User talk:AESprojects

Notes on voltage multiplier starting circuit:

1. Construction must take into consideration that almost 15 KV (in this case) may be available at the output should the tube not start or accidentally become disconnected. Layout the circuitry so that parts with significant voltage differences are widely separated and try to avoid sharp points in  the wiring and solder connections.

Perforated prototyping board or any other well insulated material can be  used. Smooth out all HV connections - avoid sharp points by using extra solder. A conformal coating of high voltage sealer is also recommended after the circuit has been constructed and tested. Together, these will minimize the tendency for corona - which can greatly reduce the available starting voltage (particularly on damp days).

2. Diodes D1 to D7 must be rated at least 3 KV. Fast recovery types are probably required. (The multiplier described in the section: "Sam's line  powered HeNe laser power supply" using normal 1N4007s does not appear to   generate adequate starting voltage when driven by this inverter). If 3 KV  diodes are not available, build them up from four 1 KV diodes (to add   margin if no equalizing resistors and capacitors are used).

3. Capacitors C1 to C7 are 100 pF 3 KV disk type.

HeNe inverter power supply using PWM controller IC: --

This power supply was found in a bar code scanner driving a .5 mW, 85 mm long HeNe tube. Fortunately, only the high voltage section was potted and some icky disgusting rubber material was used which could be removed by picking, chewing, clawing, and scraping, without any serious damage to the underlying circuitry. This is a very compact unit with total dimensions of: 3/4" (W) x 1/2" (H) x 5" (D).

The input voltage range is about 5 to 12 VDC though the minimum will depend on the size of the HeNe tube powered. The output is current regulated and fully protected against a variety of fault conditions.

The power supply has been tested on a variety of HeNe tubes up to 2 mW:

operation is probably possible. Output voltage is about 1,300 VDC. This size tube is what the power supply came with for its intended use as part of a bar code scanner.
 * For a .5 mW, 135 mm tube, nothing becomes excessively hot and continuous

become quite warm but not too hot to touch. The DC output voltage of the supply in this case was about 1,350 VDC. With at most a small heat sink, I would expect the supply to drive this size tube continuously.
 * A 1 mW, 150 mm tube caused the driver transistor and ferrite transformer to

supply). However, a heat sink would definitely be required on the driver  transistor for continuous operation and the ferrite transformer would likely  become hot enough to be damaged in a short time.  Therefore, this is not  recommended.
 * It was able to drive a 2 mW, 200 mm tube (requiring about 1,750 V from the

The current was maintained near the calculated value of 3.2 mA in all cases.

The basic design is quite nice and could be easily modified to drive much larger tubes. The only non-standard part - the ferrite transformer - is also relatively simple to construct (as these things go) with only two windings on a circular bobbin in a gapped pot core.

The design uses an integrated circuit, the Philips SG3524. This is a Pulse Width Modulated (PWM) switchmode power supply controller chip which incorporates a fixed frequency oscillator, ramp generator, error amplifier and comparator, and output drivers. The SG3524 provides regulation as well as over-voltage and over-current protection, and other functions. Through the use of these capabilities, this design should be quite robust in dealing with a variety of fault conditions.

If you want to construct a power supply similar to this one, the SG3524 is readily available from large electronics distributors and places like MCM Electronics and Dalbani but shop around - the price seems to vary widely ($2.45 to $12.50!). Additional information on this part may be found in: AN126 - Applications using the SG3524.

Estimated specifications (IC-I1):


 * Operating voltage: 1,000 to 2,000 V.
 * Operating current: 2 to 4 mA (by changing resistor).
 * Starting voltage: greater than 6,000 V.
 * Compliance range: 1,000 V.

For the bar code scanner application, the HeNe tube and power supply were glued together and mounted as a single unit. The red cap at the far left is a feeble attempt to insulate the high voltage to the HeNe tube (not covered by the gray rubbery potting material just visible over the left half of the power supply. You can still get zapped from under the circuit board (as I found out!).  This unit used a Uniphase HeNe tube.  Another one came with a very similar Melles Griot HeNe tube.

ICI1PS Top View shows the component side of the power supply printed circuit board after the rubbery potting material covering the high voltage section (left half) had been removed. The pot core ferrite transformer is just to the right of center with the IRF630 MOSFET next to it (separated by a filter capacitor). The SG3524 controller IC is located under the IRF630. The bright blue and orange objects are the filter and multiplier capacitors in the high voltage circuitry. The high voltage rectifiers can be seen above and below them. The 99K ballast resistor (3 x 33K) is visible at the far left.

As a result of the sophistication of the SG3524, the overall design is really quite simple. The PWM controller is shown first followed by the inverter:

2N3904                                    R3       Q3 +---+-+---/\/\---+ |  | 2.21K                       |  3.92K   |    R5        |/ C  /    +|--+---/\/\---o CS  VS o--|     \ R1 |      U1 SG3524         |              6.81K |\ E /    |   +--+     | |  |    |  1|              |16   |   Input (+5 to +12 VDC) +---+|-In  Vref Out|-+         o           |   |    |  2|              |15             |               1 o  T1         _|_  / R2 +---|+In        Vin|++-+-++ |:| C3 --- \ 2.74K 3|              |14  |    |           |         15T )|:|    .1 uF |   /     ---|Osc Out    E-B|--- |   _|_ C1      _|_ C4     #26 )|:| |  |       4|              |13  |   --- 6.8 uF  --- 100 uF   2 )|:|          +---++---|+CL Sense  C-B|--- |    |  16 V     |  16 V   +--+ |:|                   |  5|              |12  |   _|_         _|_     D  |                   +---|-CL Sense  C-A|+    -           -    .|---+ Q1                   |  6|              |11          D1           G||<--. IRF630         +-|---|RT         E-A|-+--|>|---+---'|---+         |         |  7|              |10       | 1N4148 |         S  |         |     +---|---|CT    Shutdown|---o OV  |        |            |         |     |   |  8|              |9        |      |/ E Q2        |      R4 /     |   +---|Gnd       Comp|---      +--|    2N3906    |    5.1K \    _|_  |   |              |         |      |\ C           |         /    ---  |   +--+         / R6     |            |         |  C2 |   |                            \ 4.7K   |            | |.001 |  |                            /        |            |         |  uF |   |                            |        |            | +-+---++++--o HV- _|_                  -

3            C6              C8              C9 T1 +--||--+---||--+---||--+ +---++-++||+  +-+-+---||--+--/\/\--+        |    |     |    |          |   |     |     |                  33K   | CS o--+ R7 / R8 /   _|_ C5      |  _|_   _|_   _|_                   R13 / 10K \ 430 \  --- .1 uF   |  ---   ---   ---                   33K \ SBT /    /    |          |   | C11 | C12 | C13    LT1             / |    |    |          |   |     |     |   +--+ R12 33K | HV- o---+-++--+-+-+---|-|      -|---/\/\--+ R9       R10     R11    |        _|_  Tube- +--+ Tube+ OV o---/\/\---+---/\/\/\/\---+        - 13K   |   4.7M    4.7M       D2-D7: 2 KV, fast recovery type. VS o--+                     C6-C8, C10: 1 nF, C9: 4.7 nF, all 3 KV. C11-C13: 1 nF, 6 KV.
 * ( 500T            D2   |  D3      D4   |  D5      D6   |  D7       HV+
 * ( #36          +--|>|--+--|>|--+--|>|--+--|>|--+--|>|--+--|>|--+o
 * ( o 4          |         C7    |               |      C10      |  R14
 * ( o 4          |         C7    |               |      C10      |  R14
 * ( o 4          |         C7    |               |      C10      |  R14

Notes on PWM controller for HeNe power supply (IC-I1): -

1. R4 and C2 set the oscillator frequency, roughly 1/R*C or about 200 kHz. This generates a sawtooth/ramp inside the SG3524. The output of the error amplifier (Pins 1 and 2, -In and +In) is then compared with this ramp to  control the pulse width of the drive to the switching transistor, Q1, which is enabled every other cycle resulting in a switching frequency of 100 kHz.

2. The main feedback loop is from terminal CS (Current Sense) which sets the output current based on the voltage drop across the parallel combination of  R7 (SBT or Select By Test) and R8.

With the installed values for R7 (SBT), the sensitivity is approximately .4 V/mA. The voltage on the +In pin of the SG3524 will then be equal to: 3.24 V - 146 * Iout. The 3.24 V reference is derived from Vref (5 V) and the voltage divider formed by R3, R5, R7, and R8. The factor of 146 comes from the voltage divider formed by R3 and R5 when driven by CS.

3. VS (Voltage Sense) is derived from a point about 1/3 of HV+ and will be  approximately equal to: 1/3 * HV+ * 24K / 9.4M. (11K of the 24K is input  resistance of the SG3524's Shutdown pin) or 8.3E-4 * HV+. The -In pin of  the SG3524 will then be VS - .7 V or 2.77 V (Vref through the voltage   divider formed by R1 and R2) depending on which is greater. The 2.77 V  reference will be in effect under normal conditions. However, if HV+ goes above about 4,200 V, the VS input will take over and limit output even if  no current is drawn (as would be the case before the tube starts or if the   tube were disconnected or did not start).

4. Once the tube starts, the set-point will be where:

-In = +In 2.77 = 3.24 - 146 * Iout

Thus:    Iout = 3.2 mA (for the installed value of SBT).

5. If OV (Over-Voltage) becomes too great, the Shutdown input will activate killing the inverter momentarily - the cycle will then repeat. Shutdown voltage is sensed from a point about 1/3 of the operating voltage, HV+. If  this point exceeds about 700 V, the driver will shut down. During staring this limits the rate of rise of output voltage. Should the tube fail to  start or become disconnected, excessive voltage will not be generated.

Notes on the inverter for HeNe power supply (IC-I1): ---

1. This is a flyback inverter where the length of time the driver transistor (Q1) is on determines how much energy will be transferred to the high voltage circuitry when it switches off. The SG3524 drives the MOSFET's  gate via D1. Q2 is used to turn off the MOSFET quickly by discharging the gate capacitance to ground.

2. T1 is a ferrite transformer wound on a pot core. The overall dimensions are 9/16" diameter by 5/16" height. The bobbin is 7/16" by 3/16"

There is a core gap which is about .005".

Estimated maximum effective V(peak) (since the output is not symmetric,  this isn't really precisely defined): 1000 V.

I have estimated the turns ratio for now but intend to perform some measurements to confirm. This is very rough at this point!

* Primary: 5 turns #26. The wire ends of this winding are buried so the wire size is based on required current capacity.

The primary appears to be wound first close to the core.

* Secondary: 250 turns, #43 (Yes, #43! The closest wire sample I have to     compare it to is #40 so this is what is known as a wild guess since I     have not disassembled the transformer to measure the wire precisely with     a micrometer) and is based on the resistivity of #43 wire and the bobbin dimensions.

I suspect that like a normal (TV or monitor) flyback transformer, the secondary is built up of several (single thickness) layers of windings (perhaps 40 or 50 turns each) with mylar insulating tape in between.

To somewhat confirm the the turns ratios, I measured the peak-peak input and output of the transformer while operating with a 1 mW HeNe tube: input was 15 V p-p; output was 700 V p-p.

Since this is a flyback converter and the transformer has a core gap, I  don't know how closely this correlates with winding ratios but at least it   is in the ballpark.

3. This is basically a wide compliance design and all stages of the voltage multiplier are active at all times. I do not know why the main HV filter capacitor is not at the output other than not being able to obtain a set of capacitors with an adequate voltage rating and somewhat limiting the surge current from the filter (due to lower voltage on the large filter  capacitors) when the tube starts.

4. WARNING: Despite its compact size, the output is high voltage and dangerous. Take appropriate precautions.

5.      |                         |           |    ---+--- are connected;    ---|--- and --- are NOT connected. |                        |           |

Sam's inverter driven HeNe power supplies: -

There are two variations on a similar approach which take advantage of the high compliance/poor regulation of these inverters for starting. Thus, no separate starting circuit is required.

These are both based on small flyback transformers and run on low voltage DC. For this, I use a very basic transformer/rectifier/filter capacitor power supply driven from a Variac.

No starting circuit is needed because of the wide compliance of thess circuits. With no load (tube not lit), the voltage will climb to 5 to 8 KV or more. As soon as the tube fires, the output drops to the sustaining ballast resistor voltage for the operating current. In essence, the poor voltage regulation of the inverter represents an advantage and allows this minimalist approach to be effective.

This is one type of design where monitoring of the input voltage to the tube is possible with a VOM or DMM requiring at most a simple high voltage probe. Parasitic voltage multipliers may not have enough current capability and pulse type starting circuits produce short high voltage pulses. It is possible to clearly see the voltage to the ballast resistor/tube ramp up until the tube starts and then settle back to its operating voltage. For small tubes, I can safely use my Simpson 260 VOM on its 5 KV range without a high voltage probe though it may go off scale momentarily.

The only additional components required for the HeNe laser power supplies may be one or two high voltage rectifiers and a high voltage filter capacitor. Since this is across the output at all times, it must be able to withstand the starting voltage but be large enough to minimize ripple when the tube is operating.

CAUTION: I would recommend using higher voltage capacitors than those shown unless you know that your inverter is not capable of reaching the capacitor's breakdown voltage. With some of these on a variable supply, 25 KV or more open-circuit is quite possible due to wiring problems, no tube connected, a bad or high starting voltage tube - or carelessness in turning the knob to far clockwise!

I have also tried a 500 pF, 20 KV doorknob capacitor on design #2 (I didn't have two such caps as required for design #1). While this low value works, it is a bit too small and results in about 20% ripple at an operating voltage of 1,900 V and current of 4 mA with a 15 kHz switching frequency. The minimal tube current setting for stable operation is slightly increased. At lower switching frequencies it will be worse and may prevent the tube from running stably at all. A few of these caps in parallel would be better. Or, use a stack of parallel plate capacitors made from aluminum foil and sheets of 1/8" Plexiglass :-).

WARNING: Since the voltage rating of these capacitors needs to be larger than for power supply designs with separate starting circuits, it is possible for a nasty charge to be retained especially if the tube should not start for some reason. Stored energy goes up as V*V!

Note: The difference in energy stored in the filter capacitor between the starting and operating voltages is dumped into the tube when it starts. For long tube life this should be minimized. Therefore, a smaller uF value is desirable for these high compliance designs. I do not know how much of an issue this really represents. A post-regulator can be used to remove the larger amount of ripple which results with samller capacitors. However, such a regulator must have overvoltage protection since at the instant the tube fires, it will momentarily see most of the starting voltage.

Sam's inverter driven HeNe power supply 1: -

This one is based on the inverter portion of the design described in the section: "Simple inverter type power supply for HeNe laser" but using the small B/W monitor flyback transformer option instead of a custom wound transformer. (For the doubler, the flyback must *not* have an internal rectifier.) The only differences are in the voltage ratings of the components required for the doubler and filter capacitors (to the right of points X and T in that power supply diagram).

Thus, it is an extremely simple circuit with no adjustments. Power output is controlled strictly by varying input voltage. Only a pair of high voltage rectifiers and a pair of high voltage filter capacitors for the doubler are required to complete the power supply.

It requires between 6 and 12 VDC (depending on HeNe tube power and ballast resistor) at less than 2 A and will power small HeNe tubes requiring up to about 6 mA at 2,000 V, perhaps more.

Other Version of the Modified (fried) High Side Regulator with a Wide Compliance Low Side Regulator Using PNP Transistors Instead of the more Conventional NPN type
This one is quite similar to the two Aerotech models PS1 and PS2 but is constructed entirely with parts that are readily available and inexpensive. Well, that is, except for the power transformer which you will still have to scrounge from somewhere. See the section: "AC line operated power supplies" for possible sources for these boat anchors. Also, due to low demand, the prices of high voltage electrolytic capacitors seem to be quite high (about $1.00 each for 10 uF at 450 V). I had a pair of surplus 1 uF, 1,500 V oil filled capacitors so I used them instead. A pair of microwave oven HV capacitors could also be used since these are typically around 1 uF at a minimum of 2,000 VAC (greater than 3,000 VDC). The cost of the remaining components (diodes, capacitors, and resistors) was less than $5.

1N4007s in series.
 * The high voltage rectifiers for the doubler are each constructed from five

capacitors with 10M bleeder resistors on each.
 * The main filter on the doubler is a pair of 1 uF, 1,500 V oil filled

1N4007s in series.
 * The high voltage rectifiers for the multiplier are each constructed from four

The high voltage capacitors for the multiplier are each constructed from four .001 uF, 1,000 V ceramic disk capacitors in series.

The series resistor for the parasitic multiplier is 10 M.

a Variac is used to adjust beam current.
 * There is currently no regulator - I may add that at a later time. For now,

I have left room for equalizing components on the diode and capacitor stacks but so far am running without them without any problems up to 2,500 VDC for the operating voltage.

It took me roughly 3 hours to construct the doubler and starting multiplier on an old blank digital (DIP) prototyping board.

I then tested it with a Variac and a current meter with several tubes from 1 mW to 5 mW:


 * 1 mW Spectra Physics (3.2 mA): 1,400 V with Rb=100K.

ballast resistor in laser head).
 * 1 mW Aerotech (4 mA): 1,900 V with Rb=100K, 1,700 with Rb=22K (additional

laser head.
 * 5 mW Aerotech (6 mA): 2,300 V with Rb=22K (additional ballast resistor in

The Variac was quite effective at adjusting tube current.

At 115 VAC the output of the power supply is about 2,500 VDC. This design appears to behave in all respects similarly to the commercial power supplies.

SECTION-II * Operating voltage: 1,500 to 2,500 V.
 * Operating current: 0 to 10 mA.
 * Starting voltage: 7,500 to 12,500 V.
 * Compliance range: NA - no regulator as yet.

X    R3     C3            C5              C7              C9       +/\/\||+---||--+---||--+---||--+ | 10M, 1 W   CR3  |  CR4     CR5  |  CR6     CR7  |  CR8     CR9  | HV+ |         +--|>|--+--|>|--+--|>|--+--|>|--+--|>|--+--|>|--+--|>|--+--o T1    |    CR1   |Y              |               |               |       | +---+|>|---++||+---||--+---||--+      |   +---|--++       CR1-CR2: (5 KV) (2)                    5 W / |         |    |       CR3-CR9: (4 KV) (3)                        \ |     C2 _|_   / R2    C1-C2: 1 uF, 1,500 V, oil filled           | |   1 uF ---   \ 10M   C3-C9: 250 pF, 4 KV (4)                    | | 1,500 V |    /                                  LT1             | |         |    |  IL2 LED      R4     Tube- +-+ Tube+ | HV- o--+--++|<|---+---/\/\---+---+--|-|         -|---+ Beam On |   1K    |  _|_ +-+ o - Test + o  -
 * (   |          |    |    C4            C6              C8              |
 * (   |      C1 _|_   / R1                                               |
 * (   |    1 uF ---   \ 10M   T1: 900 VRMS, 100 mA                       |
 * (   | 1,500 V  |    /         (primary not shown) (1,9)             R3 /
 * (   |          |    |                                              47K \

Notes for Sam's line powered HeNe laser power supply (Sam-L1): -

1. T1 is from (approximately 40 year) old tube type TV. By using the lowest line voltage tap and its 5 V and 6.3 V filament windings anti-phase in  series with the line input, its output has been increased from about 750 VRMS to 900 VRMS.

2. CR1 and CR2 each consist of five 1N4007s in series:

o--|>|--|>|--|>|--|>|--|>|--o

3. CR3 through CR9 each consist of four 1N4007s in series:

o--|>|--|>|--|>|--|>|--o

4. C3 through C9 each consist of four .001 uF, 1,000 V ceramic disc capacitors in series:

o--||--||--||--||--o

5. Construction is on a blank digital prototyping board which just has pads for 28 DIP locations (16 pins each). Perforated or other insulating board could have been used as well. Smooth rounded connections and a conformal insulating coating are essential to minimize corona in the high voltage and starting circuitry.

6. A Variac is used to adjust current - I will eventually add a low side regulator similar to the one described in the section: "High compliance  cascade regulator".

7. Output is about 2,500 VDC at 115 VRMS input and 3,000 VDC at 140 VAC input.

8. There is audible evidence of HV breakdown near maximum output before the tube starts. I suspect this is on the board itself since I have not coated it as yet with HV sealer. This is not surprising since the output can exceed 10,000 V.

9. WARNING: the power transformer is capable of much more than the 20 mA  required for even higher power HeNe laser tubes making it particularly dangerous - take extreme care not to touch (or even go near) the high voltage terminals of this or any other high voltage power supply.