Korad KA3005x (how much) better as Basetech BT-305?

[Pfriemler]

Got a Basetech BT-305 (complete "analogue", 0-30/5A, no fan) for 70 bucks (special single's day offer) and tend to return it (spikes, behavior, thermal management), although I like the design (core transformer, 8 raw voltage levels instead "normal" transformer and 4 levels at Korad's).
It's (too) easy to control. Display resolution is 10mA, voltage 100mA, and is not that accurate on top currents (show 4.98 when 5.00 comes out). Has a output button and starts always off with power switch. Does respecable overshoots with low voltages if not limited to output currents <200mA.

Looking for a better choice I should grab a 3005x+ (green voltage display and C.V./output LEDs) with binding posts (plus the few bucks to get a remote control on top). Help me to decide if it's the better choice.

1) BT-305 raw voltages starts from 15 up to 48V, depending on output voltage (raw voltage is 11 to 16 volts uo). Primary buffer is 4700µF - not much for a 5A device, I think. With 1A, ~25W, with 5A ~80W "loss" is on top of the output power. Is a Korad that "bad"?

2) If you shorten the output at 5A (regardless of output voltage, since it steps back to the first level in milliseconds), 10V are heated inside the three fets and warm up the cooler block outside up to 70-80 °C (no fan!). Whole unit takes 85W from AC.
And a Korad 3005?

All shots below are taken with a 1k resistor at the end of the 1m leads, with a 10:1 probe clamped on it. Untangled, of course.
3) overshoot with 3.3V 5A and output button off>on.

Peak 5.7V.
No visible overshoot occurs with a 100mA limit (with 1A, it's only 1V) ... but wait ...

WTF is that? Due to the circuit, the two 220 µF uncharged capacitors on the output (posts) (I measured >600!) are attached with a relay to the charged 33 µF on the output (circuit), and the ramps and steps looks like contact bounces until the output stages begins to charge the output capacitors. Is this a poor design, or is it a poor design? 

4) load change 2A->1A overshoot is ~250mV and comes down in 4 ms. Ok or bad?

5) Mains switch on peak (same parallel output lines to the 1k)

Using a pair of LEDs instead of the 1k limits the peaks to <2V, with absolutely no flash visible.
In the same setup, a change of transformer "stage" to the next level induces a 1µs +/- 500 mV noise at the load. And the Korad?

That's all for now ...

Somehow, the BT-305 is to me neither for electronics nor for "rough" purpose (like battery charging) - or do I ask for too much?

[Kleinstein]

85 W of power consumption with a short is quite high - it may want a fan for that.

many fine steps in the raw voltage has good an bad sides: more frequent tap switching, but less current spike from changing the voltage in the capacitor. The relatively small capacitor could be there to limit the stress to the relays.

250 mV overshoot for a 2 to 1 A step is not that bad. 4 ms to come back is a bit on the slow side but it depends how the curve is - that last 20 mV are usually less critical.

The very fast overshoor spikes (e.g. turn on mains switch)  may well be from the probe / cables. For now I would not take that too serious.

440 µF or the measured 600 µF at the output is not that bad. Capacitors have quite some tolerance and the capacitance can come down after use. Other supplies can have even more.  A output off via relay is a bit unusual. It would be bad if the voltage sensing is only before the relay. This would add the the output resistance.
A fast run on is not needed. The bouncy curve could confuse some parts, but not many.

[Pfriemler]

THX.
I would appreciate if someone simply can do the 5A short circuit test and measure the AC energy - just for comparison.

In general, I like the fine raw voltage steps, too. What I am concerned about is that the linear control overhead is a bit too much especially for high output currents, without any measureable benefit to the voltage regulator. And I do not like the output capacitor - my (switching) MATRIX MPS3206 has ~500 µV and no overshoots at start with rather slow but noiseless ramps (and a comparable overshoot with load downstep 2A->1A, but much faster), with fast voltage downsteps (there must be some discharging circuit inside). Better slow than noisy, is what I think.
Besides this, the Matrix does almost everything right, but the "noise" on the output is why I am looking for a linear supply. (btw: short circuit 5A here -> 25W AC, but it's a switching power supply).
The Voltcraft DPS-2010 (a switching power supply combined with a linear regulator) has 47µF and overshoots of 10-20% under all circumstances. This is even more  of ...
The mains switch peak may come from induction to the wires, but Matrix and Voltcraft show something lower by a factor of 10.

A output off via relay is a bit unusual.

A relay is simply easy. The whole design is from the last millennium.

It would be bad if the voltage sensing is only before the relay.

As I can see, it is. You're right, this may cause more probs if the relay contacts get odd.

[Pfriemler]

Just to answer my questions... My Korad KA3005P+ (Welectron, security plugs, the binding posts were sold out :-( )

(Basetech) With 1A, ~25W, with 5A ~80W "loss" is on top of the output power. Is a Korad that "bad"?

It is. Loaded with 30V 5A, it draws 240W from mains, +90W
And it is very noisy (humming) from the transformator. The Basetech was quiet, but tends to overheat. The fan in the Korad seems to have no problems to blow the heat out.

2) If you shorten the output at 5A ... Whole unit takes 85W from AC.
And a Korad 3005?

~100W

Nevertheless, I am much more happy with that. No overshoots (except for small peaks on repeated short load loss (12V 60W lamp) up to 14V - which do not trigger the OVP, btw), rather slow with load, but <10 ms without. 20 mA for testing LEDs without any peaks. No sparks if leads are shortened.
No competition. The Korad has a WAY BETTER behavior.