UK Vintage Radio Repair and Restoration Powered By Google Custom Search Vintage Radio Service Data

Go Back   UK Vintage Radio Repair and Restoration Discussion Forum > General Vintage Technology > Components and Circuits

Notices

Components and Circuits For discussions about component types, alternatives and availability, circuit configurations and modifications etc. Discussions here should be of a general nature and not about specific sets.

Reply
 
Thread Tools
Old 31st Jul 2020, 5:53 am   #81
Diabolical Artificer
Dekatron
 
Diabolical Artificer's Avatar
 
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 5,356
Default Re: High voltage regulator circuit.

A fet, EG the KSC5026 is a voltage operated device yes? No current flows into the gate, so how does 10uA base current translate between a BC546? This mosfet testing has me foxed on my CT, there's no xV per step and a mosfet is tested in SC mode. Transistors really make my brain hurt, there's so many parameters, a typical Q datasheet is a lot busier than a valves, big steep learning curve.

Thanks, think I tried that, will try again with a fresh 431, might have killed the one I was testing/mucking about with.

Andy.
__________________
Curiosity hasn't killed this cat...so far.
Diabolical Artificer is offline   Reply With Quote
Old 31st Jul 2020, 10:54 am   #82
Diabolical Artificer
Dekatron
 
Diabolical Artificer's Avatar
 
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 5,356
Default Re: High voltage regulator circuit.

I've wired the TL431 as indicated, it works, ta Dave.

I've been thinking how I can drop around 100v so I don't need such high V rated components and came up with this, see attached. It's a potential divider to fix the fets gate at whatever V, so a simple regulator, I'd use this to drop about 80 -100v then connect Dave's regulator circuit to the OP. So far on the bench it works though the R value's may need tweaking. Think of it as a pre regulator regulator.

Can anyone see a better way or poke holes in my circuit please? Andy.
Attached Thumbnails
Click image for larger version

Name:	DSC02674.JPG
Views:	33
Size:	145.8 KB
ID:	212258  
__________________
Curiosity hasn't killed this cat...so far.
Diabolical Artificer is offline   Reply With Quote
Old 31st Jul 2020, 2:35 pm   #83
Refugee
Dekatron
 
Refugee's Avatar
 
Join Date: Apr 2012
Location: Worksop, Nottinghamshire, UK.
Posts: 4,602
Default Re: High voltage regulator circuit.

Would you be able to add a few turns to your mains transformer to get 5 to 12 volts?
You will then be able to power your TL431 without a high value dropper resistor.
There is nothing stopping you powering it independently.
Refugee is online now   Reply With Quote
Old 31st Jul 2020, 3:18 pm   #84
Diabolical Artificer
Dekatron
 
Diabolical Artificer's Avatar
 
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 5,356
Default Re: High voltage regulator circuit.

I have a spare winding, but it's working as it is. What's wrong with the high value dropper?

A.
__________________
Curiosity hasn't killed this cat...so far.
Diabolical Artificer is offline   Reply With Quote
Old 31st Jul 2020, 4:47 pm   #85
Diabolical Artificer
Dekatron
 
Diabolical Artificer's Avatar
 
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 5,356
Default Re: High voltage regulator circuit.

I've encountered an issue, the MJE CCS is getting a bit hot, I think at 10mA Vce290v it's right on the max of the SOA. In which case I've looked for beefier PNP transistors, both mosfet and BJT, there are TO247 devices like the MJL4302A - see attached datasheet for the MJE350 and MJL4302A. if you look at both they look to me to be exactly the same as far as SOA max, the only difference being the latter is a TO264 package so able to dissipate more heat.

I've thought about other ways to get a 300v ish 10mA CCS, not sure how practical some of these are.
1) paralleled CCS's ? See attached. With matched MJE350's is this do-able?
2) Quasi complementary CCS, EG PNP tranny like a MJE350 driving a NPN tranny, think this is a daft idea.
3) Use a valve like a 12BH7
4 Use a MJE350 with biggish HS and little fan.
5)MPSA92/MJE350 darlington CCS

Ideas comments welcome TFL, Andy.
Attached Thumbnails
Click image for larger version

Name:	DSC02677.JPG
Views:	17
Size:	158.6 KB
ID:	212287  
Attached Files
File Type: pdf 1910535.pdf (97.2 KB, 3 views)
File Type: pdf MJE350-D.PDF (139.3 KB, 6 views)
__________________
Curiosity hasn't killed this cat...so far.
Diabolical Artificer is offline   Reply With Quote
Old 1st Aug 2020, 2:30 am   #86
Refugee
Dekatron
 
Refugee's Avatar
 
Join Date: Apr 2012
Location: Worksop, Nottinghamshire, UK.
Posts: 4,602
Default Re: High voltage regulator circuit.

Quote:
Originally Posted by Diabolical Artificer View Post
I have a spare winding, but it's working as it is. What's wrong with the high value dropper?

A.
In short high value droppers get hot.
Also high value resistors tend to be less reliable.
Refugee is online now   Reply With Quote
Old 2nd Aug 2020, 9:13 pm   #87
dave cox
Octode
 
dave cox's Avatar
 
Join Date: Jan 2009
Location: Bristol, UK.
Posts: 1,453
Default Re: High voltage regulator circuit.

Hi Andy, KSC2056 in an NPN medium power device, good for up to 800V, perfect for Q2 I would think.

I think I sent some 43V zeners, you could try one (or more as required) in series with the current source so reduce the voltage/power stress. I don't recall the power rating but the look like it must be more than 10mA will make (~0.5W?). A little bit of space round the zener will needed

dc
dave cox is offline   Reply With Quote
Old 2nd Aug 2020, 9:23 pm   #88
dave cox
Octode
 
dave cox's Avatar
 
Join Date: Jan 2009
Location: Bristol, UK.
Posts: 1,453
Default Re: High voltage regulator circuit.

As a backup, in case the zeners are not practicable for some reason, I'll see if I can spot a sensibly priced / available HV PNP transistor. Failing that maybe use a NPN (not sure about the KSC2056, might be pushing it a bit) if there is a dearth of HV PNP!

dc
dave cox is offline   Reply With Quote
Old 2nd Aug 2020, 9:56 pm   #89
dave cox
Octode
 
dave cox's Avatar
 
Join Date: Jan 2009
Location: Bristol, UK.
Posts: 1,453
Default Re: High voltage regulator circuit.

MJE5852G is a possible, but its a bit hefty/expensive (for what you need) and it still running not a million miles from its secondary breakdown limit.

Maybe the one you have with a few zeners is the best option after all. Usually, with a current source you would not want to cripple its compliance (*) with a string of zeners but in this case it is predictably dropping down from the supply rail to the 24V rail, so as long as there is enough compliance left to soak up any supply voltage fluctuation, it doesn't matter.

Refugee's point about a separate supply rail is sound. If that WERE available we could power the ref and other electronics from it and then there is no issue with heat. If its available, of course

dc

(*) Compliance meaning its ability to swing the output voltage in order to get to the right current. The LTSpice ones will go to multi-mega volts but in the real word they are restricted by the supplied voltage less a 'bit' to operate. If you add zeners then that 'bit' becomes a 'lot'
dave cox is offline   Reply With Quote
Old 3rd Aug 2020, 7:40 am   #90
Diabolical Artificer
Dekatron
 
Diabolical Artificer's Avatar
 
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 5,356
Default Re: High voltage regulator circuit.

Thanks Dave, I knocked the circuit board up yesterday and will do some proper testing, trying to test on breadboard with 100's of volts about isn't optimum.

Regarding the MJE350 CCS I looked into and tried a few circuits, namely a Sziklai pair CCS, and a Darlington CCS, the first baffled me and I couldn't get it to work, the second worked ok but I only had a 100v PNP smaller better beta Q, volage drop was equal between it and the MJE350. In the end I've gone for a gurt big heatsink mounted off board, Q sandwiched between chassis and said HS.

Looking at several datasheet's for high voltage Q's all show about 10mA max at about 300v for all Q's whether they were rated at 300v Vce or 1000v Vce, this even though your not exceeding the max power rating. For instance the MJE350 is rated at 20w @ 25 deg C, running at 270v Vce x 0.010A is 2.7W, however it was definately hotter than 25 deg. That said I've used them and the MJE340 at around the same level without issue, perhaps I'm worrying too much.

As things stand I've used my potential divider with mosfet "follower" to drop 100v ish, will see how it stands up today. more later, Andy.
__________________
Curiosity hasn't killed this cat...so far.
Diabolical Artificer is offline   Reply With Quote
Old 3rd Aug 2020, 12:09 pm   #91
dave cox
Octode
 
dave cox's Avatar
 
Join Date: Jan 2009
Location: Bristol, UK.
Posts: 1,453
Default Re: High voltage regulator circuit.

The reduced power at higher voltage is due to 'secondary breakthrough'. This was the PLAGUE of early power transistors and gave them a reputation of going up in smoke! Things are much better now but at HV it is poking its head up!

I've read that its due to 'defects' in the semiconductor crystal causing a slight current hot-spot. As the resistance drops with increasing temperature you get even more current in a runaway effect. The quality of crystals has improved over the years and thermal conductivity irons this out, somewhat, but not entirely MOSFETS tend to increase resistance with temperature so don't suffer, AFAIK.

dc
dave cox is offline   Reply With Quote
Old 3rd Aug 2020, 12:37 pm   #92
dave cox
Octode
 
dave cox's Avatar
 
Join Date: Jan 2009
Location: Bristol, UK.
Posts: 1,453
Default Re: High voltage regulator circuit.

Ignore my earlier post about using NPN for the HV current source, that would get complicated

There are a few P MOSFETS available up to 500V (SMC) but the only T0220 is at 400V here. A combination PNP transistor and P-MOS cascode looks possible but its getting a bit complex. Is there no way of having an extra power rail ?

dc
dave cox is offline   Reply With Quote
Old 3rd Aug 2020, 12:38 pm   #93
Radio Wrangler
Dekatron
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 14,785
Default Re: High voltage regulator circuit.

Good summary, Dave.

Think of any big semiconductor as a lot of small ones in parallel, then ask how fair sharing of currents and heat is achieved between different areas of the same part.

As we are now in the 21st century, with switch-mode everythings to the right and left of us, the focus is on high voltage semiconductors and high current semiconductors, but not on both at once - high power semiconductors. So trying to find something that will do a bit of heat is becoming difficult.

David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done
Radio Wrangler is online now   Reply With Quote
Old 3rd Aug 2020, 5:08 pm   #94
Diabolical Artificer
Dekatron
 
Diabolical Artificer's Avatar
 
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 5,356
Default Re: High voltage regulator circuit.

Think I misinterpreted Ref's suggestion Dave, what your going all going on about is chucking the current sources and powering the Vref and Q's base from a lower voltage rail yes? I do have a spare 29v AC winding, good for about an amp.

At present I have the potential divider/follower dropping 100v, which consists of a 100k and 220k from HT to gnd, junction of these two R's to a mosfet gate, HT to drain, source out to Dave's PSU, am using a IPW60R330P6 that Dave sent me as for the mosfet. Am using a SiHG30N60E as the mosfet in the regulator, both very beefy TO247 fets. this selection based on pure guesswork. Apropos of nothing, looking at mosfet datasheets and others of similar type it says some are D series, others E series, not sure what that signifies.

After correcting a few cockup's I'd made building the board and debugging it's working well so far, see pics. note some parts are temporary and tacked in for debugging. Will give it some proper gyp tomorrow but so far shorting the OP results in no smoke just reduced V out.

Andy.
Attached Thumbnails
Click image for larger version

Name:	DSC02684.JPG
Views:	23
Size:	162.1 KB
ID:	212592   Click image for larger version

Name:	DSC02685.JPG
Views:	23
Size:	162.4 KB
ID:	212593   Click image for larger version

Name:	DSC02686.JPG
Views:	22
Size:	158.6 KB
ID:	212594  
__________________
Curiosity hasn't killed this cat...so far.
Diabolical Artificer is offline   Reply With Quote
Old 3rd Aug 2020, 5:37 pm   #95
dave cox
Octode
 
dave cox's Avatar
 
Join Date: Jan 2009
Location: Bristol, UK.
Posts: 1,453
Default Re: High voltage regulator circuit.

Wow !
Good work Andy, 150% for effort and willingness to experiment! Especially shorting the output - just how many hail-mary's does that take

That 29V a.c supply would be perfect to avoid the difficult current source. With an extra rail available, there is a 'slightly different' schema that could be employed by creating a negative rail - this would have the advantage of being to regulate down to (near) zero volts. i don't know if that is something that would be useful to you ?

dc
dave cox is offline   Reply With Quote
Old 3rd Aug 2020, 11:55 pm   #96
dave cox
Octode
 
dave cox's Avatar
 
Join Date: Jan 2009
Location: Bristol, UK.
Posts: 1,453
Default Re: High voltage regulator circuit.

Way back, Silicon gave this advice ...
Quote:
Reverse biased power diodes across the output terminals and the pass transistor are usually recommended.
This will offer protection from a capacitor in the load trying to discharge into the PSU when you turn off. This would cause the output MOSFET get reverse biased. I guess the 'reverse' diode across the output is there to protect from reverse charged (film) cap when you connect up


dc
dave cox is offline   Reply With Quote
Old 4th Aug 2020, 11:38 am   #97
Diabolical Artificer
Dekatron
 
Diabolical Artificer's Avatar
 
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 5,356
Default Re: High voltage regulator circuit.

Ta. So far have 100r gate resistors and 12v zeners from source to gate. as the load in this case has little capacitance not to worried about diode protection across OP.

Putting this through it's paces but noticed regulation isn't good with increased V in. Have paralleled another 68r across R1. With just 68r as R1 and a 3k3 load from power up puts it into current limiting, however with two 3k R's as load on power up, when powered up, V out = 300v, when I shunt one 3k out, V out regulation remains good, EG V dropped by 100mV. So a 3k3 load from power up needs R1 tweaking.

Am using the present CCS with these tests, will leave it on for an hour and see how hot things get then decide if to implement Ref's suggestion.

Lastly, what sort of tests should I be after doing? Vin +/- 10%, load 50mA, 100mA, 150mA, 200mA kind of thing.?

Andy.
__________________
Curiosity hasn't killed this cat...so far.
Diabolical Artificer is offline   Reply With Quote
Old 4th Aug 2020, 2:50 pm   #98
dave cox
Octode
 
dave cox's Avatar
 
Join Date: Jan 2009
Location: Bristol, UK.
Posts: 1,453
Default Re: High voltage regulator circuit.

Andy,
Regulation is usually measure against ...
(1) Time - as things warm up from cold and after a change in output voltage (settling time)
(2) Changes to input voltage
(3) Changes in load current (repeated at different output voltages ?)

If you are seeing regulation that is not so good, it would be interesting to know how much the gate voltage is moving so I can try and match it up to the spice model. Obviously, once its hitting the current limit its much less interesting as it is not operating as a regulator!

The startup under load issue you see is typical of a 'foldback' current limit. I think I have set the limits too low in the first design (the spice model tops out at ~100mA @400V and 15mA @ 0V) so as you suggest the values need adjusting

R1, on its own, sets the short circuit current. I think your change is safe, it will give something like 30mA into a short (?) but it's the MOSFET that has to do the heavy lifting, so check you are still in the SOA at the voltage across it (and watch the heat!)

R4/R5 I will need to spice if they don't give a sensible uplift in max current as the voltage goes up (40 years ago I could do it in my head but not now ).

dc
dave cox is offline   Reply With Quote
Reply

Thread Tools



All times are GMT +1. The time now is 2:01 pm.


All information and advice on this forum is subject to the WARNING AND DISCLAIMER located at https://www.vintage-radio.net/rules.html.
Failure to heed this warning may result in death or serious injury to yourself and/or others.


Powered by vBulletin®
Copyright ©2000 - 2020, vBulletin Solutions, Inc.
Copyright ©2002 - 2020, Paul Stenning.