Sorensen DCS Series - Modification and Repair Thread

If it is helpful, I have a working Sorenson DCS 55-55 that I can check parts values, voltages, etc if you can tell me which ones to check.

(I also have some ancient linear Sorensons but their designs are unrelated to the DCS types).
 
Thanks for offering

I think once I actually put in the work to figure the calculations of that part of the circuit the answer will fall out.

That... Or an exact replacement DCS60-18E without the mod, then bruit force compare component by component.

I always feel fear/anxiety/avoidance about circuits like these. They are non-intuitive, seemingly impossible in my mind... Until I completely understand... Then CLICK - it feels super obvious.

Like that with all complicated things

* eBike controllers
* Solar all in one inverters
* Python programming language
* Bleeding hydraulic brakes
* Building stair cases
* Patching roofs

All start off feeling like mountains, then later feel like mole hills. If only it could be the other way around.

-methods
 
All start off feeling like mountains, then later feel like mole hills. If only it could be the other way around.

Sounds like the entire planet I'm still trying to find a way to land on, to get really going on my Snuggles-the-Wolf project sometime this decade:

I thought it was just a mountain, then discovered it was a series of ranges...then I found out I was still in orbit over the planet they were on....

I "know" what needs to be done, I just don't know how to do it and haven't figured out how to learn all the things to get to the starting point yet.
 
Replacement Fets
They are priced anywhere from $18-$28 each, claiming genuine. Out of China they are going for a couple bucks each in lots of 5pcs. All sorts of clones etc, so I just went for it on the first open-box-rando I could find.

8pcs for $6 open package + $4.20 shipping

They will either work or they wont, we will see. The markings are very different than those in my unit, the mosfets were BOSS in their time, time has moved on... so ... figuring Clones may be as good or better than the original shizzle.

... I WAS going to try and swap in some modern-package fets but that would make no sense. The later E-Series already does that and I could get a job, work 8hrs, and buy 8 for-parts units and be better off.

SO -> Rewinding Logic and packing the ol' girl with what very well may be counterfeit garbage.

-methods
 

Attachments

  • JenkemGarbage.png
    JenkemGarbage.png
    60.4 KB · Views: 2
Built a new bench on dirt floor, onward to DCS33-33M15 rebuild

Packing with the eBay mosfets

Fastening the Motherboard back in

Taking it to turn-on Test

Sweet!
Now it's producing output and banging off of OVP. PROBABLY SHOULD HAVE TURNED THAT DIAL ALL THE WAY DOWN...

-methods
 
DCS 40-25 Repair Thread
Good Guy to help!
 
Status

* The DCS33-33 now has replaced mosfets and fires up... but the logic is fubar. Certain the links above will lead us to a solve

* The DCS60-18EM52 is still only producing 5A of the 18A it could... BLAST that M52 option. Maybe I can solve it after I repair the above

* The DCS600-1.7 still works great. I keep it in a gasket-case otherwise moisture gives it hell on cold start. If kept in a $6 used-store suitcase it always fires up first turn. Been testing the hell out of Solar Inverters ... such a time saver!

I have been repairing Grid-Tie and Off-Grid inverters. GroWatt (IGBT) and Sunny Boi (missconfigured switches and fuses). Buying more as we speak, so if anyone needs a repaired Grid-Tie, I may have one for you.

(Moving the 600V from the Pink Biatch to the $6-Speical)

If anyone knows how to undo the M52 option that limps the 18A supply to 5A I will give you $100 USD

-methods
 
Just bid on an 8V 125A model for Parts/Repair.
Lately we are REALLY needing to create high current for connector validation. With so many counterfeit connectors floating around now (think Amazon MC4, knock-off Anderson, etc) we are down to validating counterfeit.

We use mostly 10AWG... so around 50A will speed that test up. 125A is very useful for calibrating larger shunts, heating up ring terminals, charging paralleled cells REAL QUICK, etc.

* 600V was the best thing I ever bought broken and fixed
* Hoping the 125A will treat me the same

The larger (2U) units are just too damn heavy to put in the car or tote around. The 1KW and 1.2KW are as close to reasonable as it gets. Sure there are modern supplies that are now probably a better value... But what would we learn then?

I WILL dominate the Sorensen DCS series until I can diagnose and repair every single problem they have. If you are into it, check out GroWatt inverters, Xantrax Inverters, Sunny Boy Inverters... they are simple 2 layer board designs like these, unpotted, easy to visually diagnose and repair

OK BYE
-methods
 
If anyone knows how to undo the M52 option that limps the 18A supply to 5A I will give you $100 USD
I'm not enough of an engineer to figure it all out, but (based on possibly misunderstanding some of your previous posts) since it can be programmed, is it likely that the limit is in the firmware of that part?
And if so, would getting the chip from a non-modded version (copying it if possible, if it's not read-protected) and replacing the one in this undo the limit?
 
Good idea dude!
I actually think it is all analog.... :-(
 
Anyhow - Understand the Sorensen DCS, understand the 600V MPPT
That is my goal... and MPPT is easier I am starting to think!

Sorensen just spells it out in Logic Chips... where as MPPT requires firmware

-methods
 
Good idea dude!
I actually think it is all analog.... :-(
If it's all analog, how does it get programmed over GPIB? I thought that was a digital interface between instruments? (but my tiny knowledge of GPIB is from the late 1980s or early 1990s, so....)
 
They have "OPTIONS"
* M52
* M58

These are slide-in cards with interfaces.
* Analog
* Isolated
* GPIB

The supply itself is programmed via Voltage, Current, or Resistance... so... somewhere there is a BoM change for a resistor divider, potentiometer, etc. All the power-guts are the same. ... I think if I just take the time to repair the 33-33 I will come across that circuit and we can do one-for-one component comparison

Sadly - only the OLD series have a full bill of materials and deviation description. The newer "E" series do not have these listings of components for repair - so - you have to have a second unit and measure part for part, in circuit. Not hard, just time consuming.

You are on the right track tho!
I will remember your idea for future reference (and claim it as my own....)

-methods
 
So if there's a card for GPIB, but it can program changes to operation, but the PSU itself is all analog, it means the card interface slot for the GPIB has tie-ins to the appropriate points to do those changes, and analog circuitry on the card that would accomplish that.

But...presumably the mod to reduce current capability couldnt' be overridden by GPIB commands, so those points wouldn't be able to override it either. :/
 
The 600V supply has worked great for years. I finally found its complement

* 0- 125A DC
* 0-8V DC
* 1KW from 120VAC or 240VAC
* Large but light weight

Make-Offer on Ebay for $125
Got it for $200 shipped out of Texas - PARTS ONLY


Got lucky and it pretty much ran out of the box

* All but 3 case fasteners were missing
* AC and DC screws were missing
* No DB25 interlock plug
* Loose voltage dial

... Rigged it up, fired up

* Got 42A out of it without any work
* 350V rail is solid
* Fan works
* Main input caps are good

...
* Will only produce 0-1V
* Will only produce 0-42.0A

Based on my experience with the 600V supply... my first rando-guess is that the output caps are shot. When I ran the 600V supply with (NO!) output caps what so ever... it behaved similar. It would produce a couple hundred watts but it would not hit the rails.

MY GUESS:
Is that the lack of output caps results in significant feedback noise on the control circuit, resulting in garbage. Right at turn-on it bangs off the 8V limit, but then it wont climb over a volt. My first venture will be to
* Pull the Motherboard
* Test the Output caps

They are something like 1,200uF 10V

O N W A R D!

(No update on the 33V 33A since I got the mosfets swapped. That one still needs logic troubleshooting)

-methods
 
Purpose
We are starting to feel the pain of MC4 miss-match in the field. Long Story Short... it is basically impossible to avoid counterfeit MC4 these days unless you want to retrofit EVERYTHING (sigh) with EXPENSIVE (sigh) name brand connectors. If you want to plug-n-play, its hit and miss (just like Anderson PP45 used to be).

So
We are going to set up a Screening Test. I am thinking 42A Continuous will ferret out the stinkers real quick


-methods
 
Note how beautifully that Jenkum connector failed!

* No big blue arks of lightning
* No exposed 420VDC
* It did OPEN

So the thermo plastic did what it was supposed to do, even tho the contacts did not

-methods
 
DCS8-125
1kw 8V 125A
Repaired

Sold for Parts / Repair $200 shipped

RCA
* Counterfeit output caps 2/3 stone cold dead and silicone glued in
* Someone snatched the remote control daughter card BUT did not put the SW1 switches back into LOCAL
* Biscuit fan leaking goo and hard start
* Rust internal

Presented as
* Lid had been popped, missing screws, no DB25, no whip
* Main 1,800uF caps sweating electrolytes
* Not much dust
* Only output 1.2V 47A
* Set to 240V

Changed
* Set to 120 mains
* Add Dsub jumper
* Add mains whip
* Adjust SW1 to local

Throws full voltage and full current rail to rail, still needs a 1kw burn in, $10 in Output caps, some housing fasteners, a biscuit fan, Validation


-methods
 
I think the DCS33-33 will solve now, we just have to replace that overheating resistor and/or figure out what is loading it down beyond its design intent.

That supply will be used for fast-charging 4S LiFe packs on the bench / BMS testing

-methods
 
If the 33V solves,
Credit to Amberwolf for looking back at the daughter card interface. ... That is DIRECT control over the dials and a definite place to look in the future for issues.

Manual explains the interface and settings
 
From the original Manual
* Differential parts, power assembly

C51 - C53
10x30x35mm Snap In
10V 22mF


-methods
 
Unit measures out to
* Accept 35mm wide
* 30mm tall

So a real oddball. Whatever was in there seems non-legit.

-methods
 
Digikey has a poor selection
Mouser has MANY more

* Max the length at 30mm, pitch at 10mm, snap in, 10V, diameter from 30-35
(Many Results! Much higher density)


But... That is without the IN STOCK filter running (which I ran on Digikey)

Then you get down to about the same 3 options
//// Sadly those are 6mm terminals, so no dice

:)

Back to the Digit'key


***********************

I settled on these units which are actually 32mm high (may hit lid) but sum to the original capacitance and would/should/may/maynot work. 2nd day shipping, $20 total

#PRODUCT DETAILSQUANTITYAVAILABILITYUNIT PRICEEXTENDED PRICE
1‎495-B41231A3159M000-ND‎
‎B41231A3159M000‎
CAP ALUM 15000UF 20% 10V SNAP TH
1Immediate2.07000$2.07
2‎495-B41231C3479M000-ND‎
‎B41231C3479M000‎
CAP ALUM 47000UF 20% 10V SNAP TH
1Immediate3.70000

Could have solved it any number of ways... but I chose to add up to the original ~66mF(ish).

Close enough for Government work, $20 fix... so 10% investment into my $200 investment to make it worth $900

-methods
 
Back
Top