Test Equipment Anonymous

I previously noted that I bought a modern AIM-TTI PL303QMD-P power supply pleasingly cheaply. It promptly replaced my previous PSU on the shelf above the workbench.

I then went on to test several PSU's current limit response, and it performed well.

I had found that occasionally one of the outputs was not on as indicated, and that it turned on when I turned the output on and off again. No major problem, but an irritation. Over the weekend I was measuring its output capacitance with a different meter; sometimes it showed a plausible capacitance, sometimes zero. Touching the banana lead or the PSU's socket was sufficient to flip between the two; clearly this needed investigating. Perhaps that was why the meter was sold on Gumtree rather than fleabay.

So off the shelf it came, with me being surprised (again) that it didn't have a handle on the case's top. Took out the traditional screws on the rear panel, but the lid wouldn't budge. After a while I spotted these wretched plastic snap rivets on the side.
It wasn't difficult to spot the problem: I used fingertips to unscrew the left hand banana plug's nut.
I couldn't get a spanner in there to tighten it up, so the front panel had to be shifted.
Looking at the entire chassis, the base and rear panel are traditional solid sheet metal, but the top, two key regulator PCBs and the front panel are held in place by a three small L-sections and more of those wretched plastic slivers. No wonder there wasn't a carrying handle on the lid.
Half of me admires the Ikea-style value engineering, half of me doesn't.

Anyway, back together and back on the shelf. Still like it and think it was good value

It has a few nice bells and whistles, and I haven't spotted any showstopper vices.

The untorqued nut and those plastic rivets are disappointingly less than HPAK etc quality. Still better than one of their ancestors, Sinclair - but that isn't much of a challenge!

Is this said power supply?

Yes, that's a good picture illustrating all the bells and whistles except for the programmability over USB/LAN/GPIB/RS232. GPIB is optional, and mine doesn't have it.

The only thing I've spotted that isn't visible on the front panel is over-voltage and over-current protection:

OVP (over-voltage protection) and OCP (over-current protection) are implemented in firmware and can only be set and used when under remote control via the RS232, USB, LAN (LXI) or GPIB interfaces. Setting resolutions are 10mV and 1mA and typical response times are 500ms. However, in local mode, OVP and OCP are still active but automatically default to 105% of the instrument's range maximum. This usefully provides shut-down protection in the event of prolonged application of a CV or CC control voltage which attempts to set the output beyond 105% of the range maximum.

Overall I like it, and wouldn't warn anybody not to get one.

AVGresponding wrote: ↑Wed Apr 01, 2026 8:24 am Oh, it's newer than I thought.

At least it's not one of their switchers: https://www.eevblog.com/forum/testgear/ ... msg6101225

I've seen such ugly behaviour in other PSUs, as illustrated in viewtopic.php?p=12663#p12663 . I tend to doubt that such ugly behaviour is a consequence of it being a switch mode PS; in any modern SMPS the frequency is several orders of magnitude higher than the observed "1s to finally get it right". (I did once have an SMPS that used thyristors at mains frequency, with large capacitors and inductors in the output filter!)

I see that EEVBlog post speculates about current limiting being done in software rather than hardware. I've wondered the same, particularly in light of my PSU having OVP and OCP in firmware.

tggzzz wrote: ↑Wed Apr 01, 2026 9:07 am I see that EEVBlog post speculates about current limiting being done in software rather than hardware. I've wondered the same, particularly in light of my PSU having OVP and OCP in firmware.

And I see your earlier quote suggests that the overcurrent and overvoltage typically takes 0.5s to act; that's easily enough time for magic smoke release. Nevertheless, I like the (linear) TTi supplies. And I do like the fact that TTi realise that you might want an earth terminal.

EC8010 wrote: ↑Wed Apr 01, 2026 9:26 am

tggzzz wrote: ↑Wed Apr 01, 2026 9:07 am I see that EEVBlog post speculates about current limiting being done in software rather than hardware. I've wondered the same, particularly in light of my PSU having OVP and OCP in firmware.

And I see your earlier quote suggests that the overcurrent and overvoltage typically takes 0.5s to act; that's easily enough time for magic smoke release. Nevertheless, I like the (linear) TTi supplies. And I do like the fact that TTi realise that you might want an earth terminal.

Agreed. It seems to be difficult, but not impossible, to find PSUs without a front panel earth connector.

For the avoidance of doubt, the 500ms OVP and OCP are not the basic "front panel" voltage and current limits - which work well.

The 50µs transient response and programmed voltage changes aren't stupid either...

I suspect you meant to type "with" an earth terminal. Thank you for clarifying the doubt, I had decided you probably didn't mean the front panel controls...

I've looked at 7 PSUs recently. All but one had a front panel banana and/or spade earth connector.

It's different here. Neither of my HPAK low voltage dual supplies has an earth terminal. Nor do my Keithley 2400. My TTi supplies do have earth terminals, and anything I made myself does.

Huh. Pretty much every one of my HP supplies , single or dual, has an earth terminal (though as you might guess, the newest one is likely still at least 30 years old...)

-Pat

tggzzz wrote: ↑Wed Apr 01, 2026 9:07 am I see that EEVBlog post speculates about current limiting being done in software rather than hardware. I've wondered the same, particularly in light of my PSU having OVP and OCP in firmware.

It's not speculation, the manual hints at it, and the tech support that they spoke to explicitly stated it to be the case, and that the behaviour was "a feature and not a bug", saying that the PSU in question is intended for larger, inductive loads, with the software optimised for that scenario, without any option to switch between response profiles.

AVGresponding wrote: ↑Thu Apr 02, 2026 9:13 am

tggzzz wrote: ↑Wed Apr 01, 2026 9:07 am I see that EEVBlog post speculates about current limiting being done in software rather than hardware. I've wondered the same, particularly in light of my PSU having OVP and OCP in firmware.

It's not speculation, the manual hints at it, and the tech support that they spoke to explicitly stated it to be the case, and that the behaviour was "a feature and not a bug", saying that the PSU in question is intended for larger, inductive loads, with the software optimised for that scenario, without any option to switch between response profiles.

Point taken, but the post is poorly written and somewhat self-contradictory.

I imagine if I re-read the whole sub-thread carefully, I might be able to catch up on exactly what is/isn't being referred to.

tggzzz wrote: ↑Thu Apr 02, 2026 11:04 am
I imagine if I re-read the whole sub-thread carefully, I might be able to catch up on exactly what is/isn't being referred to.

"For more information please reread"

(The reference awaits spotting)