The AM Forum
March 28, 2024, 09:39:56 AM *
Welcome, Guest. Please login or register.

Login with username, password and session length
 
   Home   Help Calendar Links Staff List Gallery Login Register  
Pages: [1] 2   Go Down
  Print  
Author Topic: My first MOSFET design — but, will it work?  (Read 14992 times)
0 Members and 1 Guest are viewing this topic.
KK4YY
Member

Offline Offline

Posts: 565


Your best isn't as good as you can be.


« on: November 25, 2018, 10:05:27 PM »

Before I turn a perfectly good MOSFET into a dispiriting wisp of smoke, I thought I'd run my simple as I can make it circuit design past the group.

It's a voltage regulator, I hope, used to drop the voltage on an RF final while maintaining the higher voltage on the modulator. This, to improve the effective voltage ratio of a modulation transformer that has a bit too much step-down. The fact that it may regulate the voltage is secondary in this application, its primary function is to just drop the voltage. Perhaps a simple resistor would work to this end, but what fun would that be? Especially when this circuit also provides a whole string of cool, glowing, regulator tubes! Cool

Okay, I plan to use the same circuit to regulate lower voltages too. Maybe just one 0A2 to get 150V regulated. That's more sensible 'eh? Yeah, yeah, I could use zener diodes too, but I feel bad enough already just using a 'FET. Sad

Anyway, with keeping it simple in mind, and despite all those VR tubes, please see the attached.

Comments and suggestions will be graciously received.

Don


* STP7N95K3 mosfet voltage regulator kk4yy 600.png (801.63 KB, 600x455 - viewed 853 times.)
Logged

All your worries won't add a day to your life, or make the ones you have any happier.
DMOD
AC0OB - A Place where Thermionic Emitters Rule!
Contributing
Member
*
Offline Offline

Posts: 1768


« Reply #1 on: November 25, 2018, 10:42:38 PM »

To avoid any parasitic oscillations i would use a 47 ohm resistor between the gate and the top of the 0.1 uf cap.

It looks as if you have a 16.5 ms turn-on delay built in.

Otherwise it should work fine.

Phil
Logged

Charlie Eppes: Dad would be so happy if we married a doctor.
Don Eppes: Yeah, well, Dad would be happy if I married someone with a pulse.NUMB3RS   Smiley
VE3ELQ
Contributing
Member
*
Offline Offline

Posts: 258


« Reply #2 on: November 26, 2018, 08:28:55 AM »

MOSFETs as source followers make very good simple regulators as you have proposed. Am using them in several of my projects including the station PS for my AM gear which holds to .1 V regulation, not precision but plenty good enough.

Looks good to me with one suggestion.  Ensure there is a bleed path for the gate reference V from gate to source otherwise when the main PS filter caps and output caps bleed down on power off the gate may be left hanging at the reference voltage by the cap which will quickly exceed the gate to source V rating and blow out the gate.  The 2 built in gate zeners are back to back, not sure if that will protect the gate, never used one like that. If not a small 9V zener in series with 100 ohm R from gate to source is all it needs. Does nothing until power off.

73s  Nigel
Logged

A "Tube Free" zone.
WD5JKO
Member

Offline Offline

Posts: 1996


WD5JKO


« Reply #3 on: November 26, 2018, 03:02:58 PM »

Don,

   I attach two other FET regulators I have built. These have been operational in a Sherwood S8000 tube amplifier since 2003. These are for comparative reference to illustrate another mans approach.

   The CCS to the zeners really help out. With the CCS feeding the zener string, this helps with line voltage regulation.

   The closed loop +375v regulator uses a 6.2v zener reference diode. This zener with the base-emitter junction of the error amplifier have equal and opposite temperature drift (each about 2.2mv/degre C) such that when thermally bonded together, the regulator is stable as the circuit warms up. High voltage zeners drift UP in voltage quite a bit as they warm UP.

   I like the 51 ohm resistor on the drain. If the supply is switched on when the output is shorted out, the idea is to insure the FET is in saturation such that something up stream dies first (hopefully the fuse). I was always chicken to try this once the power was turned on. Might work there too. But remember the SS regulators like the ones here usually die when the output power is suddenly shorted to ground..

Jim
Wd5JKO


* SS_HV_REGULATOR.jpg (170.12 KB, 2160x1660 - viewed 450 times.)

* SS_HV_REGULATOR2.jpg (160.39 KB, 2160x1660 - viewed 388 times.)
Logged
PA0NVD
Member

Offline Offline

Posts: 615


Nico and Chappie (Chappie is the dog...)


« Reply #4 on: November 26, 2018, 03:46:43 PM »

I always put a current limit / short circuit protection by adding a resistor in the source of the series FET and a PNP transistor over that resistor Emitter to the input side, Base via a current limiting resistor to the FET side. The collector can be used to pull up the drive voltage of the series FET when the PNP transistor starts conducting effectively limiting the drive to the FET. The extra series resistor is 0,6/Imax
That circuit did save me a lot of FET's. I made a lot of RF generators for plastic welding and they had all this current limiter.
Logged
KK4YY
Member

Offline Offline

Posts: 565


Your best isn't as good as you can be.


« Reply #5 on: November 26, 2018, 05:29:31 PM »

Great replies! I've updated the schematic with the suggestion offered. So far, it's still a nice simple circuit. Smiley

To avoid any parasitic oscillations i would use a 47 ohm resistor between the gate and the top of the 0.1 uf cap.
Done!

The 2 built in gate zeners are back to back, not sure if that will protect the gate, never used one like that.
Those built-in zeners are one of the reasons I selected that device. I hope they work.
From the datasheet...

"The built-in back-to-back Zener diodes have specifically been designed to enhance not only the device’s ESD capability, but also to make them safely absorb possible voltage transients that may occasionally be applied from gate to source. In this respect the Zener voltage is appropriate to achieve an efficient and cost-effective intervention to protect the device’s integrity. These integrated Zener diodes thus avoid the usage of external components."

I like the 51 ohm resistor on the drain.
Done!

I always put a current limit / short circuit protection by adding a resistor in the source of the series FET...
Hopefully the 51 ohm resistor added in the drain will be enough.

Thanks,
Don


* STP7N95K3 mosfet voltage regulator kk4yy 600 a.png (801.63 KB, 600x455 - viewed 631 times.)
Logged

All your worries won't add a day to your life, or make the ones you have any happier.
DMOD
AC0OB - A Place where Thermionic Emitters Rule!
Contributing
Member
*
Offline Offline

Posts: 1768


« Reply #6 on: November 26, 2018, 09:00:39 PM »

The 47 ohm resistor needs to be placed as shown below:

I don't think you need a 51 ohm value for the CL resistor, too much Vdrop.

The 16.5 ms time constant will allow for a graceful ramp-up of voltage, limiting current rate.


Phil - AC0OB


* HV Regulator.pdf (43.13 KB - downloaded 171 times.)
Logged

Charlie Eppes: Dad would be so happy if we married a doctor.
Don Eppes: Yeah, well, Dad would be happy if I married someone with a pulse.NUMB3RS   Smiley
KK4YY
Member

Offline Offline

Posts: 565


Your best isn't as good as you can be.


« Reply #7 on: November 26, 2018, 09:27:00 PM »

Phil,

Thanks for the schematic! Looks like I don't need to redraw mine again. Grin

I have LTspice on my PC but I haven't taken the time to learn how to use it. If you already have that drawing as a .asc file can you email it to me? Might be a good time for me to learn. My email is on my profile here.

Thanks,
Don
Logged

All your worries won't add a day to your life, or make the ones you have any happier.
KK4YY
Member

Offline Offline

Posts: 565


Your best isn't as good as you can be.


« Reply #8 on: November 27, 2018, 06:05:03 PM »

I don't think you need a 51 ohm value for the CL resistor, too much Vdrop.

Pondering this point while at work today, I wondered why. The regulator drops 100V across the FET and dissipates the associated power. The current limiting resistor on the drain drops voltage too. Why couldn't it be made a relatively large value? Let's say the CL drops 50V (half the total drop) at the rated output current. Then, the CL and the FET would equally share the power dissipated. Moreover, the short circuit current is lowered to a safer value. Yes, this would require a heftier power resistor (there's no free lunch) But the FET would run cooler, or need less heat sink, or it could even be downsized.

Would it adversely effect voltage regulation?

Am I missing something? Huh

Also, thanks for the .asc file Wink

Don
Logged

All your worries won't add a day to your life, or make the ones you have any happier.
DMOD
AC0OB - A Place where Thermionic Emitters Rule!
Contributing
Member
*
Offline Offline

Posts: 1768


« Reply #9 on: November 27, 2018, 06:48:41 PM »

Let's look at the math:

With a 51 ohm CL resistor it drops 12.75 volts, 800-12.75 = 787.25V-700V = 87.25Vdrop across FET.

87.25VX0.25A = 21.8W of FET dissipation.

With a 10 ohm CL resistor it drops 2.5 volts, 800-2.5 = 797.5V at the drain of the FET gives a 97.3Vdrop across FET.

97.5VX0.25A = 24.4W of FET dissipation.

The difference is 2.6W of dissipation.

If you are concerned about dissipation use a physically larger MOSFET with greater power dissipation such as the

https://www.onsemi.com/pub/Collateral/FQA11N90_F109-D.PDF


Phil - AC0OB



* HV Regulator.pdf (43 KB - downloaded 114 times.)
Logged

Charlie Eppes: Dad would be so happy if we married a doctor.
Don Eppes: Yeah, well, Dad would be happy if I married someone with a pulse.NUMB3RS   Smiley
KK4YY
Member

Offline Offline

Posts: 565


Your best isn't as good as you can be.


« Reply #10 on: November 27, 2018, 07:58:33 PM »

Phil,
I muddled my point. Let me try again. Embarrassed

If the purpose of the CL resistor is to protect the FET from a short circuit (in particular), then a higher value would offer better protection.

   CL          Short
 value       circuit
(Ohms)    current (A)
 —                —
  10               80
  51               15.43
200                 4

So the greater the CL resistance, the lower the maximum short circuit current would be.

And, taken to extreme (given a maximum desired steady-state regulator output current), the CL can drop almost all the voltage (with the FET in saturation) not exceeding the point where the regulator drops out of regulation. Additionally, allowing the regulator to drop out of regulation at a current just above that which is desired, it would offer some measure of protection to the load (a mistuned RFPA, for example).

Making the CL large is at the expense of a larger power rating of the CL (a big expensive power resistor) but does seem to have some advantage for certain applications.

My question: Although it may be impractical in most cases, would this be detrimental to the operation of the circuit as a regulator? Huh

Thanks for your patience,
Don
Logged

All your worries won't add a day to your life, or make the ones you have any happier.
DMOD
AC0OB - A Place where Thermionic Emitters Rule!
Contributing
Member
*
Offline Offline

Posts: 1768


« Reply #11 on: November 28, 2018, 01:54:36 AM »

A simple solution is to place a 500 mA AGC Fast Blow fuse in the Drain circuit if you are concerned about shorts.

Too large a RCL may not allow the cold cathode regulator tubes to "light -off" under certain circumstances. For the RCL, you should also consider the 15 mA or so  current in addition to the forecasted 250 mA.

Any circuit is a compromise so run the numbers and make a choice.


Phil - AC0OB



 

* HV Regulator.pdf (45.41 KB - downloaded 127 times.)
Logged

Charlie Eppes: Dad would be so happy if we married a doctor.
Don Eppes: Yeah, well, Dad would be happy if I married someone with a pulse.NUMB3RS   Smiley
WBear2GCR
Member

Offline Offline

Posts: 4135


Brrrr- it's cold in the shack! Fire up the BIG RIG


WWW
« Reply #12 on: November 28, 2018, 09:12:57 AM »


Another factor is the effective impedance of the regulator, as seen by the load. Usually
a lower impedance is considered to be better. A bit depends on what the load acts like,
but keep in mind that "voltage" alone is not the only thing being regulated.

Another idea here is to add a shunt stage... (more complexity)

Or, alternately, put a "sloppy" regulator in front - aka a "pre-regulator" - such as a capacitor
multiplier set to drop say half the requisite Vdrop, then go to the regulator. The preregulator
voltage will go up and down with the line and load a bit, but the actual regulator will take
care of that, and actually the result will be less hum and ripple... plus you've reduced the
Pd per device.

Alternatively, parallel two devices if the current starts to look high, in any event.

Logged

_-_- bear WB2GCR                   http://www.bearlabs.com
KK4YY
Member

Offline Offline

Posts: 565


Your best isn't as good as you can be.


« Reply #13 on: November 28, 2018, 06:09:22 PM »

My thought process (hold on to your hats):

I, and many others, have, an old boat-anchor transmitter with a common PS and a modulation transformer which has a less-than-optimum turns ratio that limits positive peaks. What to do?

Scratching my head to find a solution, my first thought was to use a resistor, between the PS and the mod xfmr, to drop the voltage going to the RFPA. This would "work" but I was thinking that it would be modulating the resistor along with the RFPA. Then, I though of hanging a big filter capacitor after the resistor to prevent that. Maybe that would work better, but a second filter bank is too big and expensive.

This led me to the MOSFET "regulator". Thinking that it may have a low enough impedance to work better in this application. This, I never stated directly, but Bear figured it out!

Another factor is the effective impedance of the regulator, as seen by the load. Usually
a lower impedance is considered to be better. A bit depends on what the load acts like,
but keep in mind that "voltage" alone is not the only thing being regulated.

That's exactly what I need to know. Can the MOSFET rival the original PS (an LC filter in the DX-100, for example) as far as its output impedance which will be seen by the modulator?

Another idea here is to add a shunt stage...
Maybe that would be better.

I haven't the first clue. Anyone?


Too large a RCL may not allow the cold cathode regulator tubes to "light -off" under certain circumstances.
Phil, I must be on the slippery-slope having chosen to use a MOSFET — I'm ordering a 300pc assortment kit of zeners now. So, it's good-bye to ionized gas and hello to avalanche breakdown. I'll change the dropping resistor network to get a proper current for a zener string (probably closer to 5mA).

Don
Logged

All your worries won't add a day to your life, or make the ones you have any happier.
steve_qix
Contributing
Member
*
Offline Offline

Posts: 2599


Bap!


WWW
« Reply #14 on: November 28, 2018, 06:37:40 PM »

I think you just need an "amplified zener" and put the circuit between the modulated DC output and the RF amplifier.  Simple and effective.

Sorry for hand drawn circuit - but something like this may do it - and it's a whole lot simpler.

The group will please let me know if I made any mistakes  Wink


Logged

High Power, Broadcast Audio and Low Cost?  Check out the class E web site at: http://www.classeradio.org
PA0NVD
Member

Offline Offline

Posts: 615


Nico and Chappie (Chappie is the dog...)


« Reply #15 on: November 28, 2018, 07:01:53 PM »

simple current limit / short circuit protection


* current limit.JPG (1693.69 KB, 2592x1944 - viewed 294 times.)
Logged
PA0NVD
Member

Offline Offline

Posts: 615


Nico and Chappie (Chappie is the dog...)


« Reply #16 on: November 28, 2018, 07:36:12 PM »

Amplified zener with short circuit protection / current limit and a stabilizing circuit that can use a low voltage zener, also with current limit and short circuit protection.
R1 defines zener current. R2 should lower the collector voltage of the transistor and is equal to R3. Operating they will take a maximum of 2 mA and the voltage drop over R2 and R3 should leave at least 10 Volts Vc-e for each transistor'
The resistor in the base of the PNP transistor is to protect this transistor during short circuit transients. All transistors are just small transistors that can handle the input voltage of the circuit.
All values are approx. and can be adjusted to the closest normal value


* amplified zener with current limit.JPG (1691.55 KB, 2592x1944 - viewed 318 times.)

* stabilizer with current limit.JPG (1651.79 KB, 2592x1944 - viewed 305 times.)
Logged
steve_qix
Contributing
Member
*
Offline Offline

Posts: 2599


Bap!


WWW
« Reply #17 on: November 28, 2018, 08:00:48 PM »

Has to be a 2 terminal device to be a zener replacement :-).

Logged

High Power, Broadcast Audio and Low Cost?  Check out the class E web site at: http://www.classeradio.org
KK4YY
Member

Offline Offline

Posts: 565


Your best isn't as good as you can be.


« Reply #18 on: November 28, 2018, 08:43:55 PM »

...put the circuit between the modulated DC output and the RF amplifier.
Looks like an elegant solution!
But I don't understand the placement you suggest. There would still be the same voltage on the primary and secondary of the modulation transformer that way, offering no advantage.

I can see using it between the power supply and the "bottom" of the mod transformer secondary (while keeping the full HV on the primary). That would effectively reduce the step-down of the transformer, which is my goal.

Don
Logged

All your worries won't add a day to your life, or make the ones you have any happier.
DMOD
AC0OB - A Place where Thermionic Emitters Rule!
Contributing
Member
*
Offline Offline

Posts: 1768


« Reply #19 on: November 29, 2018, 01:39:26 AM »

Here is a schematic if you want to use zeners:

The Gate node of a Mosfet Source follower is a high impedance node, whereas the Source node of a Source follower is a Low Impedance node.

I have built these types of regulators for Viking and Valiant modulator screen voltage regulators and they work very well.

Of course here I am using the KISS principle.


Phil - AC0OB

* HV Regulator with Zeners.pdf (50.91 KB - downloaded 144 times.)
Logged

Charlie Eppes: Dad would be so happy if we married a doctor.
Don Eppes: Yeah, well, Dad would be happy if I married someone with a pulse.NUMB3RS   Smiley
MikeKE0ZUinkcmo
Member

Offline Offline

Posts: 444



« Reply #20 on: November 29, 2018, 09:18:23 AM »

Another approach would be to use a sensitive gate SCR and just turn off the FET, hopefully saving the "load" as well as the PS.   At shut down only the very few mils of FET bias current would flow to the load.   The trip is set by selecting the value of the "10" Ohm resistor at upper right.    

Of course the regulator could be scrapped and replaced with the OP's stacked VR tubes, and a couple 100K 1W resistors from the VR tubes to the gate.

Logged

Mike KE0ZU

Bold Text and PICS are usually links

https://mikeharrison.smugmug.com/
VE3ELQ
Contributing
Member
*
Offline Offline

Posts: 258


« Reply #21 on: November 29, 2018, 10:11:45 AM »

Here is a schematic if you want to use zeners:
Phil - AC0OB

Now we're talking. Best yet IMHO. Dont need the 47 ohm gate R, V gain is unity it cant oscillate. KISS, build this one.

Adding a low value source R and SCR as Mike suggests is exactly what I do in my LDMOS linear amp MOSFET regulator, uses two SUP90140E FETS, 50V with 50A trip out. Different values but very fast trip and works great.

73s Nigel
Logged

A "Tube Free" zone.
Opcom
Patrick J. / KD5OEI
Contributing
Member
*
Offline Offline

Posts: 8308



WWW
« Reply #22 on: November 29, 2018, 12:10:39 PM »

I think you just need an "amplified zener" and put the circuit between the modulated DC output and the RF amplifier.  Simple and effective.

Sorry for hand drawn circuit - but something like this may do it - and it's a whole lot simpler.

The group will please let me know if I made any mistakes  Wink




This works very well with a bipolar transistor as well. Very versatile and has the possible advantage of tracking the HV supply as the line or other conditions make it vary.

====

Series resistor was mentioned.. maybe with a large cap across it so the load can 'see' the power supply's filter impedance. But no definite regulation except at a certain load.


Logged

Radio Candelstein - Flagship Station of the NRK Radio Network.
KK4YY
Member

Offline Offline

Posts: 565


Your best isn't as good as you can be.


« Reply #23 on: November 29, 2018, 07:11:09 PM »

Wow! Lots of great ideas here. Many thanks!

Phil's "HV Regulator with Zeners" looks very straight forward and would make a good regulator. I'd go with that when I need regulated voltages.

Regulation is not required in my proposed application, and Steve's "Amplified Zener" is a cool two terminal circuit that would provide the Vdrop I need with the fewest parts. When I did a search for "amplified zener" I realized I'd seen the circuit before, used as "B+ reducers", but it didn't come to my mind for this project. Wish I had a good excuse for not recalling it. Embarrassed

I've got 'FETs and zeners on order now, so I'll be anxiously checking my mailbox. My poor DX-100 has been sitting on the desk here, up on her side, for months. I've been modding it on-and-off and any documentation that I did do is scattered around here... somewhere.

I've spent hours watching Jordan Peterson videos recently, but I still keep procrastinating and haven't yet cleaned my room.

Maybe tomorrow... Wink


Don
Logged

All your worries won't add a day to your life, or make the ones you have any happier.
KK4YY
Member

Offline Offline

Posts: 565


Your best isn't as good as you can be.


« Reply #24 on: December 01, 2018, 12:59:51 PM »

In an earlier post, I suggested using a high value current limiting resistor ahead of a MOSFET 'regulator' used to reduce the HV. My thought was to have the resistor dissipate most of the power while also protecting the 'FET.

Although it may actually do this, upon further reflection, I believe that the 'FET won't isolate the resistor from the load. The load, an RFPA, would still 'see' the resistor as part of the load and modulate it — wasting modulator power.

I, contritely, withdraw the suggestion. Embarrassed


Don
Logged

All your worries won't add a day to your life, or make the ones you have any happier.
Pages: [1] 2   Go Up
  Print  
 
Jump to:  

AMfone - Dedicated to Amplitude Modulation on the Amateur Radio Bands
 AMfone © 2001-2015
Powered by SMF 1.1.21 | SMF © 2015, Simple Machines
Page created in 0.086 seconds with 19 queries.