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 23rd Jul 2020, 11:36 am   #61
dave cox
Octode
 
dave cox's Avatar
 
Join Date: Jan 2009
Location: Bristol, UK.
Posts: 1,468
Default Re: High voltage regulator circuit.

A very salient point there about using a MOSFET as a cascode, in this case as Q2.

It 'should be' a drop in replacement for Q2 !

I did a quick parametric search of RS and Farnell / CPC for a suitable BJT for Q2. There is an MPSA44 which is a 400V version of the MPSA40 but that voltage rating is very close to the edge of its specification. There are a few TO92's up to 500V but these are not well characterized.

The KSC5026MOS (TO126 20W) did jump out at me as good choice for the HV current source and is available in small quantities from several big suppliers. It could likely be used at Q2 as well but its a bit of an overkill.I guess the current source could also be done with a MOSFET but its not something I've played with.

dc

EDIT.
Come to think of it, I think my 1st schematic was a folded cascode.
dave cox is offline   Reply With Quote
Old 24th Jul 2020, 7:30 am   #62
Diabolical Artificer
Dekatron
 
Diabolical Artificer's Avatar
 
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 5,452
Default Re: High voltage regulator circuit.

Thanks both. Killed a few mosfets playing with this, mardy devices.( Had to look "mardy" up, thinking I'd dreamed up the word, haven't heard it used for years, another word spellchecker hates) I'm testing at 310v at present, shoving 420v up it's clacker being a daft idea, looking to use a LND150 to drop 100v ish. finding decent HV Q's isn't easy, i keep finding my eyes sliding over to a box of valves.

A question about I3, I'm using a 100k resistor as a sink, would i be better implementing a Q based CCS?

Andy.
__________________
Curiosity hasn't killed this cat...so far.
Diabolical Artificer is offline   Reply With Quote
Old 24th Jul 2020, 11:02 am   #63
dave cox
Octode
 
dave cox's Avatar
 
Join Date: Jan 2009
Location: Bristol, UK.
Posts: 1,468
Default Re: High voltage regulator circuit.

I'm still waiting to see someone post up a valve based regulator !

Andy, I3. A current source in the tail of a LTP isolates the LTP from variations in the power supply rails and greatly improves the common-mode signal rejection. For this design there won't be any varying 'common-mode' signals since one input is the reference and the other is the output voltage (divided down) which won't be changing either, at least when the reg is working! The supply to this LTP is not going to be very good so the tail current will change a bit with supply voltage. That will result in the LTP shifting its balance point as it still needs to maintain the same drive to the output so any change in tail current need to go down the other leg of the LTP. There are scary looking equations describing how an LTP creates increaing non-linearity as it goes out of balance but its sufficient to say balanced = good ! I think for simplicity an R is OK but you're about 10X too high on the R .

There are likely other effect that will limit performance more than the LTP's tail. The voltage ref for example. For a forward biased diode is ~2mV/Centigrade, say at 2.5V, gives ~ 0.1% per C so a 20C cycle would be 2% variation. This is multiplied up to the output voltage so at 300V that's a 6V error (plus whatever error the rest of the electronics/passives adds).Self heating in the output divider chain will likely contribute depending on the type/power of resistors. The next step up for the reference would be a 'bandgap' reference which cleverly eliminates that 2mV/C by using 2 diodes with different current densities, but they have the disadvantage that they don't light up to tell you that they are working

dc
dave cox is offline   Reply With Quote
Old 24th Jul 2020, 1:52 pm   #64
orbanp1
Heptode
 
Join Date: Nov 2009
Location: London, Ontario, Canada
Posts: 526
Default Re: High voltage regulator circuit.

Quote:
Originally Posted by dave cox View Post
I'm still waiting to see someone post up a valve based regulator !
dc
Hi Dave,

Ar you looking for a "new" design or existing equipment?

I have the following unit:
http://bama.edebris.com/download/pac...ion%20B-12.pdf
It came to me as "for parts only". It had a faulty 6AU6 tube in it.
Since then some the electrolytics have expired, I replaced all of them.
Also added a bucking transformer because of the increased mains voltage here.
It is a useful piece of equipment, and has served me well on several occasions.

Regards, Peter
orbanp1 is offline   Reply With Quote
Old 24th Jul 2020, 4:08 pm   #65
dave cox
Octode
 
dave cox's Avatar
 
Join Date: Jan 2009
Location: Bristol, UK.
Posts: 1,468
Default Re: High voltage regulator circuit.

Peter, that looks quite interesting!

Its got an extra -ve rail (or 2) to get the output down to zero (pretty near impossible without that extra rail).

dc
dave cox is offline   Reply With Quote
Old 24th Jul 2020, 8:32 pm   #66
Ed_Dinning
Dekatron
 
Join Date: Jan 2004
Location: Newcastle upon Tyne, Tyne & Wear, UK.
Posts: 6,164
Default Re: High voltage regulator circuit.

Hi Andy, plenty of circuits on the web using 807 and similar. Some of the gas regulator tubes used with these units can be difficult to get these days but zeners can be pressed into service.

As a previous poster said, to go down to zero you need to have a neg rail. It is also helpful to have a voltage dependant current limit circuit, so that you don't kill the valve if you are looking for max current with full volts across the valve (min output). It will at least last longer than a mosfet.

Ed
Ed_Dinning is offline   Reply With Quote
Old 25th Jul 2020, 7:55 am   #67
Diabolical Artificer
Dekatron
 
Diabolical Artificer's Avatar
 
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 5,452
Default Re: High voltage regulator circuit.

Thanks again Dave for the thought and effort your putting into this. Still plugging away at this, got distracted by making a mosfet testing rig. Having to test mosfets as my curve tracer has gone gaga - no step bias.

Re I3 I just divided 376/4mA =94k or 100k, will try something smaller.

A valve based regulator would be easier here however I have space constraints and efficiancy also plays a part. Valve reference's are easily available Ed and look good what with a lot glowing pretty colours, just haven't the space here.

I started down this rabbit hole after the simple fet based "regulator back in post one was found lacking. Several other regulators found on the web and elsewhere were found to be less than perfect, the floating regulator type's suffering from instability.

Found an old schematic Miguel knocked up for me years ago, see attached, this has an extra rapid discharge circuit on it, which could be ommited. Anyhoo, with the ammount of work Dave has put into this I'll get it working.

Andy.
Attached Thumbnails
Click image for larger version

Name:	Complete circuit.jpg
Views:	62
Size:	55.8 KB
ID:	211727  
__________________
Curiosity hasn't killed this cat...so far.
Diabolical Artificer is offline   Reply With Quote
Old 26th Jul 2020, 11:31 pm   #68
dave cox
Octode
 
dave cox's Avatar
 
Join Date: Jan 2009
Location: Bristol, UK.
Posts: 1,468
Default Re: High voltage regulator circuit.

Andy, I love a challenge AND it keeps me entertained while I'm doing any dull paid work e.g. waiting for something to finish! All kudos to you for sticking at this. The fact that there are few designs published might suggest its not so easy. But, It must be possible because there is a 0-350V monster sat under my bench! I would bet this is MOSFET based ...

I see where your calculated value is coming from. Maybe we are crossed up on which is which current source is which. I4, the one on the +400V rail will be dropping most of the voltage (~375V if you used a 24V zener). From that +24V, I3 is supplying the LTP (Q5/6) with the tail (emitters) at ~+3.5V (ref+Vbe) so it needs to drop more like 20V and give 4mA.

BTW, I had a rummage and turned up a big 900V 2SK something MOSFET - no idea where that came from Once I finished searching I'll pop some bits in the post ?!

dc
dave cox is offline   Reply With Quote
Old 27th Jul 2020, 7:36 am   #69
Diabolical Artificer
Dekatron
 
Diabolical Artificer's Avatar
 
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 5,452
Default Re: High voltage regulator circuit.

Bless you Dave...had another crack at this yesterday and have still have a low OP voltage, so pulled out Q2 as suggested, no change. something is dragging the gate down, measured across the current sense R - R1 = 44mV, therefore don't think it's going into current limiting. So pulled Q7 for good measure, still SNAFU'd.

This was after changing the resistor I'm using for CCS I3, which is now 6k, I'm still getting 23/24v across this so now we have 3.8mV, better. I also rebuilt it twice to make sure I'd made no mistakes, the gate is still sitting low, without a load about 80v, so OP about 80v.

Gate still low so at a loss I pulled out R7 - 330k, this is with Q2 & 7 replaced, no OP V. I then changed R7's value to 220k, haha, OP V now higher.

After mucking about changing other component's mainly the resistor divider I've found the circuit does work, twiddling R2 does change OP V, but something is keeping the gate low.

Not being that good with transistors it's taking me a while to get my head round things but I'm slowly creeping up on this. Some thoughts which could be completely wrong. Q2's base is always at 24v, I have a 24v zener as D2, is this right? Looking at a blown up schematic it looks like 30v Whichever this is if Q2's base is at a low voltage surely the gate of the mosfet will also be low? The gate V of the fet seems to me to be mainly dictated by a potential divider D2/Q2/D3/R7 D2 & D3 are fixed (ish) so Q2 Vce and the voltage drop across R7 are the only variable's. However Q2's collector voltage will be set by Q2's base current, but if Q2's base is always at a constant current due to the CCS, so how can Q2 Vce change?

Just thinking out loud here but looking at this another way my circuit differs from yours Dave mainly because I'm only running this off 320v instead of 400 odd V, but don't think that matters.Also V2 and D2 (maybe) are different, but am pretty sure Q4/5/6 are working as per at least for quick tests. Q5 & 6 bases are @ 2.5v so there's no disparity there, if they're right Q4 should be. Q4 is the next candidate for closer inspection, i'll get there....

Andy.

PS thanks for the offer of bits, much appreciated.
__________________
Curiosity hasn't killed this cat...so far.
Diabolical Artificer is offline   Reply With Quote
Old 27th Jul 2020, 10:50 am   #70
dave cox
Octode
 
dave cox's Avatar
 
Join Date: Jan 2009
Location: Bristol, UK.
Posts: 1,468
Default Re: High voltage regulator circuit.

Andy, good detective work! As you discovered that gate resistor is the only thing that turns the MOSFET on.

I'm a bit surprised 330k didn't turn the MOSFET fully on, maybe its a bit damaged! If you go below 100k the power dissipation in the R and Q could get a little high when you dial DOWN the output voltage so beware. I notice the basic Sussex uses 470k in series with a 500k pot on the gate ...

dc
dave cox is offline   Reply With Quote
Old 27th Jul 2020, 11:34 am   #71
dave cox
Octode
 
dave cox's Avatar
 
Join Date: Jan 2009
Location: Bristol, UK.
Posts: 1,468
Default Re: High voltage regulator circuit.

Ahhh, finally figured out how to make a pdf from LTSpice (fingers crossed) ...
Attached Files
File Type: pdf HighVoltage2.pdf (23.9 KB, 42 views)
dave cox is offline   Reply With Quote
Old 27th Jul 2020, 4:20 pm   #72
Diabolical Artificer
Dekatron
 
Diabolical Artificer's Avatar
 
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 5,452
Default Re: High voltage regulator circuit.

Thanks again. After more work am pretty sure my CCS isn't right, the diff amp isn't working right, it's unbalanced and somethings not right round Q2. Q2 is pulling an awefull lot of current.

I pulled out my CCS and subbed it for a resistor as well as subbing R3 for a blue LED so the diff amp is balanced. i've noticed a few odd things happening on the breadboard ( breadboards not the best for HV PSU's) so I could have some oscillation going off, will have a sniff round with the scope tomorrow.

Best, Andy.
__________________
Curiosity hasn't killed this cat...so far.
Diabolical Artificer is offline   Reply With Quote
Old 27th Jul 2020, 5:01 pm   #73
dave cox
Octode
 
dave cox's Avatar
 
Join Date: Jan 2009
Location: Bristol, UK.
Posts: 1,468
Default Re: High voltage regulator circuit.

Q2 could be breaking down.
An alternative is that its oscillating, if so, try ~100R in its base connection.
You could temporarily add a few k ohms into Q2's emitter and/or collector to make measuring the current possible.

Its not going to regulate with R3 replace with a blue LED. The LTP will have equal inputs but there will be no feedback to control the output. Hopefully its Q2 that's the problem.

dc
dave cox is offline   Reply With Quote
Old 28th Jul 2020, 11:57 am   #74
dave cox
Octode
 
dave cox's Avatar
 
Join Date: Jan 2009
Location: Bristol, UK.
Posts: 1,468
Default Re: High voltage regulator circuit.

Quote:
Not being that good with transistors it's taking me a while to get my head round things but I'm slowly creeping up on this. Some thoughts which could be completely wrong. Q2's base is always at 24v, I have a 24v zener as D2, is this right? Looking at a blown up schematic it looks like 30v Whichever this is if Q2's base is at a low voltage surely the gate of the mosfet will also be low? The gate V of the fet seems to me to be mainly dictated by a potential divider D2/Q2/D3/R7 D2 & D3 are fixed (ish) so Q2 Vce and the voltage drop across R7 are the only variable's. However Q2's collector voltage will be set by Q2's base current, but if Q2's base is always at a constant current due to the CCS, so how can Q2 Vce change?
Nearly right

Ignore Q7 and the gate for now, normally no current flow THROUGH the gate and Q7 is turned off until significant current is flowing into the load during an 'over-current' situation so this is a valid simplification. The gate voltage is then determined by R7 and the current flowing through D2/D3/R7 and is calculated as R7 * current (and subtracted from the rail voltage).

You correctly pointed out Q2 base is FIXED at 24V. With more common transistor configurations we think of the base voltage being modulated, resulting in some current/voltage change, but that is only true when the emitter voltage is FIXED. For Q2 this is reversed, the base voltage is fixed and the emitter voltage is modulated! Contrast this with Q4 where the emitter voltage is FIXED and the base voltage is modulated.

Q2 is configured as common base and Q4 is as common emitter. When a common base is in series with a common emitter its typically called a 'cascode', a variation on the word 'cascade' which means one after the other so 'cascode' mean one on top of the other

Dave
dave cox is offline   Reply With Quote
Old 29th Jul 2020, 6:56 am   #75
Diabolical Artificer
Dekatron
 
Diabolical Artificer's Avatar
 
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 5,452
Default Re: High voltage regulator circuit.

Got you Dave. I'm probably thinking in the wrong terms, being more familiar with working with valves. As The Dude said, "this is a complicated case, a lot of ins and outs and what have yous" Sorry I'm being a bit slow, thanks for sticking in there.

Andy.
__________________
Curiosity hasn't killed this cat...so far.
Diabolical Artificer is offline   Reply With Quote
Old 29th Jul 2020, 12:50 pm   #76
dave cox
Octode
 
dave cox's Avatar
 
Join Date: Jan 2009
Location: Bristol, UK.
Posts: 1,468
Default Re: High voltage regulator circuit.

Ooops,
Quote:
and the current flowing through D2/D3/R7
should read Q2/D3/R7 !
dave cox is offline   Reply With Quote
Old 29th Jul 2020, 1:34 pm   #77
dave cox
Octode
 
dave cox's Avatar
 
Join Date: Jan 2009
Location: Bristol, UK.
Posts: 1,468
Default Re: High voltage regulator circuit.

It's interesting to contrast the B12, previously posted by Peter, with this design with respect to how they define the output voltage.

The B12 defines a current, by applying a fixed voltage (145V) to a variable resistance (R20+R19+fraction of R18) and applies that current to a variable resistance (fraction of R18+R16+R21).

This design defines a fixed current by applying a fixed voltage to a fixed resistor (R3) then applies that current to variable resistance (R2+R8). This is simpler only because it does not worry about adjustments of zero (it can't do it) or span. If I were changing this to get to zero volts it would be quite similar!

dc
Attached Thumbnails
Click image for larger version

Name:	contrast.jpg
Views:	19
Size:	46.8 KB
ID:	212129  
dave cox is offline   Reply With Quote
Old 30th Jul 2020, 2:10 pm   #78
Diabolical Artificer
Dekatron
 
Diabolical Artificer's Avatar
 
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 5,452
Default Re: High voltage regulator circuit.

Sussed it!!! After more frustration trying to measure voltages which is made nye impossible by transistor manufacture's playing silly b*ggers with Q labelling , IE EBC, or CBE or BCE or CZQ, aghhhh! Where was I, or yes, sussed it, it was down to my choice of tranny for Q4. At first I tried a BC184, then a BD139, stuck a BC337 in, it all worked, no idea why, must be beta I guess.

After getting frustrated by trying to figure out which Q was on, thus pulling current through R7, I went for the good old "lets pull out wires till something happens (EG gate voltage goes up)" approach. popped a 337 in, worked like a charm.

This design had me foxed "This design defines a fixed current by applying a fixed voltage to a fixed resistor (R3) then applies that current to variable resistance (R2+R8)." all the variable HV PSU circuits I've come across so far use a pot to vary voltage on the base or a gate, couldn't for the life of me work out how varying R2 varies the gate V.

I'm still using a LED for V2 instead of a TL431, I really have no luck with these. I re-read all Dave had posted which helped suss things out, but can't figure how to connect a TL431 without R's. I tried cathode to R10, anode to gnd, ref to Q6 base, anode to gnd, cathode R10/Q6base and get 0.6v?

So now i have things working i'll knock the circuit up on a board and do a bit of testing. Dave has very kindly sent me a few nice bits to set me on me way.

Thanks, Andy.
__________________
Curiosity hasn't killed this cat...so far.
Diabolical Artificer is offline   Reply With Quote
Old 30th Jul 2020, 3:31 pm   #79
dave cox
Octode
 
dave cox's Avatar
 
Join Date: Jan 2009
Location: Bristol, UK.
Posts: 1,468
Default Re: High voltage regulator circuit.

Good news Andy!
Will be interesting to see how it performs.

I threw a KSC5026 (800V To126 NPN) at the curve tracer to see the gain, or lack of. I've attached a comparison to a jelly bean bc546b that was already in the 2nd socket. This is a 30V sweep with 10uA base current steps, the display uses the same scale for both ...

I think the hysteresis on the bc546 traces are due to self heating

dc
Attached Thumbnails
Click image for larger version

Name:	bc546b.jpg
Views:	18
Size:	60.8 KB
ID:	212208   Click image for larger version

Name:	ksc2056.jpg
Views:	17
Size:	50.9 KB
ID:	212210  
dave cox is offline   Reply With Quote
Old 30th Jul 2020, 4:07 pm   #80
dave cox
Octode
 
dave cox's Avatar
 
Join Date: Jan 2009
Location: Bristol, UK.
Posts: 1,468
Default Re: High voltage regulator circuit.

This should produce the required 2.5V ref.
Figure 9 in the TI datasheet shows the test circuit operating at quite a bit more current, ~8mA, with some decoupling in the supply. Lucky them not having to operate from a 400V rail 1mA is specified as the minimum cathode current so 2mA should be good. In fact, figure 4 on the datasheet shows everything flattens off above 0.4mA so 2mA is plenty.

dc
Attached Thumbnails
Click image for larger version

Name:	tl431-2V5.jpg
Views:	23
Size:	35.1 KB
ID:	212214  
dave cox is offline   Reply With Quote
Reply

Thread Tools



All times are GMT +1. The time now is 12:07 am.


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.