Weird current limiting fault on an Instek PSP2010

[philpem]

Right, here's a weird one...

I've got a pair of Instek PSP2010s on the bench with more or less the same fault. With the current limit set below 2A or so, they fail to correctly regulate. The output voltage ripples like crazy, even though the output is completely unloaded. Voltage regulation works fine, as long as the current limit is 2A or higher. I have no idea what damaged one of the power supplies (it's had this fault since I got it), however the second unit was damaged after I accidentally connected a battery desulphator to the output at the same time as the battery. Turns out the battery was open-circuit. Fifty ish volt spikes up the power supply's output terminals.    (As Lance-Corporal Jones is oft to say, "they don't like it up 'em!")
I've a sneaking suspicion the first supply may have suffered a similar fate. Perhaps at the hands of being used to test a faulty switch-mode PSU or some other inductive load.

This power supply is widely rebranded by Voltcraft as the DPS-2010 and DPS-2010PFC, and there are doubtless a dozen or so other rebrands. The DPS-4005 and DPS-4005PFC are close twins, with different voltage ratings (40V/5A instead of 20V/10A).

I found the DPS-4005PFC schematics at http://list.hw.cz/pipermail/hw-list/attachments/20050719/c6530da3/attachment-0003.pdf but haven't found anything for the 2010. An Instek dealer offered to send me a partial schematic (input circuitry only) under NDA, but only for part of the power board, and with all the component values removed. Needless to say, I politely declined.

[ EDIT -- there's a better version of part of the schematic here: http://www.mikrocontroller.net/attachment/146656/512020-sp-01-en-Switching_Power_Supply_DPS-4005PFC.pdf ]

I've figured out enough of the schematic to tell that it's a fairly standard TL494 switch-mode converter, but I can't make head nor tail of the current sensing and regulation side. Looking at page 7, my instinct is that it's using Q7 as a sensing component, V27 thru V34 are doing some form of biasing on Q7 (but how and what eludes me) and the current sense signal somehow creeps out onto pin 3 of X6A. I guess N5B is the current regulation loop?

N7's function also eludes me... some form of servo loop, but clamping its own input? Perhaps an overcurrent limiter of some kind?

Anyway, there's no theory of operation, no documentation of any kind... Can anyone shed some light on this thing and maybe suggest how it might work, or some components which might be worth testing?

I'm thinking N5B would be worth a check for a start (it's clearly something to do with current regulation; ICOM is "current commanded", IF is "current feedback", so it's comparing the two and driving V24 appropriately... I think) along with the current sensing circuitry (if I ever figure out how it works...)

Thanks,
Phil.

[Andy Watson]

I think you have to factor in the very real possibility that the circuit diagram is wrong. I believe the accuracy of the schematic has already been questioned on this very forum!

I would say that N7 is some form of clamp - except that the inputs to the opamp are reversed. Consider that the top of R89 goes to +12V and the top of R90 is connected to the sense wire. The circuit would now operate as a voltage clamp - albeit at an operating point of 60-70V.

There are (at least)  two control loops in operation, one will be controlling the switching regulator and the other is controlling the fine, linear regulation of the output - I assume it's the latter that is of interest? Q7 is the main series pass element - controlled via V34, V33 and N5A.  N5A controls the voltage, via feedback through R60, and the current via R62 - both compared with the demand which presumably is derived via R58 and what ever connects to X6A,1.

The current is sensed in R117. This voltage is amplified by N6A - Note, I am assuming that the schematic is again wrong in that the bottom end of R71 should be connected to the negative input of N6A.

Edit: The above gives a fast response to the current limit. The controlled/variable current limit is obtained by further amplification of the output of N6A, via N6B. This signal goes via X6A,3 to a monitor, and is also compared to the demanded current on X6A,2 by N5B. The presence of C27 and V25 suggested that this part of the control is intended to respond slower to the peak current output and it responds by reducing voltage demand via R58 and by partially shutting-down the switching regulator drive.

[philpem]

I think you have to factor in the very real possibility that the circuit diagram is wrong. I believe the accuracy of the schematic has already been questioned on this very forum!

Oh wonderful :-/

There are (at least)  two control loops in operation, one will be controlling the switching regulator and the other is controlling the fine, linear regulation of the output - I assume it's the latter that is of interest? Q7 is the main series pass element - controlled via V34, V33 and N5A.  N5A controls the voltage, via feedback through R60, and the current via R62 - both compared with the demand which presumably is derived via R58 and what ever connects to X6A,1.

The current is sensed in R117. This voltage is amplified by N6A - Note, I am assuming that the schematic is again wrong in that the bottom end of R71 should be connected to the negative input of N6A.

D'oh! I completely missed the existence of R117...

Edit: The above gives a fast response to the current limit. The controlled/variable current limit is obtained by further amplification of the output of N6A, via N6B. This signal goes via X6A,3 to a monitor, and is also compared to the demanded current on X6A,2 by N5B. The presence of C27 and V25 suggested that this part of the control is intended to respond slower to the peak current output and it responds by reducing voltage demand via R58 and by partially shutting-down the switching regulator drive.

That makes sense -- switch-mode gets rid of the bulk of the voltage, linear handles the fast-response current limiting...

Thanks for the explanation. I think I'll have to look at the schematic again and see if I can come up with any possible failure causes for the weird "anything less than 2A has tons of ripple" fault. I'm suspicious of the FET as that's right in the output path, though it could equally be one of the opamps or transistors.

Some more head-scratching required...

thanks,
Phil.

[Ordinaryman1971]

Hi guys, I have the same power supply with a bit different problem. When I turn the power on it just clicks.. the backlight on the display dims out a bit and it clicks again.. all the controls are disabled, the screen is "locked" and I can't do anything.... any clues. Where do I start troubleshooting.

[mahmut47]

hi guys i have the same problem when i turn the power on it just clicks and restarts again and again. where is the problem i cant find it please someone help me

[pegase1985]

I got PSP-2010 and m'y problem is that it isn't accurate anymore, when i set 1,39volt output for instance I do really  got 1,79volt instead at the output. Do you guys have an Idea of the problem ?