How to improve voltage stability of a linear power supply.

[Fflint]

I'm looking for a way to improve voltage stability of a linear (non regulated) 3. 5kW power supply.

The power supply powers a vacuum tube amplifier that requires 3300V DC. It is built using 4 small 900W transformers with 2350V secondaries. The tube requires 1A of current.

It is a straightforward linear power supply with a full wave bridge rectifier and 33uF of filtering capacitors. This produced a no load voltage of around 3300V which quickly dropped under load down to 2300V DC. So additional 470V and 230V transformers were connected in series raising the AC voltage to 3050V. This of course drops to 3050V when 1A DC is flowing.

I thought the only way to increase the DC voltage was to increase the AC side, but then I was told about other peoples power supplies using for example a single 4kW transformer with a 2800V secondary and only 8uF of filtering and this PS voltage never drops below 3500V with 1A of current.

So I'm looking for ways to improve my PS so my voltage doesn't drop that much. I realise the rectifier and the transformers are charging the filtering capacitors only for a very brief period of time during each AC cycle while DC voltage is close to AC peak value. During this brief time the current flowing is many times the current being supplied by the power supply. So capacitor ESR, connecting cable resistances and perhaps the leakage inductance of the transformer will all limit the speed at which the capacitors can charge. I can try to improve all those things to some extent.

But I'm looking for information on anything else I can do. Would using super fast rectifier diodes help? Is there anything else I can do other than go buy a new 50kg transformer?

I could convert the PS to 3 phase(drop the secondary down to 2000V by connecting the primary in opposing way, then across 3 phases I would get 3400V AC) but I would prefer to keep it single phase for portability.

Please don't suggest in this thread I use a SMPS. This is for a "retro style" vacuum tube project. A linear PS is the only way to go for this.

[Kleinstein]

Normally a large 900 W transformer has pretty low output impedance. So the voltage schould not drop that much.

33 µF and 1 A over some 8 ms would be a voltage drop of about 240 V. So quite some ripple, but still not enough to explain the voltage drop. So the size of the filter capacitor can also not explain the drop in voltage.

Ar those transformers more normal ones of microwave transformers ?. Those have a high series inductance and this offer some kind of build in power factor improvement. This comes at the price of the voltage dropping, but it helps to get more power (closer to 900 W).
A normal 900 VA transformer with rectifier and filter cap would only be good for about 500-600 W and one would thus overload the transformers in this case.

Using 3 phase would help a little in needing less filter capacitance, as the cap is recharged 3 times more often.

[bob91343]

As mentioned, you have to find out why the voltage is dropping.  There is the ripple factor and the duty cycle and the source impedance and so on.  Does the secondary ac drop much under load?  I doubt that the rectifiers are a problem at such high voltages, but do they get hot?  Do the capacitors heat up?  The transformer?

Working with sine waves is always a problem with regulation.  (That is the proper word, not stability.)  The longer the conduction angle, the better the regulation, but the more heat generated.  It depends.

[tunk]

No expert on this - are you sure the 33uF caps are ok?

[Fflint]

No expert on this - are you sure the 33uF caps are ok?

In what way, as in are they damaged? I doubt that. There are two series strings of 13 caps. I had one initially (18uF) and I added one more to lower the ripple. Now the ripple is fine (150V at most - with a 330mH choke at output side), but the DC drop is there.

My understanding always was this is normal. I always expected the no load DC to be 1.41*AC(RMS) and at load to drop roughly to AC(Rms), but then I was told about those other PSs that don't drop as much.

I've managed to simulate the entire circuit in ltspice to test few things. I've measured the transformers parameters (winding inductance, series resistance) and I adjusted the coupling factor for the simulation no load secondary voltage and primary current to match.

Ltspice shows the same dc voltages I'm getting in reality. Unfortunately I can't reliably measure AC at 2.3kV so I have to go on what LT spice tells me. Based on that there is barely any voltage drop on the transformers (50V or so).

The capacitors discharge from peak, and on the next peak they don't charge fully. It is as if something is preventing higher current from flowing. I'm guessing transformer'secondary inductance, connection lead resistance, capacitors ESR. Connection leads are at most few tens of milliohms, the inductance of primaries is pretty large (15H each, 4 in parallel are 3.75H), also the resistance of secondaries is pretty large at 140ohms (35 across 4). Capacitors are electrolytic caps (13 in series). If everyone od them is 10 ohms, they contribute 100 ohms total.

Charging currents are slightly under 2A per transformer.

People that have those power supplies that stay close to AC peak at their rated current all seem to use oil filled capacitors rather than electrolytes, but I'm not sure this is relevant, because my ltspice simulation uses ideal caps (no esr) and still there is same voltage drop. I'm leaning towards blaming transformer inductance, or diode overshoot.

So I'm thinking, perhaps if I used fast, or ultrafast rectifiers, that would help? Typical recovery time of normal rectifier diodes is on the order of 1uS. One cycle at 50Hz is 20mS. If the charging time is 5% it would have 1mS to charge. It seems a very long time in comparison with 1uS, but then there is overshoot that normal rectifiers have (they continue conducting for a brief period after the AC waveform crosses zero allowing reverse current flow). Ultra fast rectifiers don't have overshoot allegedly.

I can add more transformers to deal with a nductance, but I wonder if there is anything else anyone can suggest (othen than buying a bigger transformer)

Edit: also the problem with 3 phase is that no load DC voltage would be very high - with ac of 3500V it would be almost 5kV. I would need to waste half of the power of the PS in a bleeder resistor to bring that voltage under 4kV (max tube voltage). A large choke in front of the rectifiers would bring the voltage down, but it would cause all sorts of problems at rapid changes of load (ringing, voltage spikes etc)

[NiHaoMike]

Please don't suggest in this thread I use a SMPS. This is for a "retro style" vacuum tube project. A linear PS is the only way to go for this.

Use a pair of thyristors to modulate the primary voltage, as Carver did for the legendary "Magnetic Field" power supply?

[TheMG]

with a 330mH choke at output side), but the DC drop is there.

How is this choke connected? Is this a choke input rectifier power supply design?

A schematic would be helpful.

[Kleinstein]

A fast diode should not make a big difference. The possible problem with a slow diode is the possibility to get short reverse recovery current peaks, that may cause EMI problems, but not a significant drop in the voltage under load. One could look at the difference in the simulation.

Using 3 Phase is not about using 400 V (or 208 V in the US) input to the transformer, but about having the phase shift and thus only 1/3 the time to bridge for the filter capacitor. A transformer does not work well with a voltage much higher than the nominal voltage due to saturation of the core.

[Fflint]

with a 330mH choke at output side), but the DC drop is there.

How is this choke connected? Is this a choke input rectifier power supply design?

A schematic would be helpful.

No, not choke input. You could call it choke output rectifier :-) I attach a hand drawn schematic below, because I'm not where I have my laptop at the moment. It is a very simple design. A choke input design delivers 0.9*AC(rms) as DC voltage. That is too low for me.

This way to connect the choke is a result of ltspice simulations telling me that ripple is lowered by 50% with it like this. Voltage swing is not affected (it was the same before I added the choke.

Diodes are 5*1n5408

A fast diode should not make a big difference. The possible problem with a slow diode is the possibility to get short reverse recovery current peaks, that may cause EMI problems, but not a significant drop in the voltage under load. One could look at the difference in the simulation.

Using 3 Phase is not about using 400 V (or 208 V in the US) input to the transformer, but about having the phase shift and thus only 1/3 the time to bridge for the filter capacitor. A transformer does not work well with a voltage much higher than the nominal voltage due to saturation of the core.

Yes, that makes sense (about the diodes).

Regarding 3 phase of course I realise that. What I meant is that if I use 3 phase (with 3 transformers, I would have to drop 1) and I connect each primary between a phase and a neutral then I make a full wave 3 phase rectifier it will be as if my AC input to the rectifier is 1.7 times the secondary voltage Rms with the DC voltage equal to ((AC line voltage Rms)*2.449) Even if I wire the secondaries in opposition to primaries in series to drop the secondary voltage down by 230V this still results in 2070*1.7=3517V line voltage AC on the rectifier (star connected primaries, and secondaries) so the DC becomes almost 5kV! (3517*1.41).

[magic]

I've managed to simulate the entire circuit in ltspice to test few things. I've measured the transformers parameters (winding inductance, series resistance) and I adjusted the coupling factor for the simulation no load secondary voltage and primary current to match.

Ltspice shows the same dc voltages I'm getting in reality. Unfortunately I can't reliably measure AC at 2.3kV so I have to go on what LT spice tells me. Based on that there is barely any voltage drop on the transformers (50V or so).

The capacitors discharge from peak, and on the next peak they don't charge fully. It is as if something is preventing higher current from flowing.

At this point you should be able to figure out what causes it, at least in SPICE.
Then try to confirm if it can be the problem IRL too.

[m k]

I'm looking for a way to improve voltage stability of a linear (non regulated) 3. 5kW power supply.

The power supply powers a vacuum tube amplifier that requires 3300V DC. It is built using 4 small 900W transformers with 2350V secondaries. The tube requires 1A of current.

It is a straightforward linear power supply with a full wave bridge rectifier and 33uF of filtering capacitors. This produced a no load voltage of around 3300V which quickly dropped under load down to 2300V DC. So additional 470V and 230V transformers were connected in series raising the AC voltage to 3050V. This of course drops to 3050V when 1A DC is flowing.

I thought the only way to increase the DC voltage was to increase the AC side, but then I was told about other peoples power supplies using for example a single 4kW transformer with a 2800V secondary and only 8uF of filtering and this PS voltage never drops below 3500V with 1A of current.

So I'm looking for ways to improve my PS so my voltage doesn't drop that much. I realise the rectifier and the transformers are charging the filtering capacitors only for a very brief period of time during each AC cycle while DC voltage is close to AC peak value. During this brief time the current flowing is many times the current being supplied by the power supply. So capacitor ESR, connecting cable resistances and perhaps the leakage inductance of the transformer will all limit the speed at which the capacitors can charge. I can try to improve all those things to some extent.

But I'm looking for information on anything else I can do. Would using super fast rectifier diodes help? Is there anything else I can do other than go buy a new 50kg transformer?

I could convert the PS to 3 phase(drop the secondary down to 2000V by connecting the primary in opposing way, then across 3 phases I would get 3400V AC) but I would prefer to keep it single phase for portability.

Please don't suggest in this thread I use a SMPS. This is for a "retro style" vacuum tube project. A linear PS is the only way to go for this.

For 450V drop with 1A and 8uF you need 140Hz.

[Fflint]

I'm looking for a way to improve voltage stability of a linear (non regulated) 3. 5kW power supply.

The power supply powers a vacuum tube amplifier that requires 3300V DC. It is built using 4 small 900W transformers with 2350V secondaries. The tube requires 1A of current.

It is a straightforward linear power supply with a full wave bridge rectifier and 33uF of filtering capacitors. This produced a no load voltage of around 3300V which quickly dropped under load down to 2300V DC. So additional 470V and 230V transformers were connected in series raising the AC voltage to 3050V. This of course drops to 3050V when 1A DC is flowing.

I thought the only way to increase the DC voltage was to increase the AC side, but then I was told about other peoples power supplies using for example a single 4kW transformer with a 2800V secondary and only 8uF of filtering and this PS voltage never drops below 3500V with 1A of current.

So I'm looking for ways to improve my PS so my voltage doesn't drop that much. I realise the rectifier and the transformers are charging the filtering capacitors only for a very brief period of time during each AC cycle while DC voltage is close to AC peak value. During this brief time the current flowing is many times the current being supplied by the power supply. So capacitor ESR, connecting cable resistances and perhaps the leakage inductance of the transformer will all limit the speed at which the capacitors can charge. I can try to improve all those things to some extent.

But I'm looking for information on anything else I can do. Would using super fast rectifier diodes help? Is there anything else I can do other than go buy a new 50kg transformer?

I could convert the PS to 3 phase(drop the secondary down to 2000V by connecting the primary in opposing way, then across 3 phases I would get 3400V AC) but I would prefer to keep it single phase for portability.

Please don't suggest in this thread I use a SMPS. This is for a "retro style" vacuum tube project. A linear PS is the only way to go for this.

For 450V drop with 1A and 8uF you need 140Hz.

You're talking about ripple. I'm talking about voltage drop. They are two different things. Tetrode vacuum tube amps don't care about ripple (that much). However, there was 16uF there not 8uF. My mistake. There ware 4 oil filled caps 4uF each. Also the secondary was 3kV there, not 2800V.

Here is a schematic.


The transformer is a 6kV (center tapped so can be used as 2x3kV) rated at 0.8A.

I see it even has a 50 ohm resistor between the caps and the rectifiers. I checked the numbers with the owner. At 750mA current it drops to only 3850V!

[jonpaul]

The formulas and charts for cap/indr filtered rectifiers are Schades's curves, from GE, 1930s.

See the RCA Radiotron Designers manual, back issues of ARRL, QST, etc on plate supply design.

Unreg PSU regulation depends on the DCR of the chokes and transformer and leakage L.

20% reg is normal,


Jon

[TimFox]

An excellent tool for traditional power supply design is "PSUD2" freeware from Duncan Laboratories.
https://www.duncanamps.com/psud2/
To use this tool (which includes many rectifier models for solid state and vacuum rectifiers), you need to know the open-circuit voltage of the transformer and its effective secondary resistance, as well as the ESR of the capacitors and choke.  (It does not handle "swinging chokes" where the inductance varies greatly with DC current.)  It includes C, L-C, and C-L-C filter circuits.
It allows a choice of resistive, current, and stepped current loads.
I find it very useful to estimate ripple current in the capacitors, rectifier current, primary current, start-up waveform of the output, and output ripple voltage.

[m k]

I'm looking for a way to improve voltage stability of a linear (non regulated) 3. 5kW power supply.

The power supply powers a vacuum tube amplifier that requires 3300V DC. It is built using 4 small 900W transformers with 2350V secondaries. The tube requires 1A of current.

It is a straightforward linear power supply with a full wave bridge rectifier and 33uF of filtering capacitors. This produced a no load voltage of around 3300V which quickly dropped under load down to 2300V DC. So additional 470V and 230V transformers were connected in series raising the AC voltage to 3050V. This of course drops to 3050V when 1A DC is flowing.

I thought the only way to increase the DC voltage was to increase the AC side, but then I was told about other peoples power supplies using for example a single 4kW transformer with a 2800V secondary and only 8uF of filtering and this PS voltage never drops below 3500V with 1A of current.

So I'm looking for ways to improve my PS so my voltage doesn't drop that much. I realise the rectifier and the transformers are charging the filtering capacitors only for a very brief period of time during each AC cycle while DC voltage is close to AC peak value. During this brief time the current flowing is many times the current being supplied by the power supply. So capacitor ESR, connecting cable resistances and perhaps the leakage inductance of the transformer will all limit the speed at which the capacitors can charge. I can try to improve all those things to some extent.

But I'm looking for information on anything else I can do. Would using super fast rectifier diodes help? Is there anything else I can do other than go buy a new 50kg transformer?

I could convert the PS to 3 phase(drop the secondary down to 2000V by connecting the primary in opposing way, then across 3 phases I would get 3400V AC) but I would prefer to keep it single phase for portability.

Please don't suggest in this thread I use a SMPS. This is for a "retro style" vacuum tube project. A linear PS is the only way to go for this.

For 450V drop with 1A and 8uF you need 140Hz.

You're talking about ripple. I'm talking about voltage drop. They are two different things. Tetrode vacuum tube amps don't care about ripple (that much). However, there was 16uF there not 8uF. My mistake. There ware 4 oil filled caps 4uF each. Also the secondary was 3kV there, not 2800V.

Here is a schematic.


The transformer is a 6kV (center tapped so can be used as 2x3kV) rated at 0.8A.

I see it even has a 50 ohm resistor between the caps and the rectifiers. I checked the numbers with the owner. At 750mA current it drops to only 3850V!

DC component + RMS of AC component?
Then "never under" is not accurate.

So 3555V minimum at plate connection.
Maybe it works with 3300V but then it must be pretty DC.
I think you are quite a bit stuck if your unloaded power is 3300V.

[Fflint]

The transformer is a 6kV (center tapped so can be used as 2x3kV) rated at 0.8A.

I see it even has a 50 ohm resistor between the caps and the rectifiers. I checked the numbers with the owner. At 750mA current it drops to only 3850V!

DC component + RMS of AC component?
Then "never under" is not accurate.

So 3555V minimum at plate connection.
Maybe it works with 3300V but then it must be pretty DC.
I think you are quite a bit stuck if your unloaded power is 3300V.

I suppose he measures it with an analog voltage meter built into the RF deck. The meters in such devices usually have a small capacitor across (1nF) to get rid of any RF. So it is probably the DC component plus half of the ripple which is pretty large with 16uF (390V, he is on 60hz AC)

My unloaded power was 3300DC before I added 705V AC on top. Now my AC is 3000V (same as that PS I spoke to the owner of). My unloaded voltage is 4200V (but I have an oil immersed100k bleeder resistor to bring it down to 4kV). At full current it drops to somewhere around 3100V.

My goal is to have it no lower than 3300V and no higher than 4000V.

The formulas and charts for cap/indr filtered rectifiers are Schades's curves, from GE, 1930s.

See the RCA Radiotron Designers manual, back issues of ARRL, QST, etc on plate supply design.

Unreg PSU regulation depends on the DCR of the chokes and transformer and leakage L.

20% reg is normal,


Jon

Thanks. I'll check out the Radiotron and ARRL books. Regarding QST etc, I have to find some catalogue of old articles first.

I was shocked to find out the power supply I mentioned has 50ohms between the rectifiers and caps. The capacitors I used have a max datasheet ESR of 0.5ohm. So two strings of 13 will have only 3.5ohm. Let connecting cables add another ohm.

Then the secondaries of the transformers are 140ohms (DC) so the total (except the rectifiers) is 39.5ohms. Still less than 50 he has above.

So I have to conclude the deciding factor is the transformer inductance. So the only solution is to add more transformers in parallel. Unfortunately I'm out of transformers with this secondary voltage...

Still I can somewhat compensate for the lack of those 200V by upping the screen voltage. At least until I get more of those transformers.

[m k]

With 33uF you should be fine.

Is your bleeding resistor like those 5x30kohm 30W of original power?

[Wolfram]

What transformers are you using? As suggested earlier, they seem to have pretty significant leakage inductance, causing large variations in the output voltage as a function of the load current. The power and voltage rating sounds similar to that of a microwave transformer, as suggested earlier, and those have magnetic shunts installed between the primary and secondary windings to deliberately increase the leakage inductance.

If this is the case here, you could improve the situation by removing the shunts, or resonating away the leakage inductance using some series capacitance on the transformer output. Keep an eye on the no-load input current and AC voltage/current ratings of all the parts in the secondary circuit in either case.

[jonpaul]

Rebonjour

Normal plate transformers have a tight coupling, low leakage inductance and losses perhaps 1..5 %

Microwave oven and neon sign transformers are designed for current limiting on short circuit with magnetic shunts. The leakage inductance is intentionally high, and the efficiency is lower, with losses 5..25%

As MOT have short duy cycles and lowest cost, the voltage regulation is not,an issue.

Magnetizing current due to primary inductance is not a big factor.

Most power transformer testing and parameters can be found by an open circuit and short circuit test.

Bon courage


Jon