Forums :: NEW! Web Resources :: Features :: Photo Gallery :: Vintage Radio Shows :: Archives
Support This Site: Contributors :: Advertise


It is currently Oct Mon 18, 2021 3:36 am


All times are UTC [ DST ]





Post New Topic Post Reply  [ 83 posts ]  Go to page Previous  1, 2, 3, 4, 5  Next
Author Message
 Post subject: Re: High Voltage Regulator Design Issue
PostPosted: Aug Tue 03, 2021 7:54 am 
Member
User avatar

Joined: Nov Mon 02, 2009 7:01 am
Posts: 5376
Location: Lincoln City, OR 97367
Greetings to Ted and the Forum:

The transistors do have graphs for DC applications. I believe that I am within them. As far as I know, dissipation is just that.... whether is comes in pulses or steadily should not matter as long as the peaks are within the maximums; energy is energy.

I have no idea what the abbreviations that you used in your post mean. An explanation would be helpful.

As far as FET's are concerned, I looked into FET's.... what I found is that without exception, every one of them that I could find actually in stock with a major supplier were specified for pulse service exclusively. All of the graphs in the data sheets were of the nation of turn-on and turn-off times and saturation characteristics. Nowhere could I find graphs of drain current (or source current) versus gate voltage or any other form of transfer function. Without such data, I don't see how I could design (or model) a working linear regulator. If you wish to supply some part numbers with linear operating data available and where to buy them, I will consider the idea. In the meanwhile, I already have the transistors in stock and I will be bread-boarding the circuit as soon as I get a few more bits in the mail.

Your comment about the fold-back is intriguing.... I will take your word for it, but I have no idea how to implement it. As I said in one of my previous posts, I am unhappy with the current limit scheme but I have no idea how to improve it. I was looking for a design that would completely shut down the regulator in case of over current until power was removed and re-applied. Unfortunately, I can't figure out how to trigger an SCR on over-current either.

Thanks for your input.

Regards,

_________________
Jim T.
KB6GM


Top
 Profile  
 
 Post subject: Re: High Voltage Regulator Design Issue
PostPosted: Aug Tue 03, 2021 10:47 am 
Member

Joined: Jan Thu 16, 2020 12:29 am
Posts: 1561
Jthorusen wrote:
As I said in one of my previous posts, I am unhappy with the current limit scheme but I have no idea how to improve it. I was looking for a design that would completely shut down the regulator in case of over current until power was removed and re-applied. Unfortunately, I can't figure out how to trigger an SCR on over-current either.

Thanks for your input.

Regards,


The circuit I posted for you had a few advantages, one that it was short circuit protected, plus it is dead easy to add current limiter (leveler) if you wanted. Additional components shown in red.
If you put a push switch in series with Rb, than you would have to push it to start. If the supply dropped out due to short circuit, or a power outage, it stays latched out until the button was re-pushed.


Attachments:
reg3.jpg
reg3.jpg [ 93.48 KiB | Viewed 832 times ]
Top
 Profile  
 
 Post subject: Re: High Voltage Regulator Design Issue
PostPosted: Aug Tue 03, 2021 11:49 am 
Member
User avatar

Joined: Jun Sun 22, 2008 2:03 am
Posts: 1058
Location: Burke, VA 22015
Jthorusen wrote:
So many beautiful theories wrecked by ugly facts. I cannot put a diode across the choke. It adds 80 volts to the filter output...


I don't see how that is possible, unless the diode is incorrectly put in the forward direction bypassing the choke. In normal operation, the diode in the correct reverse position shouldn't conduct and its effect is negligible. It should only conduct when the choke current is disrupted to clamp the back EMF caused by the collapsing magnetic field. If this simple diode operation is "wrecked by ugly facts" as claimed, all basic circuit theory taught in schools until now needs to be rewritten.

_________________
Binh


Top
 Profile  
 
 Post subject: Re: High Voltage Regulator Design Issue
PostPosted: Aug Tue 03, 2021 4:35 pm 
Member

Joined: May Tue 30, 2006 4:46 pm
Posts: 10923
Location: Santa Rosa, CA
I haven't followed or read all of this sequence but wanted to comment about datasheets.

Quote:
As far as FET's are concerned, I looked into FET's.... what I found is that without exception, every one of them that I could find actually in stock with a major supplier were specified for pulse service exclusively. All of the graphs in the data sheets were of the nation of turn-on and turn-off times and saturation characteristics. Nowhere could I find graphs of drain current (or source current) versus gate voltage or any other form of transfer function


Please see Fig. 1 "Output Characteristics" for this once-popular power FET. http://www.irf.com/product-info/datasheets/data/jantx2n6770.pdf

Quote:
As far as I know, dissipation is just that.... whether is comes in pulses or steadily should not matter as long as the peaks are within the maximums; energy is energy.


Unfortunately, at higher voltages, Bipolars (or FETs) can't handle the same current or dissipation as at lower voltages. "Second Breakdown" is a "bend" in the dissipation (Safe Operating Area) curve.

Data sheets show short pulse widths on the current and power graphs because that's how they are tested in production. The pulse and duty cycle are short to prevent heating, so the manufacturer can say these are 25 C ambient ratings.

A power FET is just a voltage-controlled resistor. It can be used in linear regulators, but these days, almost no one builds linear regulators as a production item.

Rich


Top
 Profile  
 
 Post subject: Re: High Voltage Regulator Design Issue
PostPosted: Aug Tue 03, 2021 9:25 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 4672
Location: Perrysburg, OH, 43551 U.S.A.
Jthorusen wrote:
As far as FET's are concerned, I looked into FET's.... what I found is that without exception, every one of them that I could find actually in stock with a major supplier were specified for pulse service exclusively. All of the graphs in the data sheets were of the nation of turn-on and turn-off times and saturation characteristics. Nowhere could I find graphs of drain current (or source current) versus gate voltage or any other form of transfer function. Without such data, I don't see how I could design (or model) a working linear regulator. If you wish to supply some part numbers with linear operating data available and where to buy them, I will consider the idea. In the meanwhile, I

Littelfuse has linear rated MOSFETs: https://www.littelfuse.com/products/power-semiconductors/discrete-mosfets/n-channel-linear/l2.aspx
For example, the IXT*15N50L2 series, a 500v, 15A part: https://www.littelfuse.com/~/media/electronics/datasheets/discrete_mosfets/littelfuse_discrete_mosfets_n-channel_linear_ixt_15n50_datasheet.pdf.pdf
You can get them from Mouser: https://www.mouser.com/ProductDetail/IXYS/IXTP15N50L2?qs=%2Fha2pyFaduikb82wS2LpIWIH2HJisbxDXxtEW1b6XnTLzebJbYeehg%3D%3D
John

_________________
“Never attribute to malice that which can be adequately explained by stupidity.”
― R. A. Heinlein


Top
 Profile  
 
 Post subject: Re: High Voltage Regulator Design Issue
PostPosted: Aug Tue 03, 2021 10:17 pm 
Member

Joined: Jan Thu 16, 2020 12:29 am
Posts: 1561
The was one thing about energy stored in inductances, like the choke. There is a situation where this can cause transients without intended power cycling.

The classic scenario happens in a car with an alternator. If there is a high alternator current, and a bad connection to the battery (which otherwise was acting a a 14V low impedance voltage clamp), the energy stored in the Alternator's Stator's field, results is a large positive going voltage transient. This is called a "Load Dump" .

An inductor opposes a change in current, so it attempts to keep the current flowing. If the load goes High R then the voltage can peak very high. This is one reason why the environment in a car's electrical system is hostile, and many circuits require TVS devices to save them. Also why many people have built add on project devices using chips like Arduinos and wonder why the fail in that environment some time after they get installed.

In general electronics though, this creates an interesting problem, whenever it is possible that a power supply is in one location and its load in another, separated by cables & plugs & sockets. Then the a load on a power supply, both its transformers & its chokes can be disconnected, for example the load itself switched on and off, or with a plug & socket that can simply have a bad connection, or get disconnected while the supply is running. So "Hot Plugging" if you like, can induce failures by this mechanism.

Ideally an isolated power supply unit can withstand hot plugging, without failing.

In vintage radio work now many electronic versions of the elecro-mechanical vibrator have been made. In the case of the synchronous type they require diodes to replace the secondary contacts. Unfortunately most of the designers thought that a 1N4007 would be suitable and these are what generally get used. But, if there is a poor connection to the pins in the socket, or the unit gets unplugged & plugged in white running, the transformer's collapsing field causes the diodes to break down. It is another example of the load dump effect. So for my units I use two BY448's in series, (total of 4 diodes) to prevent this problem.

Other classic examples of the effect occurred in vintage hybrid car radios, that had a Class A audio output transistor, typically a 2N441 with a choke as the collector load. Without the speaker load absorbing energy and with high drive the output transistor could saturate and the choke store enough energy to peak the collector voltage high enough to destroy the transistor. Also in some cases if the transistor's collector (its case) got momentarily shorted to the radios chassis, this would energize the choke, and when the short broke again, the peak voltage destroyed the transistor.These sorts of Hybrid radios often turned up at service centers with destroyed output transistors for that reason.


Top
 Profile  
 
 Post subject: Re: High Voltage Regulator Design Issue
PostPosted: Aug Wed 04, 2021 3:02 am 
Member
User avatar

Joined: Nov Mon 02, 2009 7:01 am
Posts: 5376
Location: Lincoln City, OR 97367
Greetings, All!

Thanks again for your input. I will try to answer a few queries and supply a bit more information; hopefully there is light at the end of the tunnel.

For bb.odin: I believe the simulation is correct. The reason for this is that there is considerable ripple at the input of this choke. It is not dealing with steady-state DC. The addition of the diode effectively halves the inductance of the choke. On positive peaks of the input, the choke resists a change in current by storing energy in its magnetic field. On negative peaks, the energy in that field comes back out as the choke tries to maintain the current flow to the load. Only, with a diode across it, it can't. Thus, the output of the filter shows considerably more ripple and a higher voltage, since only one half of the smoothing action is available.

For Rich: The low power device that I am using (KSC5026) does have a "Safe Operating Area" graph which I am within (I think). The pass device (KSC5603) has a number of graphs with DC parameters on them, but there is none with a safe operating area.... I guess one is supposed to interpret that for themselves. Hopefully, the 800 volt Vceo rating moves it out of consideration for this supply (peak collector voltage around 400 volts or so). Here's the data sheet for your (and anyone else interested) perusal:

Attachment:
ksc5603d datasheet.pdf [313.93 KiB]
Downloaded 13 times


Just in case anyone is interested in data on the other device I am using, here is its data sheet:

Attachment:
ksc5026m-d.pdf [228.97 KiB]
Downloaded 10 times


For John: I will keep the FET you referenced in mind. At $10 a throw, I can't afford to blow up too many of them! :D
I think I will try to go with the bipolar devices I have for now and see how it all turns out.

I had no idea about how to design a true fold-back current limiter. I went with a compromise solution: a 10 ohm resistor in series with the output and a 10 uF capacitor after it (but before the fuse) to ground. According to the simulator, while none of the transistors are stressed much by a sudden short on the output, the addition of this resistor and capacitor results in a number of 5 to 6 amp current spikes through the 1/2 amp fuse.... which should solve the problem.

In the meanwhile, however, my brother Jerry found the design details for a fold-back current limiter.... including the formulae for determining the resistors required. I will be playing with this in the simulator this evening; hopefully I can eliminate my rather crude fuse blower. :D Here's the URL for anyone interested:

https://www.electronics-notes.com/artic ... ircuit.php

Again, I will keep everyone posted as things develop. Thank You to all who have posted and thus helped with the progress.

73,

_________________
Jim T.
KB6GM


Top
 Profile  
 
 Post subject: Re: High Voltage Regulator Design Issue
PostPosted: Aug Wed 04, 2021 8:59 am 
Member
User avatar

Joined: Nov Mon 02, 2009 7:01 am
Posts: 5376
Location: Lincoln City, OR 97367
Greetings, All!

According to the simulator, I have a finalized design. It seems to meet all the criteria. I will have to order the parts and see what happens. I will let you know what happens. In the mean time, here is the schematic:

Attachment:
Regulator All Discrete Darlington V11.JPG
Regulator All Discrete Darlington V11.JPG [ 293.26 KiB | Viewed 773 times ]


Here's how it looks in the simulation:

Attachment:
Regulator All Discrete Darlington Vtest7 graph.JPG
Regulator All Discrete Darlington Vtest7 graph.JPG [ 348.68 KiB | Viewed 773 times ]


The test circuit was set up with a time controlled switch across the load resistor. The switch closes after 1 second for 1/2 second and then opens again. The 1/2 second short condition is shown in the center of the graph. All of the transistors have their dissipation shown; however mind the scales.... the dissipation for Q6 for example is almost too low to show. All of the active devices have separate scale factors. Q4 is the pass transistor, Q5 is the first half of the Darlington and Q6 is the current limiting gadget. The regulated output voltage is in red and the input to the regulator (power supply filter output) is in blue. These are on the same scale.

It turns out that it is a non-trivial juggling act to arrive at the values of the three resistors involved in the fold-back scheme. They all interact and further, they interact with the regulation, so that some ratios that are valid according to the design mathematics do not work in the real world. If it were not for simulation software (and if I were working for the defense department) then this would be one of those classic cost and schedule overruns. :D

As I mentioned earlier, WFP (Waiting For Parts). I will keep y'all posted.

Thanks to all,

_________________
Jim T.
KB6GM


Top
 Profile  
 
 Post subject: Re: High Voltage Regulator Design Issue
PostPosted: Aug Wed 04, 2021 9:05 am 
Member
User avatar

Joined: Nov Mon 02, 2009 7:01 am
Posts: 5376
Location: Lincoln City, OR 97367
Greetings to the Forum:

Another case of foot-in-mouth disease. I forgot to check the regulation with the lower load current represented by the receive condition. It is well out of regulation at around 300 volts. It looks as though I will be juggling some more resistor values. If I can't arrive at a point where it can handle the differing load, I will revert to the brute force approach.

I will keep you posted.

Thanks,

_________________
Jim T.
KB6GM


Top
 Profile  
 
 Post subject: Re: High Voltage Regulator Design Issue
PostPosted: Aug Wed 04, 2021 2:08 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 21396
Location: Warner Robins, GA
Instead of a diode across the choke why not try a higher value resistor.

I cannot remember for sure, but I think I read somewhere that in some designs there's a cap and resistor in series across a choke to limit the high peak voltage that normally happens when the load is removed from a choke.

Also in some instances a capacitor is placed across a choke which forms a filter and makes the choke do a better job at removing the power supply ripple.


Top
 Profile  
 
 Post subject: Re: High Voltage Regulator Design Issue
PostPosted: Aug Wed 04, 2021 5:46 pm 
Member

Joined: Jan Thu 16, 2020 12:29 am
Posts: 1561
Tube Radio wrote:

Also in some instances a capacitor is placed across a choke which forms a filter and makes the choke do a better job at removing the power supply ripple.


They did this in early transistor Japanese portable transistor TV's, like your Sony Micro TV. I have others. It was a popular technique once to raise the impedance of the choke at the ripple frequency.

Although the idea of tuning it up goes against the notion of damping it, it helps because the high value of the capacitance, compared to the choke's much smaller self capacitance,lowers both the amplitude and frequency of the voltage that appears across the choke-capacitor terminals, when the choke's core is energized and the choke suddenly finds itself in a condition of no load. So it may be a better solution.


Top
 Profile  
 
 Post subject: Re: High Voltage Regulator Design Issue
PostPosted: Aug Wed 04, 2021 6:59 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 21396
Location: Warner Robins, GA
Perhaps so.

Would take some experimenting with the value of capacitor to select the right value that works the best in circuit.


Top
 Profile  
 
 Post subject: Re: High Voltage Regulator Design Issue
PostPosted: Aug Wed 04, 2021 10:18 pm 
Member

Joined: Jan Thu 16, 2020 12:29 am
Posts: 1561
Tube Radio wrote:
Perhaps so.

Would take some experimenting with the value of capacitor to select the right value that works the best in circuit.


One would start by calculating the required capacitor with the usual formula for resonance, typically at twice the line frequency, if its being driven by a full wave rectifier.

Before silicon diodes and zeners clamps were possible & popular, a capacitor was commonly used on coils and chokes to lower the resonant frequency when the core's magnetic field collapsed.

The best example of this is the Kettering Ignition system (magnetic discharge or MDI), where generally a 0.2uF capacitor is added to the primary circuit. Without it, with a standard ignition coil, the peak primary voltage is initially a narrow spike about 450V peak, but with the usual 0.2uF capacitor, it is about 200V and wider at its base. This helped prevent sparking when the contacts opened. It did slow the high voltage rise time though on the coil's secondary, compared to say CDI systems. So the people marketing CDI's in the 1960's used this as a "marketing angle" arguing there was less energy lost in fouled plugs. Later though, when a high voltage silicon transistors were used, the capacitor was removed, the high voltage rise times of CDI and MDI became near equal, because the high voltage rise time is limited by the coil's properties of inductance, leakage inductance for CDI, and the secondary's self capacitance and the capacitance of the EHT wiring & plugs.


Top
 Profile  
 
 Post subject: Re: High Voltage Regulator Design Issue
PostPosted: Aug Thu 05, 2021 5:56 pm 
Member
User avatar

Joined: Nov Mon 02, 2009 7:01 am
Posts: 5376
Location: Lincoln City, OR 97367
Greetings to the Forum:

After fourteen versions, each of which had at least ten iterations, I think I have it. I have punished it with the simulator numerous times, which resulted in the addition of some resistors, capacitors and protection diodes, but I think it now satisfies all the criteria.

It regulates properly at the expected loads, although it tends to drop out of regulation (resulting in about four volts of ripple) at 25% over-current. It can withstand a short and goes into fold-back, limiting the dissipation on the pass transistor. It can withstand an open circuit, although the dissipation on the zener diodes is close to their maximum ratings in that case. It can withstand an open followed by a suddenly applied load followed by another open, although there is a spike to almost 300 volts on the output when the second open occurs. I doubt that this will bother the radio, but it may blow the fuse if both zeners in the over-voltage protection circuit are 5% low in rated voltage.

The schematic in its final test version (with a timed switch to open and close the load) is supplied below.

BTW, if anyone is familiar with MicroCap-12, I'd appreciate an answer to the following question: How does one make a normally closed switch? The help files state that the switch control syntax can select either, but nowhere (that I can find) does it tell one how to do this. This is why the test I ran consists of the load being removed, applied and removed again, rather than the other way around. Many thanks to anyone who knows how to do this and will let me in on the secret..... or can assure me that the help file is wrong and that it can't be done.

Here's the final (I hope) version:

Attachment:
Regulator All Discrete Darlington Vtest14.JPG
Regulator All Discrete Darlington Vtest14.JPG [ 274.22 KiB | Viewed 721 times ]


I'll let y'all know how well it works in the real world as soon as I get it built.

BTW, I don't know if the Forum will let me post a MicroCap-12 circuit file. If it will, I will add it to one of my posts. If not, and anyone out there in radio-land has MicroCap-12 and wants to try it for themselves, PM me with your E-mail address and I will send you the circuit file.

Thanks again to all for all the helpful ideas and comments.

73,

_________________
Jim T.
KB6GM


Top
 Profile  
 
 Post subject: Re: High Voltage Regulator Design Issue
PostPosted: Aug Thu 05, 2021 7:29 pm 
Member
User avatar

Joined: Jun Sun 22, 2008 2:03 am
Posts: 1058
Location: Burke, VA 22015
Congrats on your final design. The simple two-resistor voltage divider that allows the constant-current error amplifier to fold-back is clever. Have you simulated your circuit at different temperatures? Especially at freezing temperatures to ensure a cold start.

Have never used MicroCap-12. To make a voltage-controlled switch, you can use a power N-channel MOSFET on the low side of the load. Drive the gate with a DC voltage to turn on the switch continuously or pulse it as you wish.
Quote:
For bb.odin: I believe the simulation is correct. The reason for this is that there is considerable ripple at the input of this choke. It is not dealing with steady-state DC. The addition of the diode effectively halves the inductance of the choke. On positive peaks of the input, the choke resists a change in current by storing energy in its magnetic field. On negative peaks, the energy in that field comes back out as the choke tries to maintain the current flow to the load. Only, with a diode across it, it can't. Thus, the output of the filter shows considerably more ripple and a higher voltage, since only one half of the smoothing action is available.

Since I was not so sure about this, I ran a simulation of the following pi filter to study its behavior when its input is not DC but looks more like a full-rectified sinewave. The following discussion doesn't add any usefulness to your design. My apology if it makes this topic's discussions longer than necessary.

Attachment:
HVSupply_pi_filter.jpg
HVSupply_pi_filter.jpg [ 60.77 KiB | Viewed 714 times ]

The figure below shows the pi filter's input voltage V(a) with its output B loaded. The red plot is for the choke without the clamping diode. The blue plot is for the choke with the clamping diode.
Attachment:
HVSupply_pi_filter_Vin.jpg
HVSupply_pi_filter_Vin.jpg [ 152.28 KiB | Viewed 714 times ]

When V(a) is higher than the output voltage V(b), current flows at a positive rate through the choke to charge all output capacitors. When V(a) is lower than V(b), V(a) continues the follow the shape of a full-rectifed sinewave as shown in the red plot. With a diode across the choke, the negative-rate current in the choke flows mostly through the diode. Therefore V(a) cannot fall below V(b) plus the diode foward voltage as shown in the blue plot. The diode effectively half-wave rectifies V(a) with respect to V(b). The input ripple is thus significantly reduced.

The current in the choke is shown below.
Attachment:
HVSupply_choke_current.jpg
HVSupply_choke_current.jpg [ 144.97 KiB | Viewed 714 times ]

With the diode, the steady-state ripple current through the choke is reduced by about a factor of four. This can be explained by the fact that the choke's stored energy is discharged at a small and almost constant voltage which is equal to the diode forward voltage. As expected the ripple current looks more triangular than sinusoidal. Since the ripple current is reduced, one can argue that the clamping diode magnifies the choke's apparent inductance by the same factor.

The clamping diode current is shown below.
Attachment:
HVSupply_diode_current.jpg
HVSupply_diode_current.jpg [ 125.12 KiB | Viewed 714 times ]

The diode current is quite large. The energy to sustain this current is supplied mostly by the choke. Since the choke's energy must be stored and released at the same rate to maintain equilibrium, its average DC current settles to a higher value as seen in a previous plot. This implies that filter's average output voltage is also higher with the diode.

The filter's output voltage V(b) is shown below.
Attachment:
HVSupply_pi_filter_Vout.jpg
HVSupply_pi_filter_Vout.jpg [ 112.92 KiB | Viewed 714 times ]

The output ripples are about the same in both cases. However with the diode, the triangular ripple is less desirable for RF circuits since it has more high frequency contents.

The following plot shows the filter's output voltage with a load initially applied for 150ms and then disconnected for 150ms. The diode removes the transients when the load is disconnected.
Attachment:
HVSupply_Vout_with_load_on_off.jpg
HVSupply_Vout_with_load_on_off.jpg [ 123.71 KiB | Viewed 714 times ]

The diode's effectiveness is more drastic when a 47uF capacitor is added to the filter's input. The diode removes the transients on turn-offs as well as turn-ons. This is the case where the output ripples are the lowest and the diode provides the most effective protection against accidental load removal when the output capacitor also fails open. The downside is the in-rush current will be higher because of the higher input capacitor. That will put more stress on the transformer.
Attachment:
HVSupply_Vout_with_large_Cin.jpg
HVSupply_Vout_with_large_Cin.jpg [ 118.05 KiB | Viewed 714 times ]

With a 4.7uF capacitor at the filter's input and a 47uF capacitor at the output (same values used in your final design), there are no transients. The diode is not needed and it does more harm than good especially when the transmitter PA is on. The higher ripples are due to the asymmetrical loading of the high voltage source which is splitted in two. When the PA is on, the bottom half is loaded more than the upper half. There would be less ripple if the two halves are equally loaded or the two supplies do not share the same transformer winding.
Attachment:
HVSupply_Vout_with_PA_load_&_large_Cin_Cout.jpg
HVSupply_Vout_with_PA_load_&_large_Cin_Cout.jpg [ 112.96 KiB | Viewed 714 times ]

One may ask: what's the big deal about this pi filter? Things may get complicated and difficult to understand when there is a non-linear element acting like a switch involved. Another example of a circuit that is simple and yet very hard to analyze is a class-E circuit. It has only a voltage-controlled switch, two inductors, two capacitors and a resistive load. To this day, I believe there are still no analytical solutions for that circuit. Thanks to all circuit simulators which allow us to quickly test and understand the behavior of these circuits under various conditions.

_________________
Binh


Top
 Profile  
 
 Post subject: Re: High Voltage Regulator Design Issue
PostPosted: Aug Thu 05, 2021 10:45 pm 
Member

Joined: Jan Thu 16, 2020 12:29 am
Posts: 1561
Binh:

Since you have the simulation set up, it would be really interesting to see what the on-off transients looked like with just a capacitor across the choke, that tuned it to around 120hz.


Top
 Profile  
 
 Post subject: Re: High Voltage Regulator Design Issue
PostPosted: Aug Thu 05, 2021 11:35 pm 
Member
User avatar

Joined: Jun Sun 22, 2008 2:03 am
Posts: 1058
Location: Burke, VA 22015
Hugo:

Here it is. I replaced the diode with a 1.759uF capacitor which resonates with the 1H choke. It's interesting that the transients on load removal are suppressed as effectively as the diode. The ripples are lower too!
Attachment:
HVSupply_Vout_with_resonant_LC.jpg
HVSupply_Vout_with_resonant_LC.jpg [ 110.67 KiB | Viewed 642 times ]

With a 4.7uF capacitor added to the input to mitigate the start-up overshoot and a 47uF output capacitor, the result is amazing. The tuned capacitor cleans up the ripples even with the transmitter PA load applied. This looks like a well-behaved lowpass filter.
Attachment:
HVSupply_Vout_with_resonant_LC_PA_load_&_large_Cin_Cout.jpg
HVSupply_Vout_with_resonant_LC_PA_load_&_large_Cin_Cout.jpg [ 103.87 KiB | Viewed 638 times ]

_________________
Binh


Top
 Profile  
 
 Post subject: Re: High Voltage Regulator Design Issue
PostPosted: Aug Fri 06, 2021 9:41 am 
Member

Joined: Jan Thu 16, 2020 12:29 am
Posts: 1561
Binh,
Thanks for that simulation, very interesting.
Hugo.


Top
 Profile  
 
 Post subject: Re: High Voltage Regulator Design Issue
PostPosted: Aug Fri 06, 2021 11:08 am 
Member

Joined: Jan Thu 16, 2020 12:29 am
Posts: 1561
It seems like this linear regulator problem is already solved.

It may be worth mentioning that there is a completely different way to get the 275V @ 200 mA, from off the shelf parts. If the notion of it takes your fancy.

In the dawn of mobile Tube radio TX/RX gear in the mid 1960's, there was a requirement to generate HT voltages for Tube gear in cars. The basic configuration invented for the task was known as the Royer Oscillator. (it is covered in the ARRL handbooks)

The idea was that you have a pair of power transistors (typically germanium types) driving the primary of a transformer. The oscillations (in the 100Hz to 10kHz range) were sustained by feedback from the transformer windings to the transistors bases. The circuit timing was dependent on the time it took for the primary winding/core combination to begin to magnetically saturate. This is because the induced voltage in the feedback windings is proportional to the rate of change of current with time and this tapers off as saturation begins. One of the transistors, the conducting one, after a time, would start to fall out of saturation. As it did the feedback would diminish and accelerate the process, bringing the other transistor into conduction repeating the 1/2 cycle and generating a square wave drive for the transformer's primary.

The circuit only required a transformer, two transistors and a few resistors. Elegant and simple. And, one wonderful feature, the HT generator using this method was intrinsically short circuit protected, because if the output is overloaded, the feedback is diminished and the transistors shut down.

Many transformers for this task were made by Triad and were quite compact pot core ferrite configurations and they ran around 400 to 1000Hz. They made one for 12V to 375V @ 200mA. For example if the supply voltage was 9V it would produce around 280V. So with this circuit, by varying the supply voltage to the primary circuit you can also control the output voltage.

Here is an example of such a transformer, normally the circuit/schematic sheet comes in the box with the transformer:

https://www.ebay.com/itm/304059591601?h ... SwygVg5vlE


Top
 Profile  
 
 Post subject: Re: High Voltage Regulator Design Issue
PostPosted: Aug Fri 06, 2021 1:05 pm 
Member
User avatar

Joined: Oct Sun 15, 2006 12:57 pm
Posts: 7887
Location: Liberty, Missouri
Late to the party, but I'd also recommend using a FET, and am a big fan of the 11N90 device.

I've used it in SEVERAL HV SUPPLIES and it has performed well. The nice thing about FETs is that there is virtually no static drive current.

I'd also recommend some current limiting, no point in destroying parts needlessly.

_________________
Pics, Bold & underlined text, are usually links.

https://mikeharrison.smugmug.com/

Mike


Top
 Profile  
 
Post New Topic Post Reply  [ 83 posts ]  Go to page Previous  1, 2, 3, 4, 5  Next

All times are UTC [ DST ]


Who is online

Users browsing this forum: Radioroslyn and 5 guests



Search for:
Jump to:  


































Privacy Policy :: Powered by phpBB