HMC8042 parallel operation low current hunting output?

[m0rjc]

I bought an HMC8042 - nice compact unit that fits on a desk and is quiet and should suit my needs for most things albeit seriously expensive!
It came without its box which is odd - waiting for the seller to get back to me on that one.

Most uses for me are within 15V and 3A, so a single channel is fine, but I accepted the smaller outputs thinking if I needed to I could parallel this. I have a somewhat cheaper 9-15V 30A switcher if I want to run higher power amateur radio equipment, so this is for microcontroller projects and low power radio projects.

When I do parallel it and current is low it switches back and forth between sides (non-professional video of display here https://photos.app.goo.gl/K8KK9Sj6e9tne6qK8). On the output on the scope I see the voltage move up and down as it does so. The current indication on the display includes what I think is referred to in the manual as "compensating currents between the channels", about 15mA varying.

I didn't screenshot the scope in DC mode. The deviations were noticeable above the noise (which I assume was picked up on my 50cm test leads and the amount of stuff effectively dangling off a non isolated USB oscilloscope). With the scope set to AC Coupled it looked like this:



The load is shared once the current increases. I tested with radio transmitting into dummy load so pulling a couple of Amps.

Is this normal for any DC twin power supply operating in parallel mode at low current? So switching to a different one won't solve it? (Though I guess a less precise unit will always end up with one side taking charge as they can never be set the same unless it has a tracking mode?)

I can mitigate it by setting 10mV difference between the sides, then the higher voltage channel takes the strain until limit before the second channel kicks in.

Or I can decide that it arriving with no box is really bad on such an expensive bit of kit, A TTi 200D is cheaper and I wouldn't have to worry about running out of current. Given the induced noise around my desk, the noise figure is arbitrary (though the new office may improve that) and nice as the automation features are I'll never use them?

Thanks

[Kean]

I've not paralleled the channels on my HMC8042 from memory, but the manual does suggest this behaviour is expected and that it is better to set one channel 50mV lower than the other and let it supply a smaller fraction of the current (if practical) until it hits its limit, at which point the other channel will take over.  Read section 6.6 of the user manual.

One of the reasons I bought the HMC8042 was specifically for test automation via ethernet along with a pair of HMC8012 and my existing equipment.  In my case if I needed higher current and was worried about regulation then I'd probably just use a different beefier supply.  I've not used the TTi supplies, but I am quite happy with my EEZ BB3's.

[m0rjc]

Thanks. The BB3 looks an interesting project. Looks difficult to import into the UK though - buyers need a valid VAT number so not private buyers.

At the moment I'm sharing workspace with kids. When I can get them off it I'll have a go at aligning an old kit radio I built and could do with getting back on air. That will give me time to use the 8042 in anger on a more typical project for me and see what it's like to actually use.

Still feeling grumpy about it arriving loose in a large carton without its own box, but that's another story. I'll see if my supplier replies to questions about that. Slight fan rattle - from bumping around in a delivery van or normal? That kind of thing.

[Kean]

Yeah, I hear import into UK is a pain right now across the board for smaller sellers/buyers.  I don't know if you can buy a BB3 via Mouser in the UK?

I ran some quick tests with the channels paralleled on my HMC8042, and one of my DC loads.  The AT8612 loads I have on the same bench are limited to 3A, so I couldn't try high currents.

• If I used "matched" cables from the channels to the load (i.e. piggy backed banana jacks at the load) then the current draw was somewhat balanced between the channels.
• If the cables were different, e.g. I paralleled them at the PSU and then ran a test leads to my load, then one channel was definitely significantly favoured.
• If I adjusted the voltage on one channel up or down, then that would allow me to balance the current between channels.  Of course that would end up with voltage variance at the load.
• If the channels are set at different voltages, then with the load off/disconnected you could see up to about 200mA flowing between the channels.  This current sinking to try regulate the lower voltage channel would stop once the voltage difference between the channels got too high.  Maybe via software control to limit power dissipation when sinking
• When the channels are set to identical voltages, then even with no load you may see some bouncing around at low current just as you saw as it tries to keep both channels in regulation.  I saw up to 10mA with no load, but it could depend a lot on your wiring.  I used heavy duty Hirschmann test leads

So in your video where you saw the approx 200mA bouncing back and forth, what was the load and how was the wiring done between channels and load?

I suspect that a PSU that does the parallel mode via an internal relay, similar things happen but via lower resistance connections.  And the load balancing is hidden from the user.

Regarding the fan - I don't think I've ever heard it come on apart from a short operation at start up.  Admittedly I'd rarely put more than 30W load on it.  I've just run it at 50W into a DC load for 15 minutes and only the fan in the load has come on.

If you purchased it as new, then it definitely should have come well packaged in its original box.  Otherwise it should have been sold as used and probably at a discount from list price.

[exe]

Thanks. The BB3 looks an interesting project. Looks difficult to import into the UK though - buyers need a valid VAT number so not private buyers.

I'd ask prasimix (the guy behind BB3) about that: https://www.eevblog.com/forum/projects/eez-h25005-a-possible-successor-of-eez-h24005-programmable-power-supply/  .

[m0rjc]

Thanks.

Re: Mouser - I checked their web site and didn't find anything looking for EEV or BB3, apart from a module for it which wasn't a power supply module. I'll keep looking and check that link.

Re: Wiring - I had it paralleled with short but not particularly heavy duty leads, then took from one side. So my wiring wasn't balanced. The load current was around the 150mA to 200mA mark, so you were seeing that swapped from side to side. Moving the settings did stop that happening. With both sides set equals when load current was increased to about 2.5A as the radio transmitted the current was shared. At low current the voltage fluctuation could be seen at the load as it swapped back and forth.

I do wonder if the devices with a "run it in parallel switch" which turns half of the control interface off hide all this internally.

My choice is interesting - accept this really quite nice power supply? It's compact, quiet, should alongside my 9-15V 30A radio specific supply do what I need as long as I stick to low power control or audio work and don't do something like try to build a power amplifier. Or I go the other end of the scales and buy something like a TTi CPX400D which will replace my 30A supply too. No automation or digital control so would be much like the ancient fully analogue linear supply I had from the 80s. Pretty physically big and I don't know how loud that fan is. The fan in this one seems to have settled and I didn't notice it at all while focusing on the radio. Testing things like "how does it behave as the battery goes flat [1]" was a little clunky with the "live mode" interface and the automated way using EasyArb seemed very fiddly. Easy read current display in mA is a nice feature with a radio that only draws a couple of tens of mA on receive - but then I have an external ammeter I can use for that.

[1] The radio - behaved dramatically below about 9V, such that it would transmit out of band if battery dips under load. Something to be aware of. Came back when turned back up again so when receiving all would seem fine. But that's another story.

[Kean]

On Mouser look for CS-BB3-04 - they were originally distributed by Crowd Supply (owned by Mouser)
Also see https://mouser.com/manufacturer/envox/ for their product range (mostly not currently in stock)