LAB PSU Schematic Explaination

[xoom]

Little update on old project

Nice job. How did you get the enclosure?

I found it in metal scrap yard.. actualy found 3 of them

Hi!

Can I replace D11 D12 with LEDs to indicate voltage mode and current mode?

Thats what i did

Simply replacement? Is color have matter? Why original circuit uses 1A diodes?

Yes, color doesn't matter .. maybe becouse they are popular and cheap

[xavier60]

I hope those blue Pots aren't the same as mine . They started going intermittent open circuit.
This will cause the output voltage to increase in some power supply designs.

[xoom]

i have another PSU with those pots.. so far so good hope they dont start to die

[ferdieCX]

I hope those blue Pots aren't the same as mine . They started going intermittent open circuit.
This will cause the output voltage to increase in some power supply designs.

This is a quite common design error.

[xavier60]

I hope those blue Pots aren't the same as mine . They started going intermittent open circuit.
This will cause the output voltage to increase in some power supply designs.

This is a quite common design error.

I would rather that the Pot adjusted the reference voltage into the error amplifier.
With the addition of bleed resistor, the power supply becomes fail safe to a Pot failure.

[001]

I hope those blue Pots aren't the same as mine . They started going intermittent open circuit.
This will cause the output voltage to increase in some power supply designs.

This is a quite common design error.

I would rather that the Pot adjusted the reference voltage into the error amplifier.
With the addition of bleed resistor, the power supply becomes fail safe to a Pot failure.

Can You draw mod for this shematic?

[Kleinstein]

When replacing the diodes (D11 / D12) in the regulator with LEDs, chances are the LEDs can see quite a bit of reverse voltage, possibly more than specified for most LEDs. Chances are it can still work, though outside the specs. The diodes only need to conduct a small current (about 1  mA), just enough for LEDs. With slightly smaller value for R10 the current could be increased a little if needed.

Changing the circuit, so that the reference voltage is adjusted instead of the divider is a good idea. This is not just because of possibly bad contacts, but also to have an essentially constant regulation loop. So regulation can be faster at high voltage and less ringing at low voltage
It likely needs an extra buffer (e.g. 1 more OP) at the upper end of R3 and the pot as a divider before the OP.

There are a few more odd points in the circuit (e.g. C1,  no resistor at the inverting input of voltage regulator, DW3 not fully avoiding spikes, not fast current limit). So the general concept is OK, but the details could have been solved better.

[001]

Thanx!

Is it clone of something wellknown circuit? Any impruwements?

[Kleinstein]

The circuit is one classical type of regulator. Hard to tell if it is a clone of another circuit - there are lots of lab supplies (something like 75%) that use a similar type of circuit, with the floating auxiliary supply for the regulator.  Especially in the old days it was quite common to adjust the divider as shown, though it is not really good.

The main points to improve on this circuit would be:
1) add a fast (usually fixed) current limit, as the present current limit via the OP is rather slow to react. The speed of the current limit is a 2 sided point: a fast reaction can also be confusing. Ideally one would choose slow / fast or have an independent (higher) fast limit.
2) A better lockout for insufficient auxiliary supply, to prevent spikes on turn off / turn on.
3) adjust from the reference side and thus constant loop gain. This comes naturally when using digital control via DAC is used.
   So modern, digital controlled ones usually make it right.
4) LEDs to show CV/CC mode active - usually a comparator from the outputs of the 2 regulating OPs. Directly using LEDs for the diodes is tricky.
5) Some windup limiting, especially for the voltage (e.g. diodes across the OP inputs would be a 1st step).
6) possible 2 power transistors parallel - Chinese supplies tend to be rather optimistic with power handling capabilities
7) some adjustments to the frequency compensation to alow faster reaction and reduced output capacitance (e.g. 100 µF instead of 470 µF).

[Kleinstein]

The transition from CV to CC mode is a tricky part. In the simple circuits (like shown) it take quite a bit of time until the CC regulator comes out of saturation. Some 100 µs are a little on the long side and 2 µs is already rather fast.  There are a few tricks to speed up this time down to the µs range. However there is a downside to the CC mode to take over so fast: to counteract fast load transients the CV sometimes need to give short pulse of higher current. So quite often a short time higher current is actually good.  Ideally one would have 2 current limits: an accurate one, that is a little slow and may take some 10-100 µs to engage and an additional fast one that is a little higher (e.g.  150%+1A). The common solution is to have a fast, but fixed limit at something like 1.5-3 time the maximum current and than a moderately slow enacting adjustable limit.

There is also the charge from the output capacitor, that can deliver quite some current on a short, with no control. A 100 µF cap charged to 20 V can deliver 20 A for 100 µs in case of a "short". So there is limited use of having the current limiting much faster than this, though it gets more moderate at lower voltage.

[YaakovK]

Hello Xoom,
Just wondered if you can help me to find a schematics of the HY3003D model?
Many thanks