Lucky or not ?

[janengelbrecht]

I have designed a Laboratory Powersupply: 0-33V, 0-3.3A for a school.
Problem is: It seems to work on breadboard...but am I just a lucky son of a bitch ? Will diversions in parameters make it fail big time ?

[janengelbrecht]

Errors:
1) U3 isnt connected to 56V but to 35V
2) J3 and J4 are in fact connected directly to Q2 and Q3
(Sorry ive uploaded old versions!)

Digital Voltmeter connected to: J6
Digital ampmeter connected to: J7
Trafo connected to J1 (2 outputs are: 40V and 2 are: 10V)
D4 are infact conected to J11 (not on PCB)
Output is J5 (Infact is is connected to a switch....one position connects output+ to gnd, the other to output terminal: Output ON/OFF)
Fuse is infact connected to J8 (not on PCB)
Potmeter RV1 are in fact connected to J10 (not on PCB)
Potmeter RV2 are in fact connected to J9 (not on PCB)

[Benta]

I'd like to see the heatsink for Q2 and Q3. They'll be dissipating 100 W worst case, so it should be 0.6 K/W or lower. That's BIG.

[janengelbrecht]

well....(55V-10V)*3.3V = 115.5W divided by two 2N3055
Yes its big and equipped with 12V FAN

It could be better to divide the input voltage further....

[Le_Bassiste]

jan,
not quite sure, but it seems that the two 2N3055 are not sufficient to get your lab supply down to 0V at 3.3A. seems that you are exceeding the SOA at that operating point. adding a third one would give you some margin.

[Benta]

jan,
not quite sure, but it seems that the two 2N3055 are not sufficient to get your lab supply down to 0V at 3.3A. seems that you are exceeding the SOA at that operating point. adding a third one would give you some margin.

No problem.

Real men use 2N3773 anyway: "More power! argh, argh, argh!"   

[Audioguru]

Your LM317 used for U3 is missing the important protection diodes shown on its datasheet.

[janengelbrecht]

jan,
not quite sure, but it seems that the two 2N3055 are not sufficient to get your lab supply down to 0V at 3.3A. seems that you are exceeding the SOA at that operating point. adding a third one would give you some margin.

In fact...it works:) But again....it could be a good idea to distribute the power to three 2N3055 instead of two. Then each only has to pass 1.1Amps and then the max effect for each is 38.3 W. Well within the limits for 2N3055.

Sendt fra min EVA-L09 med Tapatalk

[Le_Bassiste]

  cool!
one thing i'd definitely add, though: a 100V/10A "freewheeling diode" between J5 (cathode) and j7(anode). firstly, that enables your design to run safely in series-connected mode, if you plan to double your design and connect them in series for higher O/P voltage. and secondly, as a protection against inductive loads that can generate serious (reverse) overvoltages in case your supply shuts them off rapidly. 

[Kleinstein]

The current measuring shunt is not compensated by the voltage feedback - so the output resistance will not be that low, not below 0.1 Ohms.

With a 32 V supply for the OPs, the output voltage might not reach 30 V, maybe 27 V are possible (depends on the current and gain of the transistors).

Switching the transformer tap based on the set output voltage is not reducing the worst case power loss. In case of a short, it would still use the high transformer tap.
Also switching from something like 20 V to 40 V causes a significant current spike with so large filter caps - this at least is a significant load to the relay contacts. There is no hysteresis on the switching - thus worst case the relay could be turning on/off quite frequent worst case. The transformer taps might not work as intended with the second 55 V supply used and also connected to GND - it might need the lower tap (e.g. 20-30 V) to be fully isolated.

For a 3 A supply no more than 10000 µF should be enough.
With a large filter capacitor the transformer is seeing a significant RMS current - may be more than 6 A for a 3A output.

The LM317 is not really happy the 55 V at the input - especially if the supply rises fast. Chances are it can blow.

[Damianos]

Some more comments on the design:
- The output voltage will be varying with the load current: the current sensing resistors and the ammeter are in series with the load...
- What is the purpose of the C10 and C11? These will make the circuit to oscillate at high frequencies. It is needed the opposite: to slow down the control circuit because the output transistors cannot react fast.
- What is the purpose of the D2? If it fails, due to a capacitive load, by opening, the circuit will be corrected automatically! If it fails in short circuit then there are consequences.
- The values of the components around the U3 indicate that the data sheet is not studied! Note that is used the LM317 and not the LM117, there are some differences.
These from a first glance at the circuit.