hyland psu questions

[p.larner]

on the schematic,it looks like u3+u2 have a negative feed on pin4 and the +ve supply on pins7,how do i measure the negative supply to them?,also can i feed the+ve opamp rails from say an 7812 reg,or wll that screw up the operation of the psu?,it seems the max voltage on the opamps is on the edge of what a tl081 can take.

[Jwillis]

You measure the op amp power rails relative to the common ground.
If you pre-regulate the positive supply rail to the op amps to 12V then the maximum output voltage will be aproximately12V.

[p.larner]

so where are the common grounds?,on u3 do i measure between pins 4 and 7,i am trying to work out how too see if the -ve rail is working,but not sure what points to probe.

[p.larner]

can 741 opamps be used ok in this?.

[p.larner]

anyone.

[Jwillis]

Connect your negative probe to the common ground that I labelled  on the schematic. Probing the pin 7 on U1, U2 or U3 with the positive probe  will read a positive voltage. Probing pin 4 on U1, U2 or U3 with the positive probe will give you a negative reading. Because the positive rail and negative rail voltage are relative to the common ground. The Common Ground is the 0V potential of both the positive rail and the negative rail.
If you measure between  pin 7 and pin 4 you get the added voltage of both the positive rail  and negative rail.


A 741 op amp would work. I've also dropped in a LT1637 which is just as slow as the 741 and it works fine without further modification. Using any op amp with ±22V supply works way better than the TL081CP suggested for the design. 
Since your dealing with DC (0Hz), then op amp parameters such as Slew Rate, Unity-Gain Bandwidth and Phase Margin become unimportant. A slow op amp will work just as well as a fast op amp at DC or (0Hz)

Are you building the kit or just using the circuit diagram for your own scratch build? What is the VA and Secondary Voltage of your transformer? How many Amps do you require? How much ripple voltage is exceptable? Some changes to the input filter circuit may needed depending on requirements. 

[p.larner]

let me explaine,a couple of years ago i brought the remains of a farnell psu,it had the main transfo removed,the control board i removed but kept the filter cap (22000uf) and the 6 2n3055 pasas transistors on there heatsinks,the transfo i have fitted is from a ham psu like in the pic,i want to use the hyland kit to drive/control the pass transistors already fitted to the farnell remains,i have done away with the rectifier diodes on the hyland board and am using a 30amp off board bridge rectifier.

[kevin.gibbs]

It is better to see which op-amps are in the original circuit and then choose a replacement. If you do not have data about resistors, Zener diodes, and capacitors, repeating the circuit may become a problem.

[p.larner]

its a cheapo aliexpress kit i am using with the tl081 opamps. but i have a selection of opamps i can use.

[CaptDon]

Why no spaces after the commas in your sentences? Are you a non-native typer of english?

[Kleinstein]

can 741 opamps be used ok in this?.

AFAIK the plan is based on an old circuit from a electronis magazine and originally was using 741. So the 741 would be fine.
The TL081 is only OK for a lower voltage (e.g. < 20 V transformer).

The PCB of the kit has a slight problem with the ground layout. The ripple from the filter cap couples to the current control loop. This can be fixed with a few badges if needed.

[p.larner]

my transformer outputs about 19v unloaded, reads 40.7v on the filter cap after the bridge rectifier unloaded,is that ok for a 741 not taking into account the minus rail of 5v?.

[Gyro]

No. Read the datasheet.

[p.larner]

Hi chris,i found some MC34071AP ic's, would they be ok?.

[Jwillis]

let me explaine,a couple of years ago i brought the remains of a farnell psu,it had the main transfo removed,the control board i removed but kept the filter cap (22000uf) and the 6 2n3055 pasas transistors on there heatsinks,the transfo i have fitted is from a ham psu like in the pic,i want to use the hyland kit to drive/control the pass transistors already fitted to the farnell remains,i have done away with the rectifier diodes on the hyland board and am using a 30amp off board bridge rectifier.

 


Your PALSTAR PS-30M power supply is/was  0-15V 30A Max design. The transformer has 2 secondaries. One for high current to the power transistors and one for control circuit, which will likely be low current. Both secondaries  may not have the same current rating.
Can you post a picture of just the transformer secondary wires.
Voltage measurements need to be taken of each secondary separately to identify the transformer properties and limitations. From looks of the picture below, there appears to be a high current secondary and a low current secondary. The transformer looks to be around 400VA to 500VA. That would put the high current secondary at around 15V to 19V at best.
If the 2 secondary windings have vastly different current ratings you cannot connect them in series and expect to get 30A without burning out the low current secondary.
You may not be able to get full advantage  of the hyland board because of the transformers limitations.

I added a circuit diagram of the original PALSTAR PS-30M.

[p.larner]

Its the high current secondarys i am using,the raw ac unloaded is 19v, For some reason i get 40.5v when i connect the filtercap to the bridge rectifier output!, It seems too high to me?.

[Jwillis]

Its the high current secondarys i am using,the raw ac unloaded is 19v, For some reason i get 40.5v when i connect the filtercap to the bridge rectifier output!, It seems too high to me?.

Ok.
The unloaded 40V is probably the low current secondary. That would work out to be around 25V to 28V from that secondary. The current  would probably be rated for around 1A.
The High current secondary is probably around 19V at 30A.
So using with the hyland you can expect theoretically around 0v to 15V at 30A using the high current secondary. For that amperage some modifications would be needed to the hyland since the current limit is set for around only 3A. If you only want 3A then a direct drop in would work.
If you modified the hyland board to run on the low current and passed the high current separate then you can achieve better stability and a cleaner output. Pretty easy to do. I think I have a modified circuit diagram of the hyland already drawn that does just that if your interested.     

[p.larner]

the widings i am using are the thick heavy secondarys,the aux secondarys are unconnected,my plan is to use a low value 100w current sense resistor,i have also found some mc34071p opamps to try in this in place of the tl081 parts,with 40v on the filtercap and the -5v, its still 1v over the opamps 44v rating,should i use say a 4v zenner n the - rail charge pump?.

[Jwillis]

Something is not right if your getting 40V after the rectifier and filter capacitors supplied by the 19VAC secondary. With a 19VAC secondary you should be getting only about 27VDC after the filter capacitors.

[xavier60]

Not certain if this applies here. A capacitive load on one of the secondary ends can cause a voltage boost.

[Jwillis]

Not certain if this applies here. A capacitive load on one of the secondary ends can cause a voltage boost.

The bridge rectifier for the high current secondary has parallel caps with the diodes. Would it be plausible that a diode is open and one of the caps is causing the rectifier to act as a doubler? I've not worked with that type of rectifier or what voltage would be expected.
 

[coromonadalix]

the palstar psu does not have a good current sensing for the current limiter 

it sniff out a portion of the current shared by tr3  ...  not god to me,   i never saw that ...


filtered dc passed thru an full bridge rectifier calculation is :   Ac  x 1.414 = xxx

[Jwillis]

the palstar psu does not have a good current sensing for the current limiter 

it sniff out a portion of the current shared by tr3  ...  not god to me,   i never saw that ...


filtered dc passed thru an full bridge rectifier calculation is :   Ac  x 1.414 = xxx

From what I understand, He's using the remains of the original PS-30M without the control circuit. Just the transformer rectifier and filter caps. But your right about VAC x 1.414 = Peak DC voltage after rectification and filter caps. This is why the measured 40VDC doesn't make sense.

[floobydust]

A relative of treez? The troll voltage seems too high.
The Palstar filter cap rated 35V, hopefully it does not go bang, probably a good idea to keep one's face away from it- if the voltage measurements are real.

edit: high quality CapXon parts - DUCK

[IanB]

Your PALSTAR PS-30M power supply is/was  0-15V 30A Max design. The transformer has 2 secondaries. One for high current to the power transistors and one for control circuit, which will likely be low current. Both secondaries  may not have the same current rating.
Can you post a picture of just the transformer secondary wires.
Voltage measurements need to be taken of each secondary separately to identify the transformer properties and limitations. From looks of the picture below, there appears to be a high current secondary and a low current secondary. The transformer looks to be around 400VA to 500VA. That would put the high current secondary at around 15V to 19V at best.
If the 2 secondary windings have vastly different current ratings you cannot connect them in series and expect to get 30A without burning out the low current secondary.
You may not be able to get full advantage  of the hyland board because of the transformers limitations.

I added a circuit diagram of the original PALSTAR PS-30M.

Remember the OP does not have the original transformer. In fact the only thing that remains of the original power supply is apparently the box, some pass transistors and filter caps. Everything else is missing or removed. So any reference to the original schematic is void.

We do not know the details of the replacement transformer, what windings it has, or what size it is. There is no picture of it and no specifications given.

[Kleinstein]

The hyland circuit makes not so much sense with high power. The higher power supplies usually use transformer tap switching or a similar technique to reduce the power dissipation.
The simple circuit would limited the useful current from the heat.

[Jwillis]

Your PALSTAR PS-30M power supply is/was  0-15V 30A Max design. The transformer has 2 secondaries. One for high current to the power transistors and one for control circuit, which will likely be low current. Both secondaries  may not have the same current rating.
Can you post a picture of just the transformer secondary wires.
Voltage measurements need to be taken of each secondary separately to identify the transformer properties and limitations. From looks of the picture below, there appears to be a high current secondary and a low current secondary. The transformer looks to be around 400VA to 500VA. That would put the high current secondary at around 15V to 19V at best.
If the 2 secondary windings have vastly different current ratings you cannot connect them in series and expect to get 30A without burning out the low current secondary.
You may not be able to get full advantage  of the hyland board because of the transformers limitations.

I added a circuit diagram of the original PALSTAR PS-30M.

Remember the OP does not have the original transformer. In fact the only thing that remains of the original power supply is apparently the box, some pass transistors and filter caps. Everything else is missing or removed. So any reference to the original schematic is void.

We do not know the details of the replacement transformer, what windings it has, or what size it is. There is no picture of it and no specifications given.

OH! Ok. Well that makes sense now. I misread the post. Thank you.

[Jwillis]

The hyland circuit makes not so much sense with high power. The higher power supplies usually use transformer tap switching or a similar technique to reduce the power dissipation.
The simple circuit would limited the useful current from the heat.

Ya your right. It takes a fair bit of redesign to get it to work at high current. It can be done but tap switching is a must to keep the heat dissipation of the transistors manageable.
I would be happy to share the rework I did for 0-40V 20A. Works like a charm.

[p.larner]

I have fitted the power transformer from the old ps-30 into the remains of the farnell psu as well as a 50 amp bridge recfifier and 22000uf filter cap, The hyland board is just being used as a driver for the 6 2n3055 pass transistors that remained in what was left of the scrap farnell psu, I hope that clears things up.

[p.larner]

can you share your rework here then?,cheers Paul.

[Kleinstein]

From the plan shown the PS30 transformer seems to have 1 main winding and 1 auxiliary one with likely a lower current rating.
Before discussing more about the circuit it would make sense to check the transformer (it may have additional taps or 2 high current windings).
Another point is the wanted output - so what are the priorities / wishes.

[p.larner]

the transformer has 2 totaly separate windings one low current aux+ a high current main wnding no center tap, I would like 30v with 15-20 amps max if poss.

[Kleinstein]

15 A at 30 V would 450 W of output power. With the typical losses one would start with more like 35-40 V and thus 500-600 W. With a simple rectifier/filter the transformer would need about 750-1000 VA for this because of the more pulsed current.
A 500 VA transformer and maybe 40 V at the filter capacitor would be good for maybe 8 A.  This would likely still be too much for the transistors and heat sink, unless there is a massive fan.

[p.larner]

Well as the farnell psu is scrap as it is i want to make use of it using the hyland pcb as a driver for the remaining unit inc pass transistors,i am not to bothered about its max output and will take whatever i can get,in my eyes a bodge /lashup is better than throwing it into landfill 8a at 25v would be ok.

[factory]

A relative of treez? The troll voltage seems too high.
The Palstar filter cap rated 35V, hopefully it does not go bang, probably a good idea to keep one's face away from it- if the voltage measurements are real.

edit: high quality CapXon parts - DUCK

Not treez, see this thread; https://www.eevblog.com/forum/repair/solartron-dmm/msg5314252/#msg5314252

[Jwillis]

can you share your rework here then?,cheers Paul.

This is a highly modified version of the original Hyland design. The modification required a complete restructure of the PCB. The opamps used here are the OPA445 high voltage op amps. Other modifications were made for the extreme voltage and current requirements.But the basic design still exists.

Modified Control Board
Starting from the left side there is Positive CB which is a separate low current supply input for the control board only. In this case its around 50V.
GND (Common Ground) is the ground potential for all circuits of the entire power supply.
Negative supplies the op amps. It also  keeps the discharge transistor Q104 in an off state during operation.
Collector Rail  is connected to the Main high current supply. This is also connected to the Power Transistor Array Collectors. Q101 is the Power Transistor array driver and not the final positive output.
Base Rail is connected to the Bases of the Power Transistor Array.
The Emit Rail is connected to the Emitter Rail of the Power Transistor Array. This is required for the control board to sense output voltage for operation.
V Tap sense is for the Tap switcher operation.

The point I'm making here is the original design will work much better when you use a separate supply for the control board and keep the high current at the power transistors only.
Tap switching the transformer is a must for high current linear supplies.  You must consider the heat dissipation of  the power transistors. You calculate this with  Collector voltage minus the Emitter Voltage multiplied by output Current. So If the Collector voltage is 40V and you've adjust the supply to be 1V output  and a connected device is drawing 10A that is 40 - 1 X 10 = 390W
That's a lot of energy that the power transistors need to dissipate as heat. If your had 6 transistors you would be OK since 390 Divided by 6 is 65W per transistor.
Look what happens at 20A. 40 -1 X 20 = 780W or 130W per transistor. That exceeds the dissipation of the 2N3055 transistor and they will burn out. That why tap switching for high current is so important.

If you run your control board on a separate supply at a lower voltage like 30V instead of the 40V, then you can use the 741 or  MC34071p with ease. Use the 40V high current for the Power Transistors only. You may also need a separate high current ground rail as well. The traces on the Hyland won't handle more than 5A at the most before they start to burn.

I suggest you start with low current first. Understand how the control circuit works. Then you can make modifications for higher current as you learn more.
Making these modification wasn't done in a weekend. I started this project several years ago with little knowledge.
I'm not an electrical engineer. It took a lot of searching, reading, learning and experimentation to get it to work the way I want it to.

The modified control circuit is below plus a couple photos of the stages of prototyping the design. First photo is the "Bread Boarding" and the second photo has the Prototypes Control Board, plus it's required power supplies, and Transformer Tap Switch Board. Pardon the mess. And I'm still not finished!
 

[p.larner]

This is basicaly the design i plan to use, I am just waiting for  1 watt zenners and  the 2 watt resistors to be delivered. The main Ground track  will be beefed up and a a 100 watt current sense resistor mounted offboard on a heatsink. what

[Jwillis]

How many Amps do you want? Anything over 10A, at best, and the circuit won't work as expected. Even with a 1kVA transformer. Already tried it. It has to do with Transformer Regulation.
All transformers suffer from losses consisting of I²R copper losses and magnetic core losses. This means that as more current is drawn from the secondary the voltage drops.
This effects the output voltage of the power supply control board.

Also, the peak voltage after rectification and filters is unusable. Only the RMS voltage of the secondary is usable.
With a 28VAC secondary the Peak DC voltage after rectification and filtering will be around 40V. But as soon as any current is drawn from the control board output, the maximum voltage will drop immediately to 28VDC. You cannot expect the peak DC voltage from the filter caps to maintain peak voltage under load. As more current is drawn that voltage will drop more. You will not get the expected 30V on the output of the power supply under load.
 
By using 2 separate AC secondaries, the voltage drop on the high current secondary will have much less effect on the voltage of the low current secondary. This means better stability at higher currents.

[p.larner]

should i also incorparate the mods as shown in this schematic,but use the last schematic output section?, If that makes sense?,At the moment i am waiting on some 2 watt resistors and a 1 watt 10v zenner to arrive.

[Jwillis]

 Your pre-regulating only U3 to 10V, but it's inputs will be as high as 30V. You'll damage U3. ALL of the op amps have to be pre-regulated to the same supply voltage. 
If you pre-regulate all the opamps to 10V, then your maximum output voltage of U1, U2 and U3 will be 10V.
If you want 30V output, all the opamps positive supply voltages need at least 30V, not taking in account the Maximum Output Voltage Swing of the op amps being used.

Understanding Operational Amplifier Specifications
Op Amps forEveryone

[p.larner]

so using a second low power supply is better for the control pcb is the way to go then?.