Manson DPD-3030 PS - modifications possible?

[MarvinTheMartian]

Hi all,

I recently purchased a Manson DPD-3030 power supply (see Shock's thread for photos - https://www.eevblog.com/forum/reviews/photos-manson-dpd-3030-triple-output-power-supply/msg664001/#msg664001 ).

This PS has single turn pots (very touchy and difficult to adjust settings accurately) and does not have a 'Load On/Off' switch.

Could someone (Shock perhaps has already done this?) advise me on a couple of questions please?

• There seems to be plenty of room inside to replace the one turn pots for ten turn pots - are there any traps to doing this or is it simply replace the old with the new?
  
  
• To put in Load On/Off switches for each output (or all outputs?). For individual switches - is it as simple as putting a switch in between the output jacks and the wires connected to them or are there issues in doing this? For one switch to control them all - how would I achieve that?

I'm pretty new to working out how circuits work so I am unsure if what I am proposing is a good idea or it will just ruin my new PS! 

Thanks everyone. 

[ElektronikLabor]

1. There seems to be plenty of room inside to replace the one turn pots for ten turn pots - are there any traps to doing this or is it simply replace the old with the new?

I don't see any traps if the pot has the right resistance value and mechanicaly fits into the holes

2. To put in Load On/Off switches for each output (or all outputs?). For individual switches - is it as simple as putting a switch in between the output jacks and the wires connected to them or are there issues in doing this? For one switch to control them all - how would I achieve that?

If a PSU has an On/Off switch its nearly allways for all outputs together. I could imagine it's because it's cheaper 
I would also fit the switch between the wire and output jack, because that should not affect the voltage reading of the display.
But at the moment I can hardly imagine a reason why to switch the channels separately.

To switch all channels together you need a two positional switch with three isolated inputs which can handle the current

[MarvinTheMartian]

Thanks Medvedev. I might go ahead with that. I remember watching one of Dave's videos where he says it is good to have the load switch instead of just the power switch as the output would be unstable initially and may blow your circuit.  So the motto is don't connect your circuit if you use the power switch to avoid damage! 

Really appreciate the great information in Dave's videos and everyone on the forum (which I've only just started using). 

[Shock]

No haven't modified mine yet as I've had other things keeping me busy. When it comes to switching the load on and off, a few things about the behavior of the power supply dictates how you would need to approach this.

Firstly I think that the DPD-3030 would benefit from a simple switch on each channel shorting the outputs to allow you to set the constant current. Obviously when you automate this you would need to measure for voltage spikes/surges so you could observe if switching with a circuit load connected could damage the circuit. But in order not to present a shorted or a high current path to the circuits positive and negative rails the switch would need isolate the shorted path from the circuit at the same time. If you can do this cleanly this solves two of the problems.

The other issue is the DPD-3030 works in both single channel mode but also in parallel and series (doubling the current supply or the voltage supply). You would need to consider this before implementing any switching system to either prevent damage or functionality problems which is why I was looking at the possibility at doing this elsewhere or working out all the possible scenarios at least beforehand.

Swapping out the pots seems simple as it's just up to determining if drop in replacements are suitable for size, value, rating (important) and pins. A 10 turn pot may be an overkill perhaps 5 may be more suitable.

[MarvinTheMartian]

Thanks for the info Shock 

Firstly I think that the DPD-3030 would benefit from a simple switch on each channel shorting the outputs to allow you to set the constant current.

I understand what you're saying but when you set the current using the pots doesn't this set the maximum current delivery anyway (as if you had shorted the outputs)?
[I'm not that experienced yet  ]

Obviously when you automate this you would need to measure for voltage spikes/surges so you could observe if switching with a circuit load connected could damage the circuit. But in order not to present a shorted or a high current path to the circuits positive and negative rails the switch would need isolate the shorted path from the circuit at the same time. If you can do this cleanly this solves two of the problems.

Thanks for the tips.   

The other issue is the DPD-3030 works in both single channel mode but also in parallel and series (doubling the current supply or the voltage supply). You would need to consider this before implementing any switching system to either prevent damage or functionality problems which is why I was looking at the possibility at doing this elsewhere or working out all the possible scenarios at least beforehand.

Swapping out the pots seems simple as it's just up to determining if drop in replacements are suitable for size, value, rating (important) and pins. A 10 turn pot may be an overkill perhaps 5 may be more suitable.

Yes, I had considered the doubling of the voltage/current needed to be taken into account for the switch specs. I was also thinking that I may be swapping one hassle for another by using 10 turn pots (Dave had mentioned using these in one of his videos) but I wasn't aware there were also 5 turn pots - this sounds like a good middle ground - thanks!

[Shock]

I understand what you're saying but when you set the current using the pots doesn't this set the maximum current delivery anyway (as if you had shorted the outputs)?

You can but you cannot see what current limit you're setting without shorting the output (at least as far as I'm aware), it removes the voltage display as well which is a bit of a pain.

[MarvinTheMartian]

OK Thanks

[jobog]

I just bought a Tek Power 3005 power supply and I was thinking of adding switches to control the output the same as you are looking to do with your Manson.  I was having trouble finding switches to handle the 30V, 5A max output of the supply and adding to the problem was that I wanted lighted switchs to indicate when the output was switched on.  I think I have come up with a workable solution for very little cost but would like opinions as to the suitability of my solution.  I found a 2 channel DC 5V relay switch module for Arduino on Ebay for $3.98.  This module will handle 30V DC at 10A on both channels.  I also found 12V LED lighted switches on Ebay at 10 pcs for $7.99.  The switches are rated for 12V, 16A but the LED will light with as little as 5V DC.  My plan is to run the variable outputs of the supply to the relays and then to the front banana jacks and to tap into the fixed 5V supply in the Tek Power to the lighted switchs and then to the relay module.  The fixed 5V supply should light the LED and power the relays when the switches are switched on which will then direct the variable outputs to the front jacks.  My only concern is drilling 3/4" holes in the front panel to mount the switches.

Does this sound like a reasonable solution or is there something I am not seeing here?

[MarvinTheMartian]

Hi Jobog.

Sounds like a good plan but I'm not experienced enough to advise whether there are any issues with it.

I've haven't progressed any further with my modifications at this stage (been too busy) but I'd be interested to see how you progress. 

Hopefully someone on the forum will help you with advice.

Good luck! 

[Shock]

I just bought a Tek Power 3005 power supply and...

Sounds ok, but since you're going to that trouble you might want to look at the schematic of the power supply to see if there is an alternative way to switch the supply off before it comes to modifying it directly at the output. I suspect there is a far easier way to do it, something I will do on Manson before I attempt the brute force solution (when I get around to it). Also don't expect anything brought on Ebay from China to magically meet it's ratings, you have to do due diligence. 

I suggest if you are ready to perform the modification start a new thread on your plans and put as much info including photos and pdfs and a drawing of your modification and parts you're intending to buy so it makes it a no brainer for people to see exactly what you're doing and offer better advice if they see it.

[MarvinTheMartian]

... I suspect there is a far easier way to do it, something I will do on Manson before I attempt the brute force solution...

Thanks for helping out Shock.

If you find the 'easier way' before I get to my modification on the Manson could you let me know?

Many thanks.

PS Do you have a service manual/schematic for the Manson?

[Shock]

... I suspect there is a far easier way to do it, something I will do on Manson before I attempt the brute force solution...

Thanks for helping out Shock.
If you find the 'easier way' before I get to my modification on the Manson could you let me know?
Many thanks.
PS Do you have a service manual/schematic for the Manson?

Well not had time to look inside in any detail but I'll share my other idea for a hack. Put in an external output and shorting switch. Either break out the output to a switch box or inline cable switch.

Or even...

Drill 3 holes in a prototyping PCB (or make a PCB) the size of the outer thread on the 3 output (pos/neg/earth) banana sockets. Mount the switch to the PCB and one extra banana socket for the switched output. Attach the PCB so now you have all 3 connections as pass through straight out but one can be switch controlled if you wish. You would need to generously add solder to the used PCB traces to allow decent current flow between any connections.

Using a 3 way rocker you could have a short output position (to set current), you could also by adding more circuitry and a double throw switch make a CR2032 powered circuit to give switch status indication.

[MarvinTheMartian]

Interesting idea, hadn't considered that. You wouldn't have to make any internal changes at all then. 

Though I think i would still like to have it all built into the unit, but well worth considering if it turns out to be too messy or impractical to do so. 

I've got to go inside in any case to change the pots to 5 turn ones - I'll make a decision then on which way to go.

Thanks Shock. 

[beanflying]

Bump, it's mod time 

Not sure if either of you got around to modding your own?

Just got this wee beasty to go with my 15V 0-60A Manson (does this make me the Manson family  )

Still had the original stickers from when it was brought on top so very very clean. Did a little quick testing on the outputs regarding shorting switches for the front.

In Parallel mode only the Master Current set works to set both sides evenly. It doesn't care if it is master or slave shorted still only works on master.

In normal mode Mmmm well it operates "Normally"

In Series mode both Master and Slave can and should be set independently as per normal mode.

So spring loaded toggles or maybe heavy pushbuttons for the job on either channel would be the simple answer. BUT

As I would like to add is independent Channel switching with a default to off when first powered up so some sort of Soft Latch FET based circuit should do it. This will prevent any startup transients or spikes getting to any projects. So I am thinking the elegant solution will be a small daughter board mounted on the chassis between the display and the next board back. The nice thing is then very small and tidy pushbuttons can be used instead of big clunkers.

5 or 10 Turn pots on Voltage are a bit of a must do when I get some too.

Off to look for switches and Pots 

[Shock]

Go for it, yeah show us the circuit you decide to use and parts list etc with photos. I have two of these now I picked up another a few weeks back.

[beanflying]

Nice. Two is better than one 

Checked this one is fairly small diameter for a simple shorting switch with no other circuitry. 3A at 28V is a little underdone maybe but as the voltage drops to zero it should cope. http://au.element14.com/multicomp/r13-548b-05-bb/switch-push-button-spst-latching/dp/1634675

Mine turned up on Evilbay and looked lonely 

[Shock]

I brought one of these a few weeks back as well, was cheap and couldn't resist adopting it.

[beanflying]

Soft fet switch layout. There is room along the stripe at the back to drop switches in. From the left Voltage on/off LED and Current short to the right of the word. Repeated on the master side. The outer two buttons will fall just on the edge of the stripe not perfect but fairly well blended.

Seems someone made a video  Soft Latch switch

https://youtu.be/Foc9R0dC2iI

[beanflying]

Getting Naked 

Nice work the plugs to the display boards are labeled and the silk screen on the boards is clear part of the reason I love these supplies 

Voltage Pots are a 5K Linear taper 026T228F502A1A1 heaps of room behind for a multiturn so more about matching shafts or finding new knobs for the front.    http://pdf1.alldatasheet.com/datasheet-pdf/view/828829/CTS/026T228F502A1A1.html

** edit Current Set pots are a 501 and not a 502 like the Voltage set.

The Pot plates are held in by a single screw with a welded standoff to the front steel chassis. Tops of them are around the strip area, so a little creative metal work should see the switches and boards hanging on there too. This will leave plenty of vertical clearance on the display boards.

[Shock]

Are you going to put the circuit directly on or just before the output, or somewhere further back in the circuit?

I was considering (since the current resolution is milliohm accurate) an ideal inline switching circuit would need to have no current draw when both on and off loading the output, plus work in a low voltage scenario (the supply goes down quite accurately in voltage as well).

Or be further back in the circuit as to not effect the regulation and current limiting or be powered separately from the output to have no parasitic effect. You would need the inline switch e.g. FET or mechanical switch if loaded to handle the full voltage and current of the supply.

Dave circuit looks good on the scope but is this still suitable for a series 0-60VDC 3A or parallel 0-30VDC 6A supply?

So are you taking into consideration these problems?

Whats the control setup for the front you want to use, one momentary push button and two status LEDs?

[beanflying]

One LED only for voltage output on or off. Current set momentary short would do the job nicely so no need for an LED IMO

So default to off on startup. Push to turn outlet on and push again for off. Might pay to have it also reset on current set?

I need to do some more looking but there is a pair of fine wires going to the output boards that I suspect are sense wires go to the next board back near an 8 pin DIP chip (opamp??) that I haven't looked at. If they are then a tiny voltage drop on a fet before the final outlets won't be an issue. Set voltage unloaded should remain the same loaded?

As to accuracy I ran all the channels up against my Agilent today and on voltage were within 20mV across the range. Didn't load it so not sure of the current settings.

Should also note I am doing the nuts and bolts ahead of the electronic design while I have it stripped 

[beanflying]

Board behind the outlets. Shunts voltage relays and such. All through hole so it should be simple to hack. It may even be easier to hack into the relay circuit for the voltage control and just drive a coil?

[Shock]

This is how the front end is switched (relays) for series and parallel.

[beanflying]

Thanks I will have a good look tomorrow. Just worked out some spacing for the front panel. Silly  made the spacing 2.5mm wider on the right than on the left.

Anyway 2x8cm board is perfect. Just back one set of holes on the left gets 3mm padding from the button edge to the letters.

[Shock]

Both the master and slave supply (at the AC end) are two separate secondary windings each with multiple taps. The master and slave both each have an RL101 and RL102 relay (double throw to single pole) which selects from the taps and feeds into the AC input of the fullwave rectifier AC input. From there the DC positive rails go to two ganged series pass power transistors 2SD1975 the base of which are controlled by the voltage regulation circuit. Then it's current controlled and runs through the ammeter, voltmeter then relays to control the series and parallel (as shown in my previous post).

You could mod or switch off at the winding relays at the AC. Or turn off the power transistors at the base so the positive power rails have no supply. Or even switch off the DC anywhere, but the problem is unless I'm wrong the voltage displays right next to the output in the circuit. Switch off the voltage anywhere before that and you have no idea of what voltage you're enabling when the output is off.

Anyone else done this kind of thing feel free to chime in.