Sorensen DCS Series - Modification and Repair Thread
- Thread starter methods
- Start date
methods
1 GW
The 60-18E successfully charged a 12S 10Ah pack as well as a 12S 25Ah pack.
We rode them here:
By quadcopter or feet + swimming its only a couple hundred yards. By non-floating bike you have to round the harbor.
If your in Santa Cruz this is where you want to land. On the left you can catch the edge of the Crows Nest. Wed is all night happy hour and you proceed upstairs for that. Any other night you eat on the dock, at what used to be the bait shack.
Wood fired pizza, hot sandwiches (same as you get next door), all at a very reasonable price. For that reasonable price you also get to sit right on the water.
(No... We do NOT eat at McDonald's... EVAR!@$!!)
(Shit is rancid poison that kills kids and gives you candido)
(Seriously... F' McDonald's... I go there only to purge my bowls)
Zooming in... (Android in low light)
View attachment 1
Take the trail around the harbor.
If you explore it in full you will find a path up to a field. The field used to be wild... And now it is all packed full of signs and fences. Ignore those and turn right on the dirt trail. Head back there far enough and you will find epic oak trees to climb. You have to cut back into the trails to find the spots. When we were kids we used to smoke pot back there.
A word to the wise... If you buy any pot (which is legal) in Santa Cruz GO EASY on it. One small puff is all you need to crash your ebike and wreck your face. So... There ya go. I dont smoke it, it makes me see machine elves and hear Terrance McKenna ramblings coming from the bushes.
So
..........
Get you wona deez Power Supplies and get ready to live in outer space. The earth is becoming overcrowded and its time to explore space.
-methods
We rode them here:
By quadcopter or feet + swimming its only a couple hundred yards. By non-floating bike you have to round the harbor.
If your in Santa Cruz this is where you want to land. On the left you can catch the edge of the Crows Nest. Wed is all night happy hour and you proceed upstairs for that. Any other night you eat on the dock, at what used to be the bait shack.
Wood fired pizza, hot sandwiches (same as you get next door), all at a very reasonable price. For that reasonable price you also get to sit right on the water.
(No... We do NOT eat at McDonald's... EVAR!@$!!)
(Shit is rancid poison that kills kids and gives you candido)
(Seriously... F' McDonald's... I go there only to purge my bowls)
Zooming in... (Android in low light)
View attachment 1
Take the trail around the harbor.
If you explore it in full you will find a path up to a field. The field used to be wild... And now it is all packed full of signs and fences. Ignore those and turn right on the dirt trail. Head back there far enough and you will find epic oak trees to climb. You have to cut back into the trails to find the spots. When we were kids we used to smoke pot back there.
A word to the wise... If you buy any pot (which is legal) in Santa Cruz GO EASY on it. One small puff is all you need to crash your ebike and wreck your face. So... There ya go. I dont smoke it, it makes me see machine elves and hear Terrance McKenna ramblings coming from the bushes.
So
..........
Get you wona deez Power Supplies and get ready to live in outer space. The earth is becoming overcrowded and its time to explore space.
-methods
methods
1 GW
BVH said:
Oh yes - the topic
Hmmm... Interesting
At all the places I have worked that ship to the UK, it is certain that they hold a higher standard than we do. More shielding, more safety, and a good bit of fussing to go with it.
For mass market (due to capitalist pressures) this is good.
For hobbies and one-off... It should be "wild west"
... "Freedom to build what you wish is as fundamental as freedom of speech"
It admittedly gets out hand at times... With all the noise (audible and interference) and disruption to others and the environment.
... Until we move away from oil and gas... Let the revolution continue...
-methods
methods
1 GW
And speaking of Ruckus... Ruckus there will be now.
Just dropped my boy off at his moms so it is time to remove the kid trailer from the trike, turn up the voltage and current limit, and "2 wheel motion" the town.
I have riding it on 2 wheels dialed. Faster the better. Minimal riding gear is leather shoes, jeans, padded motorcycle jacket ($20 used... No blood stains), and a full face downhill (bicycle) helmet.
Tip: Motorcycle helmets are roo heavy with not enough peripheral vision.
Tip: traditional top helmets don't provide adequate chin protection
Rational?
When you go down you want to slide, NOT roll... And you wont have MUCH choice about how you go down... So protect your bits.
I have gone down and slid.
Walked away
I have gone down and rolled
Went to hospital, arm still sounds like cap'n crunch cereal
Find out for yourself. :wink:
... Anyhow
Work to get caught up on
Trouble to stir up
And a political agenda to work on
I will update the thread next time I get my hands on a Sorensen.
-methods
Just dropped my boy off at his moms so it is time to remove the kid trailer from the trike, turn up the voltage and current limit, and "2 wheel motion" the town.
I have riding it on 2 wheels dialed. Faster the better. Minimal riding gear is leather shoes, jeans, padded motorcycle jacket ($20 used... No blood stains), and a full face downhill (bicycle) helmet.
Tip: Motorcycle helmets are roo heavy with not enough peripheral vision.
Tip: traditional top helmets don't provide adequate chin protection
Rational?
When you go down you want to slide, NOT roll... And you wont have MUCH choice about how you go down... So protect your bits.
I have gone down and slid.
Walked away
I have gone down and rolled
Went to hospital, arm still sounds like cap'n crunch cereal
Find out for yourself. :wink:
... Anyhow
Work to get caught up on
Trouble to stir up
And a political agenda to work on
I will update the thread next time I get my hands on a Sorensen.
-methods
methods
1 GW
Working on a DCS33-33M15
* Arrived DOA
* Blown main fuse - F3 - 10A 250V Fast Blow
* R98 cooked off
Playlist (incomplete at time of writing)
* Main caps are good
* Dsub wired correct
* Switches set correct
So we need to find the original manual. The later versions do not include component values.
Combined newer manual
M15 Option Description
Original Manual
(was linked in this thread, I dont see it anymore)
All of these are saved off at SchindlerEngineering.com
-methods
* Arrived DOA
* Blown main fuse - F3 - 10A 250V Fast Blow
* R98 cooked off
Playlist (incomplete at time of writing)
* Main caps are good
* Dsub wired correct
* Switches set correct
So we need to find the original manual. The later versions do not include component values.
Combined newer manual
M15 Option Description
Original Manual
(was linked in this thread, I dont see it anymore)
All of these are saved off at SchindlerEngineering.com
-methods
methods
1 GW
methods
1 GW
methods
1 GW
Hard Copies stowed away on SE
PDF Manuals
The links keep breaking (sigh)
My links dont break
-methods
PDF Manuals
The links keep breaking (sigh)
My links dont break
-methods
methods
1 GW
You see why you need the ORIGINAL manual???
Comes with Schematics!
Comes with Schematics!
methods
1 GW
Looking for shorted transistors and mosfets
Follow the lines off the high power resistor
Also
Look under board for hard short from particulate matter
Follow the lines off the high power resistor
Also
Look under board for hard short from particulate matter
methods
1 GW
Picking up there on Q8, Q9, Q11, Q12
methods
1 GW
Those were fine
Replaced the fuse
Smoke now coming out of R116
Check the Rectum Fryer
Replaced the fuse
Smoke now coming out of R116
Check the Rectum Fryer
methods
1 GW
Best DCS video EVAR!
He immediately points out that my smoking resistor is the precharge resistor... And that relay is the pre-charge relay! Duh!
That's a really helpful video. Watch it.
-methods
He immediately points out that my smoking resistor is the precharge resistor... And that relay is the pre-charge relay! Duh!
That's a really helpful video. Watch it.
-methods
methods
1 GW
He also points out that the fuse... Which was originally blown... Is the best one to remove to separate the two primary halves of the supply.
I usually try to solve things on my own...............
But that's usually because there's no help available. Since there's now help available, I'm going to take my time to study this video and improve my troubleshooting skills.
We don't want to get stuck in the YouTube loop watching other people instead of working ourselves - but - This is pretty complicated stuff and we should take any help we can get.
-methods
I usually try to solve things on my own...............
But that's usually because there's no help available. Since there's now help available, I'm going to take my time to study this video and improve my troubleshooting skills.
We don't want to get stuck in the YouTube loop watching other people instead of working ourselves - but - This is pretty complicated stuff and we should take any help we can get.
-methods
methods
1 GW
Disassembly showed a blast-thru to case on pins of T4. Those reside right next to P1, which is the plug for changing between 120V/240V. Bunch of video in the playlist if you want to see it. Recorded at 60fps HiDef, so if it looks blocky, turn up the resolution.
For that
I think the floppy PCBA just touched case thru the paper insulator while someone was jamming the plug in while energized.
Now moving on to other findings:
Smoking Resistor next to the relay
R116 was found to be a PreCharge resistor, smokes when it cant bring up the rail, removing F7 relieves that leading to further testing
Burning Smell after removing F3
That was explained away by the main bleeders across the 1.8mF caps, nothing is burning. All the DC rails on the board are coming up, including the isolated 6.2V
*************************************
At this point I have the main board out of the housing. Housing appears to have drank salt water into the inlet fan area (sigh).
*************************************
Cooked 10W resistor R98
Still not explained, getting to that now/soon
Main Power Transistors
First test showed case is drain and drain/source are short. Off to go look those up now to confirm. Looks like they are screwed in then soldered over. Looks by the insulating washers that the aluminum anodized rail is hot to drain (hot rail, very common)
************
Youtube Playlist updated with a full teardown of the unit, removal of the main board.
* Some spots there were 3 different length fasteners in same diameter!!@#@
* Some spots had triple washer, one being Insulator
* Most fasteners are Washer + LockWasher, SUPER easy to lose one in there
Must be methodical while you disassemble!
*************
-methods
For that
I think the floppy PCBA just touched case thru the paper insulator while someone was jamming the plug in while energized.
Now moving on to other findings:
Smoking Resistor next to the relay
R116 was found to be a PreCharge resistor, smokes when it cant bring up the rail, removing F7 relieves that leading to further testing
Burning Smell after removing F3
That was explained away by the main bleeders across the 1.8mF caps, nothing is burning. All the DC rails on the board are coming up, including the isolated 6.2V
*************************************
At this point I have the main board out of the housing. Housing appears to have drank salt water into the inlet fan area (sigh).
*************************************
Cooked 10W resistor R98
Still not explained, getting to that now/soon
Main Power Transistors
First test showed case is drain and drain/source are short. Off to go look those up now to confirm. Looks like they are screwed in then soldered over. Looks by the insulating washers that the aluminum anodized rail is hot to drain (hot rail, very common)
************
Youtube Playlist updated with a full teardown of the unit, removal of the main board.
* Some spots there were 3 different length fasteners in same diameter!!@#@
* Some spots had triple washer, one being Insulator
* Most fasteners are Washer + LockWasher, SUPER easy to lose one in there
Must be methodical while you disassemble!
*************
-methods
methods
1 GW
Updated Playlist
Sorensen DCS 33-33M15 teardown, diagnosis, repair
-methods
Sorensen DCS 33-33M15 teardown, diagnosis, repair
-methods
methods
1 GW
methods
1 GW
Original HexFets, ooooowld School
www.digikey.com
500V 12A
N-Channel
Obsolete
ACK - pdf posted not the datasheet, just PN reference
Actual data sheet off Google
IR bought by Infinion
NOTE: "Repetitive Avalanche Rated" - thats pretty boss. Means you are allowed to slam it into conductance by over-volting the drain to source. Only seen that once in a recent design.
IRF450 | DigiKey Electronics
Order today, ships today. IRF450 – N-Channel 500 V 12A (Tc) 150W (Tc) Through Hole TO-204AA (TO-3) from Infineon Technologies. Pricing and Availability on millions of electronic components from Digi-Key Electronics.
500V 12A
N-Channel
Obsolete
ACK - pdf posted not the datasheet, just PN reference
Actual data sheet off Google
IR bought by Infinion
NOTE: "Repetitive Avalanche Rated" - thats pretty boss. Means you are allowed to slam it into conductance by over-volting the drain to source. Only seen that once in a recent design.
methods
1 GW
Ok, so picture is a screen capture of bottom of PCB. Confirmed the small trace is Gate, the larger trace is Source, and the big mounting buggers are Drain.
Original guess was wrong, they are a pair of totem poles.
All we have to do now is measure Drain to Source on any of them. All appear to be SHORT which... er... I better go measure again.
Hold up
Original guess was wrong, they are a pair of totem poles.
All we have to do now is measure Drain to Source on any of them. All appear to be SHORT which... er... I better go measure again.
Hold up
methods
1 GW
Went back and measured
Two of them have 11ohms from Drain to Source
Two of them have 22ohms from Drain to Source
They are "in circuit" so we have to see if that makes sense (probably not). I am going on assumption that they rest OPEN and only go to SHORT when sufficient voltage is applied gate to source (like 10V). Assuming normal N-Channel mosfet, like the one we use in Ebike Controller
This one
It drives the transformer like 1 coil of the hub motor. Same simple totem pole design. Turn on the top, puts hot on coil. Turn on the other one bottom, puts ground, coil energized. THEN Turn on the bottom, drives ground to coil, turn on the other top, energizes coil in reverse direction. So this one can go bi-polar BANG BANG BANG
Thats about all I know, going to sleep now
* Check the Gate Drive Circuits
* Calculate the Drain-Source impedance OR remove from circuit and measure
Likely it only takes one or two to give readings we saw. Maybe we see coil input resistance eh? We can look that up.
-methods
Two of them have 11ohms from Drain to Source
Two of them have 22ohms from Drain to Source
They are "in circuit" so we have to see if that makes sense (probably not). I am going on assumption that they rest OPEN and only go to SHORT when sufficient voltage is applied gate to source (like 10V). Assuming normal N-Channel mosfet, like the one we use in Ebike Controller
This one
It drives the transformer like 1 coil of the hub motor. Same simple totem pole design. Turn on the top, puts hot on coil. Turn on the other one bottom, puts ground, coil energized. THEN Turn on the bottom, drives ground to coil, turn on the other top, energizes coil in reverse direction. So this one can go bi-polar BANG BANG BANG
Thats about all I know, going to sleep now
* Check the Gate Drive Circuits
* Calculate the Drain-Source impedance OR remove from circuit and measure
Likely it only takes one or two to give readings we saw. Maybe we see coil input resistance eh? We can look that up.
-methods
methods
1 GW
Corrected pinout for measurements taken
methods
1 GW
Back to the DCS60-18EM52
This is the 60V 18A supply that is limped by the factory to only 5A. I never solved that.
Using this Revision F manual, latest is L, best is the early blurry stuff.... but those are different than the E series.
BrainStorm
They have a theory of operation. Reading that leads one to think it is not as simple as swapping the Shunt Value. I will stuff some of that below
I think before we came to this same conclusion... since the front panel reads 5A correctly. I am going on the assumption that they did the MINIMUM work possible to convert a bone-stock 18A supply to a limped 5A supply. It is undoubtedly going to be something as stupid as a single voltage reference, potentiometer, voltage divider, etc. ... Just something scaled down by a factor of 18/5.
I think last go we actually checked R91 and it was identical in hardware to the Manual. Lets check again quickly.
...
...
Bastards! The E-Series manuals do not have the same level of detail as the earlier manuals. Thanks Profiteers, for that fine stroke of genius. I am certain some money-hungry fool in a department other than Engineering had the grand idea to strip information from the documentation in order to........ Whatever.
We still do not have a standard DCS60-18E to compare side by side. I have only, in my hands...
DCS60-18EM52 (working, the one we speak of here)
DCS600-1.7 (working, but the older non-e model)
DCS33-33 (in the middle of a fet swap and non-e model)
A magnifying glass and patience will show the difference **
I guess we can attempt to Zoom In on the problem area. Another fighter for the revolution posted his findings on YouTube and a blog. I will cross-link the blog here, as I have asked a question there in hopes of glory. I offered to give away free my GPIB option and/or my Isolated Programming option to the first person who helps me get 18A out of my M52 option.
Here it is
lazyelectrons.wordpress.com
That guy is #1 friend of the revolution. Pretty much the only real "solve" of the DCS posted publicly where it can be found via google. I am sure other solves exist, but they are either poorly indexed, gone, extinct, private knowledge, in paper only, ... or maybe just on page 999 of search results.
Ok... lets see if his is E-Series or Earlier
OH YES... Ok, lets look at his pictures
This is the 60V 18A supply that is limped by the factory to only 5A. I never solved that.
Using this Revision F manual, latest is L, best is the early blurry stuff.... but those are different than the E series.
BrainStorm
They have a theory of operation. Reading that leads one to think it is not as simple as swapping the Shunt Value. I will stuff some of that below
I think before we came to this same conclusion... since the front panel reads 5A correctly. I am going on the assumption that they did the MINIMUM work possible to convert a bone-stock 18A supply to a limped 5A supply. It is undoubtedly going to be something as stupid as a single voltage reference, potentiometer, voltage divider, etc. ... Just something scaled down by a factor of 18/5.
I think last go we actually checked R91 and it was identical in hardware to the Manual. Lets check again quickly.
...
...
Bastards! The E-Series manuals do not have the same level of detail as the earlier manuals. Thanks Profiteers, for that fine stroke of genius. I am certain some money-hungry fool in a department other than Engineering had the grand idea to strip information from the documentation in order to........ Whatever.
We still do not have a standard DCS60-18E to compare side by side. I have only, in my hands...
DCS60-18EM52 (working, the one we speak of here)
DCS600-1.7 (working, but the older non-e model)
DCS33-33 (in the middle of a fet swap and non-e model)
A magnifying glass and patience will show the difference **
I guess we can attempt to Zoom In on the problem area. Another fighter for the revolution posted his findings on YouTube and a blog. I will cross-link the blog here, as I have asked a question there in hopes of glory. I offered to give away free my GPIB option and/or my Isolated Programming option to the first person who helps me get 18A out of my M52 option.
Here it is
Sorensen DCS 600-1.7E – 0-600V, 1.7A DC Power Supply Repair
This switching power supply came along with me from a flea market. Lost, and lonely and being neglected by owner as it had many burned components and a big label on it ” DO NOT POWER UP…
lazyelectrons.wordpress.com
That guy is #1 friend of the revolution. Pretty much the only real "solve" of the DCS posted publicly where it can be found via google. I am sure other solves exist, but they are either poorly indexed, gone, extinct, private knowledge, in paper only, ... or maybe just on page 999 of search results.
Ok... lets see if his is E-Series or Earlier
OH YES... Ok, lets look at his pictures
methods
1 GW
So if we go forward on this logic:
* They would not just change the shunt, since that feeds the meter display
* They would change the feedback circuit
So we know the feedback circuit must have lots of claptrap
* It has natural limits (ABSMAX limits)
* It has input from the current control knob
* It can be programmed via Analog off the back
* It can be programmed via GPIB
* etc...
(I remember going thru all this before, oh well, here we go again)
They could have intercepted the signal from either the Front Panel or the Rear panel (reading page 3-6...) by simply adding a voltage divider or buffered divider in there. We should be able to detect that, and pretty sure we did. They would have to hack both the front and rear inputs as I tested...
* Front panel goes 0-5A
* Rear panel goes 0-5A
I have a video around where I was trying to solve this (years ago) by reconfiguring the DB25 for external control and feeding in an Arduino 0-5V in PWM. I was able to still control Voltage and IIRC I was not able to run current up beyond 5A *
So
If there were a simple 18/5 hack, it would have to touch both those controls if done in that way
... sigh
A current shunt (R91) in the output return line develops a voltage proportional to the outputcurrent. This current information is amplified by U15 and compared to the setting of the currentlimit control in the current control error amplifier U4-1.
Now pulling from revision L
R31, lets go find that
PAUSE
ARG
Ok, I forgot some
* Set current via front panel
* Set current via Potentiometer
* Set current via Current Source
* Set current via Voltage Source
* Set current via GPIB
* Set current via Isolated....
So all that garbage - they would not want to fuss around in permutation hell. I think they would do it right where all those things come together.
eh hem
....
-methods
* They would not just change the shunt, since that feeds the meter display
* They would change the feedback circuit
So we know the feedback circuit must have lots of claptrap
* It has natural limits (ABSMAX limits)
* It has input from the current control knob
* It can be programmed via Analog off the back
* It can be programmed via GPIB
* etc...
(I remember going thru all this before, oh well, here we go again)
They could have intercepted the signal from either the Front Panel or the Rear panel (reading page 3-6...) by simply adding a voltage divider or buffered divider in there. We should be able to detect that, and pretty sure we did. They would have to hack both the front and rear inputs as I tested...
* Front panel goes 0-5A
* Rear panel goes 0-5A
I have a video around where I was trying to solve this (years ago) by reconfiguring the DB25 for external control and feeding in an Arduino 0-5V in PWM. I was able to still control Voltage and IIRC I was not able to run current up beyond 5A *
So
If there were a simple 18/5 hack, it would have to touch both those controls if done in that way
... sigh
A current shunt (R91) in the output return line develops a voltage proportional to the outputcurrent. This current information is amplified by U15 and compared to the setting of the currentlimit control in the current control error amplifier U4-1.
Now pulling from revision L
R31, lets go find that
PAUSE
ARG
Ok, I forgot some
* Set current via front panel
* Set current via Potentiometer
* Set current via Current Source
* Set current via Voltage Source
* Set current via GPIB
* Set current via Isolated....
So all that garbage - they would not want to fuss around in permutation hell. I think they would do it right where all those things come together.
eh hem
....
-methods
methods
1 GW
So without understanding the OpAmps yet (they are explained) we know that any of the 4 resistors co-located with that can be substituted for a value which will impact/shift/limit the range of that pot.
I suppose Offset could be it as well, at the bottom middle.
Trouble is, those really are trim pots (in my experience turning them...) and the gross setting is hard-fixed. So... looks like we have to actually think the circuit thru (which is what I generally avoid)
hahahaha
Sometimes you CANT - like USUALLY. Only in this weird case did an MFG supply Schematic and BoM for us to practice our skills... so I go at it like I would anyway - Blind as a bat and hoping for lucky. Known-Good Swap and the Compare Method.
-methods
methods
1 GW
Lets circle every resistor and Active Component mentioned in the prose and call those out so we can have visual on the schematic. Yea, Engineer in good practice can hold both in his head at once... I am not in good practice.... so I have to translate the paragraphs into graphics and see it in one spot.
Starting with this mess
Garble-Magic-No-Read_special
-methods
Starting with this mess
Garble-Magic-No-Read_special
-methods
