant AT ALL)
More importantly the hottest diode wants to conduct the most.
(Negative thermal coefficient)

In parallel they must be at the same Vf
If one dioide is hotter... It will carry more current
If the dioides are thermally isolated - this dioide will pop and you will get a cascade

In proper designs (say... Using a TO-220) where the dioides are heat-linked you will see spikes of non-equal sharing but the parallel array will converge on the same temp, So equal sharing.

This is why you spec a 10A diode for a 1A job btw... Burst handling
(And to make up for inconsistency in parts, thermal mounting, etc)

Necking your voltage down will help bring Vf down, with that P should drop
Right?

... Here is a forum link to a dioide paralling whitepaper
https://electronics.stackexchange.com/questions/53190/diodes-in-parallel-or-series

(Boarding now...)

So
We can handle 20A with a hand full of dioides in parallel
We are looking for the beast thermaly conductive package we can scrap.

Searching Google for clues:
https://www.google.com/search?client=ms-android-att-us&biw=598&bih=122&tbm=isch&sa=1&ei=bGFoW_D5DY-28AP0t56gDg&q=schottky+power+diode&oq=schottky+power+diode&gs_l=mobile-gws-wiz-img.3..35i39.9958.11423..12176...0.0..0.577.2302.0j2j3j2j0j1......0....1.KOHqrPHKKfg

Ah - so thats what we look for
Big packages with 2 legs
Packages meant for chassis mount
Etc

To get your 20W of waste heat down to something more reasonable (like 5W) you are looking at expensive parts (hence why we scrap...)

I would not bat an eye at spending $50 to prototype this (all in)
Where it can be done for cheap if margin is proper and heat can be absorbed.

In EE we always start expensive and whittle cheap
(Unless we are OG... We dont bother with starting cheap and building up)
(Thats a thread in itself ... But it has to do with margin and volume... Again a reason for there to be a VIVID scrap industry for these parts)

...

Back to the ideal dioides that Nick makes... Best bet.
(Based on the linear technology chip, capped at 90V last I saw)

Where we REALLY want to see higher voltage flavors of ideal diodes.

...

One could build a prototype easily
The linear chip is just a convenience
The whiteppaper describes (linear white paper) the considerations of the chip - namely fast transition, stability (overshoot and oscillation), and power consumption.

Yep... We could make a highly inefficient ideal diode...
But that would defeat the point
Right?

...

So to out together a "chargers kit" for Sorensen type supplies...
One would look at:

Packages that can directly bolt in to a drilled cover
A run of heat sink bars and an array (lot of work)
An ideal

Math does not lie with paralleling...
So a bunch of cheap diodes (bulk and work)
Or a very expensive diode (where the work has been done and they are...

As thermally close as possible (same package)

Argument complete for specifying higher end parts over lower end parts (plus slave labor) - even if the later is cheaper*

* support what you want to see industry drive costs down on *

...

Rolling back from the gate.

K bye

-methods
 
Why do you want to use a diode? Most power supplies have diodes in them already so the only thing you would get is no spark when you hook up the battery if the supply is turned off.
 
Not these supplies.

With a battery attached, if you shut down the supply, it immediately smoke out.
(Assuming vbat close to vsupmax)

I also pointed out a few pages back that the manual specifically States that if you want to charge batteries you need to use a blocking diode.

It's the worst kind of failure cuz it works all day everyday forever..... Until you blow a breaker or someone turns off your supply.

20180806_112103-1494x2656.jpg

I suppose if my signal to noise ratio was better you would have picked that up😁

Lately I have a lot of noise to make
We can filter later

-methods
 
Here is the battery charging guidance mentioned:

https://endless-sphere.com/forums/viewtopic.php?f=14&t=95184&p=1394127&hilit=Charging#p1394127

... Not sure if I mentioned it in the guidance... But FOR CERTAIN the supplies SHOOT SMOKE when you attach a 12S Lipo to a 60-18 empowered!

(Mine shot smoke for apx 3-5 seconds before I caught it. It still works fine).

-methods
 
Note: I charge regularly with no diode... But it requires strict order of operations*

Anyone want to unlid and video the event... It would be interesting to watch.

I watched my low power regs shoot smoke when I jumpered for 120 and applied 240. Been meaning to check the fan path. . . . its 120V... Only 18W... Don't recal how it gets that 120. Its not thru the low voltage dc transformer for certain. That's 8V to 16V.

Its. . . not directly connected. 240V would smoke it quick.

Perhaps more jumper magic

...

Remember...
Short P2 if you like big 325VDC sparkles!

-methods
 
Unbelievable value. . .



I have decided that once I have some income coming in... This will be my focal-hobby for a while. Nothing like having 25 Sorensen supplies on the desk...

Remember they stack in series...

High Voltage Electrolysis demos anyone???

...

Its housecleaning day over at methods gymnasium of 6yo's
Sigh... Our cave is majorly wrecked
Roached to the max

-methods
 
This is what you are looking for to buy with a bit more confidence

Screenshot_2018-08-12-09-19-09.png

Even the supplies with a 30 day'noDOA come in with problems.
Not worth the time/money/hassle to return

Pics of load test?
Maybe $400 for low voltage
150V - 600V are understandably more
Still could have a bad fan or harmonics

$250 lowest for low voltage - under that sure to be an issue.
Premium for GPIB

... I want to professionalize the Arduino Analog control...
That would be a $200 to $400 value add to any base model.
Integrate with BMS for safety!
Or just temp probes...

I built one and documented on another platform. It worked (read and write, voltage current & shutDown) but desperately needed a low pass filter on the pwm out. Maybe a voltage follower to be safe and crisp.

USB could be noise prone
Common ground would mean you want to isolate the comma (easy n cheap)

...

-Back to reality-
Contracting
... Figuring it out... Gotta have not two BUT THREE gigs to keep stress in check. Yea you can make more - but if you get side tracked its "long time no eat" on the regular.

Now
... To take a nap

-methods
 
WOW! Just found this thread. You're a man after my own heart - or whatever that saying is. In 2015 and 2016, I scoured Ebay in search of both DCS and DLM Sorensens - Not working/for parts. In the end, I acquired 50 supplies covering DCS 1,000 and 3,000 Watt units as well as the DLM 3,000 and 4,000 Watt (3-Phase) units. I bought a rotary Phase converter to be able to work on these units. But nothing over 120 Volts output. I kept track of everything I did with all costs and I managed to fix 33 of the 50 units and re-sell them on Ebay. I ended up making about $3,500 overall after parts costs. But it wasn't the money that interested me. I could hardly wait for UPS to arrive and deliver another dead supply. I'd pop the top and start looking for any visible damage. It was like Christmas. I really enjoyed the few years of doing this and then the re-sale costs started to fall a lot to the point where I would lose money after sellers fees and shipping costs.

I had only two very old units like yours with the two long input section caps. Most that I got had the bank of 6 smaller caps and newer output chip-style FETs.

I bought a stock of parts so I could fix them as soon as I figured out what was wrong. I have to say that at least 1/3 of them arrived with no D-Sub connector with bridge pins. As soon as a proper configured plug was installed, most fired right up and worked under full load. (I bought and still have a TDI brand 4000 Watt electronic load that I used for 20 minutes on each unit at 85% load before I sold them)

Fans, input section and output section caps, a few bridge rectifiers, a relay or two, sometimes a bad GPIB board and a few other minor component swaps where the component being burned was obvious were the problems I usually found. Only once did I find a blown fuse. The new fuse fixed it and it passed the 20 minute near full load test. I never had one returned although I did end up buying one back (didn't know it at the time until I saw some of my markings inside) It came to me with a squealing part that I was never able to find, so I guess I technically had one "come-back".

At one point, I could not find a DCS 1K unit fan for under $65.00. I ended up drilling a tiny hole in the pressed-in cap visible on the top/center of the fan. Then I used an Awl to gently pry off the cap. This allowed access to the retention circlip that once removed, gave me access to replace the tiny ball bearings. Worked like a charm.

I kept only one of the supplies for myself - The Sorensen DLM 32V/95A unit. I'm pretty sure it was brand new as seller described it. I got it for $795.00. A school bought a few of them for a grant project and didn't need them all.

I still have new fans, tons of caps, some rectifiers, relays, resistors, display digits and other misc parts. Maybe I'll do it again someday.

Just looked on Ebay for fun. The first 4 not working DCS 33/33 supplies for $100, $100, $125 and $125 are all from the same seller (one that I bought from) and all have no D-Sub connector. Good chance that's all that's wrong with them.
 
20180812_161649-1494x2656.jpg

From phone...
Working on this bad-boy so we can take a night ride.

Will return and read.

-methods
 
Below in RED



BVH said:
WOW! Just found this thread. You're a man after my own heart - or whatever that saying is. In 2015 and 2016, I scoured Ebay in search of both DCS and DLM Sorensens - Not working/for parts. In the end, I acquired 50 supplies

... :shock: ...

covering DCS 1,000 and 3,000 Watt units as well as the DLM 3,000 and 4,000 Watt (3-Phase) units.

... Three phase is cheap in every brand... but hard to make use of without a converter... solid state, cheap, 3-phase converter???

I bought a rotary Phase converter to be able to work on these units.

... this is the converter that works by https://www.google.com/search?q=rot...chrome.0.0l6.366j0j7&sourceid=chrome&ie=UTF-8turning a motor yes?


But nothing over 120 Volts output.

... DC or 3phase, I am unclear

I kept track of everything I did with all costs and I managed to fix 33 of the 50 units and re-sell them on Ebay. I ended up making about $3,500 overall after parts costs.

... LOL, minus 40,000 hours of your time :mrgreen:

But it wasn't the money that interested me. I could hardly wait for UPS to arrive and deliver another dead supply. I'd pop the top and start looking for any visible damage. It was like Christmas. I really enjoyed the few years of doing this and then the re-sale costs started to fall a lot to the point where I would lose money after sellers fees and shipping costs.

... Yes, I know this feeling. It is similar to the wood craftsman who spends 400 hours building a chair leg for a neighbor.
Its not about the capitalist perspective... its about the moments of not feeling anxious (for me).


I had only two very old units like yours with the two long input section caps. Most that I got had the bank of 6 smaller caps and newer output chip-style FETs.

... Yes, the E-Series good for more power ...

I bought a stock of parts so I could fix them as soon as I figured out what was wrong. I have to say that at least 1/3 of them arrived with no D-Sub connector with bridge pins. As soon as a proper configured plug was installed, most fired right up and worked under full load. (I bought and still have a TDI brand 4000 Watt electronic load that I used for 20 minutes on each unit at 85% load before I sold them)

... yea... the Dsub confuses new users. I try to keep fresh in my mind how confusing the supply was upon first observation. I considered spinning a little Purple PCB board for those :p I thought that could be cool.

In the end I made an Arduino Interface that reads and writes V/I. I wrote the interface for the Isolated Dsub (different pinout) so I still need to spin a variant for the non-Isolated.

... you... eh... want to part with your parts collection and E-Load? I cant buy right this second - but if contract work picks up I would not bat an eye and offloading this gear from you. I am quite passionate about fiddling these supplies into a condition where they are more noob friendly. I want to see one of these (in 120V Single) on every E-bikers bench.


Fans, input section and output section caps, a few bridge rectifiers, a relay or two, sometimes a bad GPIB board and a few other minor component swaps where the component being burned was obvious were the problems I usually found. Only once did I find a blown fuse. The new fuse fixed it and it passed the 20 minute near full load test. I never had one returned although I did end up buying one back (didn't know it at the time until I saw some of my markings inside) It came to me with a squealing part that I was never able to find, so I guess I technically had one "come-back".

... yes... the squealing is the next area of interest. I have not replaced any output caps yet. I cant tell if the noise is coming from the caps or the Transformer... easy enough to tell... with a set of "known good" swap-in's.... but to get to that lelve, I have to "tool up".

At one point, I could not find a DCS 1K unit fan for under $65.00. I ended up drilling a tiny hole in the pressed-in cap visible on the top/center of the fan. Then I used an Awl to gently pry off the cap. This allowed access to the retention circlip that once removed, gave me access to replace the tiny ball bearings. Worked like a charm.

...SWEET... I have been looking at a fan on my desk thinking about that. It is riveted together. On this fan the blade is visibly broken - I am presuming it is pressed on. I will drill it out and post pictures.

I kept only one of the supplies for myself - The Sorensen DLM 32V/95A unit. I'm pretty sure it was brand new as seller described it. I got it for $795.00. A school bought a few of them for a grant project and didn't need them all.

... Wow, good price. Funny... in my mind the DLM were always the 600W half size units. I was actually searching for reference to correct you (thinking you were mixing up the series...) but... looks like you taught me something :D Thanks man!.

I still have new fans, tons of caps, some rectifiers, relays, resistors, display digits and other misc parts. Maybe I'll do it again someday.

...Nope - they will decay in your garage until one day far in the future you will sadly toss them into the bin :oops:

Just looked on Ebay for fun. The first 4 not working DCS 33/33 supplies for $100, $100, $125 and $125 are all from the same seller (one that I bought from) and all have no D-Sub connector. Good chance that's all that's wrong with them.

... It was really great to hear from you, thanks for taking the time~! Early on I was a bit... eh... "covety" with the Sorensen information (this was a few years back). I was scared everybody would buy them up :mrgreen: Now I am over it and I really want to get a community started around this form factor (the 1U, 120V). It looks like there are a ton of them out there and they can be maintained over time.

Your post will likely be the nudge I needed to put the time into the Adruino Adapter.
If you see it (6mo or a year from now) - know that you had a part in it coming to be. :wink:

-methods
 
I bought an American Rotary 10,000 Watt capacity (depending on type of load, the rating is de-rated) motor generator unit that creates the 3rd leg.

By "nothing over 120 Volts" I meant I never bought any other high voltage DC output supplies. I think I had one with a DC output of 120 volts, give or take. Or maybe it was 150V ??? Can't quite remember. Lots of 80 Volt DC output units.

The way I looked at it, I'd have done the same thing if I could break even. It was just so much fun diagnosing and fixing them. My time is free!

It would be good to get the parts stock to someone who's going to use it. As you say, they will age just sitting there unused. The E-load is a different story. I actually have their 4000K unit and their 1.5K unit. Not the newest generation, but not the oldest either. I just can't seem to want to get rid of them even though I have no use for them - other than if I wanted to test the capacity of my batteries. The 4K unit weighs a ton.

I tired changing output caps on one squealer even though they visually looked fine and it did not fix the issue.

I forgot about the fan case being riveted so yes, I drilled out the rivets and used bolts upon re-assembly. IIRC, the fan wheel is part of the hub and central shaft - all one piece. But not 100% sure I remember correctly.

I, too, was closed mouth about the Sorensens. I was having too much fun fixing them and wanted a steady supply of them. But that was then and this is now.

I run a full capacity rated set of Blocking and Freewheeling diodes on the output of mine as Sorensen recommends for high inductive kickback loads. (Short Arc searchlight power supplies) Their screwed into a 1/4" thick, large copper plate. I remotely sense my output Voltage after the diode to get the actual output voltage after the diode loss. I also have a "Transient X X Diode" across the output terminals as is recommended. I can't remember the actual 3-word name of the device. It's a sacrificial device designed to fry under certain conditions instead of your PS frying.

I still enjoy discussing these PS's so will bookmark this thread.
 
methods said:
... Three phase is cheap in every brand... but hard to make use of without a converter... solid state, cheap, 3-phase converter???
This isn't efficient, but:

Take your biggest brushless hubmotor, and hook it up so it's driven by another motor. Power the second motor from something that lets you adjust the speed of the system to provide the right frequency out of the first motor (50/60hz). Hook up the phase wires of the first motor to your 3-phase Sorenson AC inputs. If you have to have the fourth wire, maybe you could bring out the star conection from the motor.

If the hubmotor is the right winding, you can have it running at that speed and make a lot of power. Most likely it would have to be rewound to make the proper power at the proper speed.
 
Got the pack finished and put it on the beach-runner.
Note the vintage front 9C motor fitted to the rear

... We use Anderson PP45's. To keep them together we use tiny zip-ties. For single Anderson connections (Series connections) we add empty housings. For the balance taps we either leave them open or parallel using "Parallel Boards" and board jumpers.

20180811_132820-3984x2241.jpg

And... Lots of recruiter action this week so maybe there will be money in the coffers to buy a few supplies.

...

Fan landed for my 60-18E

No feedback on the supplies delivered
(Note - as mentioned above the proper procedure is to "burn in" a repair job at various loads. Let everything get up to thermal equilibrium, test transient response (step and impulse loads), confirm noise on output and input... Where in the absense of a repair manual we would measure relative to a known good)

... See... I am already trying to turn it into a business :-( ... Business sucks and kills hobbies. All a hobby needs to do is offset its own cost to some degree. Maybe 50%, maybe 120%. Its a monetary loss by its very nature... Where the gains can not be calculated easily using capitalist metrics.

(I.e. ... How will you explain to the wife/husband how much it is worth to gain knowledge, experience, self confidence and reliance, etc. It pays off for your employers... Just not directly to your bank account)

-methods
 
from phone in woods
Missed above replies
Will return when at PC
-methods
 
BVH said:
I run a full capacity rated set of Blocking and Freewheeling diodes on the output of mine as Sorensen recommends for high inductive kickback loads. (Short Arc searchlight power supplies) Their screwed into a 1/4" thick, large copper plate. I remotely sense my output Voltage after the diode to get the actual output voltage after the diode loss. I also have a "Transient X X Diode" across the output terminals as is recommended. I can't remember the actual 3-word name of the device. It's a sacrificial device designed to fry under certain conditions instead of your PS frying.

Hmmm... Good info.
I was aware of the inductive load constraints but had not spec'ed parts for it.
So... If we are to make a "package", that updates a standard early 90's Sorensen... It needs (or should have):

1) communication retrofit (with temp probe inputs, BMS input, and LabView interface/logging/control.

2) proper reverse connection protection

3) proper inductive load protection

4) proper transient noise supression
(TVS = Transient Voltage Suppression)

Yea... Sweet
Soliciting/Collecting Requirements :)

Sounds easy enough... But it can be real tough to figure out what your customer wants/needs, even when you are your only customer.

...

-methods
 
Couldn't find mention of transient protection in manual, but it makes good sense to strap the DC output hard with a heavy duty TVS. It will crowbar any nasties that flash back and let the warning smoke out of a better spot.

Sacrificial parts are good when we know they are there.
Azure Dynamics found out they are not so good when margin is tight
... We still have that Azure Van - that as far as I can tell is probably fine...

Someone just needs to troubleshoot the problem component and pull the TVS... Then I will put the pack back in and its dialed for sale or use.

Sheeeittt... I need to go work on that :-/

-methods
 
Now that I think about it, I think I was advised to use a TVS in an email reply from one of the manufacturers (now Ametek) company techs.
 
Determining how 120V gets to the fan
(from a source which can be 120VAC or 240VAC)

Supply: DCS60-18E
(we will indicate anything with an E suffix as 2nd Gen)
(Example: DCS60-18 == 1st Gen, DCS60-18E == 2nd Gen)

Fan hits the two layer board at solder point B1 (flying leads)
Top trace leads directly to resistor pair R137||R138 (in parallel)
Relatively "small trace" of unknown weight, call it 40mil
They appear to be 1/8th watt zero ohm (single black stripe)

At the same node we touch our selector jumper at P1-4
Monster trace over 200mil (so handling power...)

(I am only viewing the top of the PCB... as it is installed)
Limited visibility into bottom

After shunting (as opposed to shunt) resistors heavy trace leads to the center tap of the main DC cap pair
(Weird...)

Reviewing our previous traceout (which NOTE... was for a Gen 1... so BEWARE!!!)
file.php


...
Back to B1

The second trace follows the first under the board
It via's up next to R115 (power divider/bleeder/balancer on main RectumFryer caps)
From there it hits P1-3 via a mid-size trace (maybe 100mil or 80mill)

OK

So... our fan directly connects to P1-4 and P1-3
Thats where we will start - ignoring all the rest

DOH

Hold the phone...
(Saw we were across that 60ohms that we were wondering about earlier... went and measured the fan DC Resistance... eh...

Original Fan:
Rated for 115V @ 170 to 190mA

115*.170 = 19.55W
That works out to 20.7W ... ok...
115*.190 = 21.85W

New fan is MUCH heavier (tho made of metal instead of plasticrap)

New fan rated at 115VAC, 19/18 W (WATT, not AMP... uh... lol... its ok... but I missed it (and caught it) before)

18.0W/115VAC = 156mA
18.5W/115VAC = 160mA
19.0W/115VAC = 165mA

So we are good (lol - I know... but mistakes like that cause missiles to crash)

What threw me off (and made me do the math above) is:
OLD FAN: 220 ohms
NEW FAN: 120 ohms

Well... its an induction motor of unknown winding pattern with an AC input so - at this point in time, dont know, dont care.
Onward

...

P1-3 we can see is tied directly to the AC input of the RectumFryer... so we know that is touching source
(be it 240V or 120V...)

P1-4 (IN THE 120V Configuration) jumpers right over to the other leg of the Fryer... so we are golden

Now... for the 240V config :D
(what we are after here... stay focused... lol... stay focused...)

P1-4 jumpers over to P1-1 IN THE 240V config... and nothing else...
So -
I am guessing we are cheating and using the primary of the LV step-down transformer in a tricky way to feed the fan?
...

Be back in a second
I need to go back to the Step-Down transformer input wiring
It has multiple configs depending on P1
I want to find that Frugging 60 ohms too...

-methods
 
Ok - here is the post where we dug into that shit-pile
(and IIRC... I had made a mistake that I only corrected pages later... lol...)

https://endless-sphere.com/forums/viewtopic.php?f=14&t=95184&p=1396672&hilit=transformer#p1396672

... see... this mess converges... but the YouTube generation just do all this (REAL WORK) behind the scenes and show you the FINAL PRODUCT

Which... is entertaining
And... makes them look cool
But... does nothing for teaching men to fish (or boys, girls, LGBT, LGBTQ, whatever... just replace Man with yourself.. k? I remember before there was a Q on the end... and before the T... and before the LGB when it was just Gay... and before that when we didnt talk about it... and before that when we DID talk about it... and... yea... sigh... you get to own it now... yes its your turn :p Just dont yell at me please. I look out for the community all the time - so - if you are looking to parse semantics, look elsewhere. I use the word MAN because ... eh... old guys thousands of years ago did and... uh... they look like Jesus... so... what?)

lol
What were we talking about?
Oh yea...

I am off to go parse the primary input windings to the fan

YES
IT IS OK TO TALK ABOUT STUFF... WE ARE NOT AT WORK

Girls
Boys
Weiners
Vaginas... sigh...

and this is NOT a kids thread!

This is not a work safe thread either!
If you are at work... call me on my Paid Line lol... dollar a minute (er... more like $2.30 - but whatever)


-methods
 
OMFG...

I just got accounted...
(audited?)

We must also address the stark difference between the words Person and Entity
(since the debacle with Entities being considered persons... and...)

Entities are not persons...

so - from here forward - if I say "Teach Men to Fish"
Enter anything (robot, alien, human)
But do not enter Business, corporation, etc

Parsing... Semantics...
This is what Politicians use to avoid ACTUAL TOPICS

... like when I asked our Representatives (when they came to town) about 1 thing: WAR?

back on topic

-methods
 
Documentation:

Documentation is important!

First: System Documentation

These days people think that the summation of individual documents equates to a System Level Wiring Diagram (or hybrid)
IT DOES NOT

When you have 50 cables, 15 complex components, 6 kinds of serial communication...
Its
Time
To
Make
A
SYSTEM
LEVEL
WIRING
DIAGRAM

lol

Back in the day we also kept our Schematics all wired together
These days lines fly off the page everywhere... which makes them easy to read... BUT HARD TO VALIDATE

At least in paper
Or... PDF...

Unless you note what page a node goes to (which good engineers do)
Another good thing to do is note how many places a node that goes off the page goes to

... Because all the other groups reading your schematics....

ARE NOT
I repeat
ARE NOT
looking at them in your development enviornment

We cant just click and highlight where your traces go off the screen!

...

We are sitting there scratching our heads with a highlighter mapping everything out!
That is expensive and error prone
It is also necessary...

Because in Quality or Test or Validation or Reliability...
The FIRST thing you do is make sure you understand what you are testing

IF
You dont understand what you are testing
OR
You dont have sufficient documentation
THEN
The results of your test... or rather the VALIDITY of your test - comes into question

Unless
You are trusting someone else to tell you what and where and how
But
I dont do that

Because... I once took part in a very important validation
and
I worked in a Nuclear Complex for a decade
and
... more importantly...

I have done a lot of design work

MISTAKES GET THRU!!!!!

Just because you are looking at a pretty PCB board that is in production
All conformal coated and nice...
With long part numbers and stickers
And... the PCB is being used all over the place...

DOES
NOT
MEAN

that it is error free.

...

Double Checking and sometimes even triple checking
(with different eyes)
Is mission critical for things that involve safety

Safety in our world means:

High Voltage or ... anything needing interlocks (like a throttle or contactor)

-methods
 
And this is justified by the old "two birds with one stone"

If
You find yourself needing to "figure out what a PCB board or Schematic does"
well...

You can charge hours - eh... "learning"
Lol
That does not look good right?

I mean... yea... you can skim a schematic in 15 minutes
Thats not what I am talking about here

...

Truly understanding what a circuit is supposed to be doing comes from 1 of 2 places
1) The Requirements

in a vacuum of requirements

2) An understanding of the circuit

so

The way you kill 2 birds with 1 stone is to VALIDATE or REVIEW or CONFIRM or otherwise do a formal "lookover" of the schematic.
This is justified by you signing off on the circuit
Providing some notes or feedback
Asking questions where questions should be asked

...

In the end -

You understand the circuit (that in this rant you have to test... cuz remember I am playing Test Engineer lately)
Your design engineer feels better because someone check their work
Your Quality team or whoever is Validating can count points

and the best part...
drum roll please.....

IF

It should later be found that there is some glaring problem with the PCB/Schematic
AND
You signed off on it
THEN
You should get a swift kick in the ass

Right?

... :mrgreen:

Accountability
One way to do it

So

When reviewing and signing off on things take LOTS OF NOTES!

IF
You are not back-calculating all of the power dissipation... note it
Perhaps suggest to the engineer that they may want to put little notes like that (calculations) in the schematic

.........
Recently I had the pleasure of reviewing some work that was done by a guy with hair in his ears
(not literally... figuratively)

In his schematic there were all sorts of really good notes
Little calculations
Little assurances...

It felt good reading the work and I had confidence in the engineer.....

so
Think about it

From both perspectives or wearing both hats
The designer
The validator/quality person/test engineer - whatever

-methods
 
Kid woke up
Mini-rant over :x
(that one had a 6mo delay on it)

Next Episode (no pun intended) we will figure out what is going on with that primary and the jumper to feed 120V to the Fan

-methods
 
P.S.

Soldered in new fan
Ran in the 11 fasteners into the lid
Fires up and works

Fan may be louder.... but I like loud fans
Blows like hell

... to celebrate...

We will charge up the Ebike's and ride them out to dinner

(McDonald's of course... because we are "balling on a budget")

-methods
 
Back
Top