methods
1 GW
Since I have not contributed much of value recently - here is some value.
Purpose: To allow enthusiasts to buy used/damaged or (otherwise in need of repair) DCS supplies for charging and general geek use.
Background: Sorensen DCS supplies are expensive, high quality, rack mount, CV/CC Variable Lab Supplies. They run off of 120V or 240V single phase (an important point!) and they can be had for 1/10th the new cost on Ebay.
* They can stack in Series or Parallel (Very important point!)
* They can be remotely controlled via Analog (via Arduino) or GPIB etc (GPIB variants cost significantly more)
* They are laid out flat, very easy to reverse engineer, and come with a parts replacement list
* They have remote sense, remote interlock, current control, analog control, remote display, etc - all standard
I have rigged one of these supplies up with an Arduino to control and read back - it was trivial...
To work with: I have the following 1U supplies in hand, as well as a list of experience with other flavors
In Hand
DCS 60-18
DCS 60-18E
DCS 600-1.7
Experience with
DCS 8-125E
DCS 50-20E
DCS 300-3.5E
And in the 3KW variant 2U @ only 240V
DCS 8-350E
DCS 80-37E
Datasheet Links
http://www.programmablepower.com/dc-power-supply/DCS/Downloads.htm
...
Focusing in on what I have in hand (so what we can troubleshoot)
DCS is the series name
First number is the Voltage Limit
Second number is the Current Limit
(The later models have an E suffix)
The 1U (very carry-able) variant can provide either 1KW or 1.2KW (on either 120 or 240)
The 2U (Very heavy) variant can provide 3KW on single phase 240 (which can be made from 2pcs 120V outlet Single Phase...)
DCS 60-18E
This is my personal supply and the one I use to charge Ebikes with. I paid apx $350 shipped from Ebay. It arrived in good working order. Under heavy load it "squeals"... which is something we want to understand and repair.
DCS 60-18
This is the property of ChargePoint. It was purchased as part of a lot for Test Engineering Lab Use (bringup - not line test) and cost only $250. It has a bad fan and that is the most common failure and easiest to fix failure.
DCS 600-1.7
This is the property of ChargePoint. It was purchased for the same as the above and costs under $600 (Check pricing on HV supplies... they are expensive and hard to find... ESPECIALLY in a 120V flavor). This supply demonstrates a new failure mode which I have not encountered before: It can reach 500V... it can produce 1.7A... but not at the same time. I can reliably get about 420V at a few hundred mA out of it. This will also be an easy repair.
The most common out of the box failure modes are as follows:
EDIT:
Dry Caps
There are two very large 1800uF 200V caps wired in Series (on the 600V units)
These serve as voltage doublers for the AC input while the supply is wired in 120VAC mode
They serve as the filtering caps (for both 120V and 240V mode), and feed the chopper circuit.
With 120VAC in (and the selection P1 in either location) a failed supply will read apx 150VDC on the P2 jumper (across the stack). You should be reading more like 340VDC (RMS IN / 0.707, or Peak not RMS). 150V is the peak out a 120VAC RMS input*
In 240V mode the supply would reach 600V and 1.7A but NOT AT THE SAME TIME.
In 120V mode the supply would not reach 600V, but could reach 1.7A, But produced only about 100W
Dead Fan
This is the most likely... You can hear the fan clearly and the supply will work but overheat quickly. it is a few fasteners and 2 wires to replace the fan. The fan can be had for apx $20... or you can rig up whatever you have got around if you MUST. (I suggest doing it up back to factory)
* Red Light
If the Dsub25 "programming jumper" is not inserted into the rear of the supply it will not operate. There are a handful of jumpers that MUST be connected for the supply to work turn-key. The jumpers tell the supply things like the following:
- Front panel control or remote control
- Interlocked or not interlocked
The manual is quite explicit so I will not re-tread that information here. You just read the list of 25 pins and follow the directions as to which pins to short.
* The internal 240/120 jumpers are not set correctly
In this case, if internal is set to 120V, and you apply 240V... smoke can be seen. Make sure you see where it comes from.
If the internals are set to 240 and you apply 120 you will see a "weak response"
* Switch Settings Incorrect
There is a standard 8 leg rocker that needs to be set to a default
The manual covers it in detail
Squealing
This is supper common and ... appears... to not affect its output. I am sure that if you looked on the scope you would see the squeal - but for lipo charging, it has not affected me. I am MOST interested in solving this particular riddle.
V and A, but no Power
This one is new to me. I am certain that it is another super easy fix.
Slow Ramp Down - cooked sugar load
These supplies have great transient response. The price of that it some output loading... SHOULD you wire up a 50V or 120V lipo pack, then turn off the supply, the battery voltage back-feeds into the output impedance and smokes off the pull-down. I call that the sugar load... and it does not impact the supplies usefulness for charging or lab use (it just slows it down a bit). We want to identify which part is smoking off - buy a batch of them - and set up a "Sorensen Repair Kit".
Dead Bands in the dials
Unconfirmed - but I swear that while bringing one up for Boran I saw a dead band in the dial. It was weird... like a nick in the pot... allowing the voltage to momentarily dip (blip dip). Problematic possibly - would need to scope the output. They are 10 turn pots and could EASILY be swapped out.
General Cruft Buildup
Most of these supplies are from the 90's. They go into a rack, they get turned on, they stay turned on (or cycle once a day). The caps should all be shot out by now - but for what I want to do with them (charging - not dynamic response to changing load) they are fine so far and I have not had to replace any caps.
...
For charging - what we NEED - is an Ideal Diode.
Nick Long produces these and I have an open invoice to pick up another 50 of them (hopefully he still has them)
They are good to 100VDC and upward of 20A if done right
They have almost no forward drop, you can wire the sense wires downstream of them, so they have no impact on performance
They WILL stop your Lipo Pack from thrashing your supply should the breaker blow or should the supply be turned off before removing the active load.
A schottky diode will NOT SUFFICE. Even the huge variants get SUPER HOT and dump TONS OF HEAT.
Example:
100mV forward Voltage
20A forward current
V*I=W
100mV * 20A = 2W
Now... 2W does not SOUND like much... but it stacks up quickly... and requires a fan or major heat sinking to keep below 60C. For reference... most balance resistors see:
4.2V / 50mA = 84ohms
4.2V * 0.05A = 200mW
200mW requires an 800mW resistor MINIMUM - better to use a 2W resistor
Even at that tiny power you will find parts skyrocketing in temp.
(This is the sort of stuff that GreenHorn engineers do not understand. After they get done re-working their spice design... they then get to learn of the second order effects of major thermal stressing over time <fatigue failure>). Poor engineering is usually answered with fans... which result in single point failures involving fire and smoke... so... pay attention to what you are doing and always shoot for mill spec. (Mill Spec means all day, every day, working with multiple failures)
...
Good place to stop.
I am not going to fill this thread up with fancy pictures and half-wit remarks. I am going to use it as my notes section while I troubleshoot and repair these supplies.
Go buy one... and come along for the ride.
-methods
Purpose: To allow enthusiasts to buy used/damaged or (otherwise in need of repair) DCS supplies for charging and general geek use.
Background: Sorensen DCS supplies are expensive, high quality, rack mount, CV/CC Variable Lab Supplies. They run off of 120V or 240V single phase (an important point!) and they can be had for 1/10th the new cost on Ebay.
* They can stack in Series or Parallel (Very important point!)
* They can be remotely controlled via Analog (via Arduino) or GPIB etc (GPIB variants cost significantly more)
* They are laid out flat, very easy to reverse engineer, and come with a parts replacement list
* They have remote sense, remote interlock, current control, analog control, remote display, etc - all standard
I have rigged one of these supplies up with an Arduino to control and read back - it was trivial...
To work with: I have the following 1U supplies in hand, as well as a list of experience with other flavors
In Hand
DCS 60-18
DCS 60-18E
DCS 600-1.7
Experience with
DCS 8-125E
DCS 50-20E
DCS 300-3.5E
And in the 3KW variant 2U @ only 240V
DCS 8-350E
DCS 80-37E
Datasheet Links
http://www.programmablepower.com/dc-power-supply/DCS/Downloads.htm
...
Focusing in on what I have in hand (so what we can troubleshoot)
DCS is the series name
First number is the Voltage Limit
Second number is the Current Limit
(The later models have an E suffix)
The 1U (very carry-able) variant can provide either 1KW or 1.2KW (on either 120 or 240)
The 2U (Very heavy) variant can provide 3KW on single phase 240 (which can be made from 2pcs 120V outlet Single Phase...)
DCS 60-18E
This is my personal supply and the one I use to charge Ebikes with. I paid apx $350 shipped from Ebay. It arrived in good working order. Under heavy load it "squeals"... which is something we want to understand and repair.
DCS 60-18
This is the property of ChargePoint. It was purchased as part of a lot for Test Engineering Lab Use (bringup - not line test) and cost only $250. It has a bad fan and that is the most common failure and easiest to fix failure.
DCS 600-1.7
This is the property of ChargePoint. It was purchased for the same as the above and costs under $600 (Check pricing on HV supplies... they are expensive and hard to find... ESPECIALLY in a 120V flavor). This supply demonstrates a new failure mode which I have not encountered before: It can reach 500V... it can produce 1.7A... but not at the same time. I can reliably get about 420V at a few hundred mA out of it. This will also be an easy repair.
The most common out of the box failure modes are as follows:
EDIT:
Dry Caps
There are two very large 1800uF 200V caps wired in Series (on the 600V units)
These serve as voltage doublers for the AC input while the supply is wired in 120VAC mode
They serve as the filtering caps (for both 120V and 240V mode), and feed the chopper circuit.
With 120VAC in (and the selection P1 in either location) a failed supply will read apx 150VDC on the P2 jumper (across the stack). You should be reading more like 340VDC (RMS IN / 0.707, or Peak not RMS). 150V is the peak out a 120VAC RMS input*
In 240V mode the supply would reach 600V and 1.7A but NOT AT THE SAME TIME.
In 120V mode the supply would not reach 600V, but could reach 1.7A, But produced only about 100W
Dead Fan
This is the most likely... You can hear the fan clearly and the supply will work but overheat quickly. it is a few fasteners and 2 wires to replace the fan. The fan can be had for apx $20... or you can rig up whatever you have got around if you MUST. (I suggest doing it up back to factory)
* Red Light
If the Dsub25 "programming jumper" is not inserted into the rear of the supply it will not operate. There are a handful of jumpers that MUST be connected for the supply to work turn-key. The jumpers tell the supply things like the following:
- Front panel control or remote control
- Interlocked or not interlocked
The manual is quite explicit so I will not re-tread that information here. You just read the list of 25 pins and follow the directions as to which pins to short.
* The internal 240/120 jumpers are not set correctly
In this case, if internal is set to 120V, and you apply 240V... smoke can be seen. Make sure you see where it comes from.
If the internals are set to 240 and you apply 120 you will see a "weak response"
* Switch Settings Incorrect
There is a standard 8 leg rocker that needs to be set to a default
The manual covers it in detail
Squealing
This is supper common and ... appears... to not affect its output. I am sure that if you looked on the scope you would see the squeal - but for lipo charging, it has not affected me. I am MOST interested in solving this particular riddle.
V and A, but no Power
This one is new to me. I am certain that it is another super easy fix.
Slow Ramp Down - cooked sugar load
These supplies have great transient response. The price of that it some output loading... SHOULD you wire up a 50V or 120V lipo pack, then turn off the supply, the battery voltage back-feeds into the output impedance and smokes off the pull-down. I call that the sugar load... and it does not impact the supplies usefulness for charging or lab use (it just slows it down a bit). We want to identify which part is smoking off - buy a batch of them - and set up a "Sorensen Repair Kit".
Dead Bands in the dials
Unconfirmed - but I swear that while bringing one up for Boran I saw a dead band in the dial. It was weird... like a nick in the pot... allowing the voltage to momentarily dip (blip dip). Problematic possibly - would need to scope the output. They are 10 turn pots and could EASILY be swapped out.
General Cruft Buildup
Most of these supplies are from the 90's. They go into a rack, they get turned on, they stay turned on (or cycle once a day). The caps should all be shot out by now - but for what I want to do with them (charging - not dynamic response to changing load) they are fine so far and I have not had to replace any caps.
...
For charging - what we NEED - is an Ideal Diode.
Nick Long produces these and I have an open invoice to pick up another 50 of them (hopefully he still has them)
They are good to 100VDC and upward of 20A if done right
They have almost no forward drop, you can wire the sense wires downstream of them, so they have no impact on performance
They WILL stop your Lipo Pack from thrashing your supply should the breaker blow or should the supply be turned off before removing the active load.
A schottky diode will NOT SUFFICE. Even the huge variants get SUPER HOT and dump TONS OF HEAT.
Example:
100mV forward Voltage
20A forward current
V*I=W
100mV * 20A = 2W
Now... 2W does not SOUND like much... but it stacks up quickly... and requires a fan or major heat sinking to keep below 60C. For reference... most balance resistors see:
4.2V / 50mA = 84ohms
4.2V * 0.05A = 200mW
200mW requires an 800mW resistor MINIMUM - better to use a 2W resistor
Even at that tiny power you will find parts skyrocketing in temp.
(This is the sort of stuff that GreenHorn engineers do not understand. After they get done re-working their spice design... they then get to learn of the second order effects of major thermal stressing over time <fatigue failure>). Poor engineering is usually answered with fans... which result in single point failures involving fire and smoke... so... pay attention to what you are doing and always shoot for mill spec. (Mill Spec means all day, every day, working with multiple failures)
...
Good place to stop.
I am not going to fill this thread up with fancy pictures and half-wit remarks. I am going to use it as my notes section while I troubleshoot and repair these supplies.
Go buy one... and come along for the ride.
-methods