spark eroder

[martin haysomParticipant @martinhaysom48469]

i built a spark eroder  to a design by Derck Lynas. very simple but does all i need to do. when i switch it off, the capacitors remain charged. i use a power supply as per Derck’s drawing. is there anything i could add  which would discharge them when switching off

[Julie AnnParticipant @julieann]

Add a resistor(s) across the capacitor(s). In order to calculate values we’d need to know the capacitance, whether the capacitors are in parallel and/or series, operating voltage and ideally the exact capacitor specification, or at least the leakage current spec if the capacitor has one.

Julie

[martin haysomParticipant @martinhaysom48469]

thanks Julie, i will have to take the power supply apart to get that information. hopefully i will have time tomorrow

[Clive FosterParticipant @clivefoster55965]

Derek unashamedly set out to build the very simplest functional spark eroder possible. Article says during testing he concluded that a discharge resistor was unnecessary.

Presumably assuming that on switch off the solenoid would shortly fall therefore earthing the capacitor and discharging it so dangerous voltages would not be present for long. Clearly this isn’t the case with your version. I believe Derek used a 470µF capacitor so 27 kΩ will discharge in around 1 minute which should be OK for safety and not seriously affect operation.

Which didn’t seem unreasonable to me when I looked at the design in the spirit of “if I need it I will build it”. Chances of me needing a spark eroder being so low that time spent on a sophisticated design would be wasted.

Mike at https://mikesworkshop.weebly.com/a-spark-erosion-apparatus.html  has some interesting comments on the design and what appear to be ways of improving performance without loosing the essential simplicity.

Clive

[BazyleParticipant @bazyle]

Mike’s webpage shows 21v which will barely electrocute a flee. If you are worried about the residual charge causing a spark that might set fire to the bottle of sloe gin you just brewed (well it is September) short it out with a croc clip.

[martin haysomParticipant @martinhaysom48469]

i built this because i originely needed to remove a 6ba tap i broke building a bonelle tool grinder. the power supply is very simple as i don’t know any other way. it runs at 60volts hence the reason i would like it to discharge itself. i original tried 24v painfully slow. the internet tells me 110volts is better, i won’t be going there.

[noel shelleyParticipant @noelshelley55608]

I find these devices very interesting and having read the 2 books on making such, I picked the basics of the 2 designs and made one ! The cost of the transformers put me off a bit BUT ! Living near a seaside town, in the the lead up Nov 5 the arcades would donate their old one arm bandits to the fire. This was before the days of LED lamps. I would turn up in the dark and rob the power supplies out of them, hefty transformers and bridge rectifiers plus other useful bits before the 5th. Time passes ! The transformers gave 5V, 12v, 33v and 50v, the last 3 at reasonable amps. 2 coupled gave me 50v at 8A and 66V at 6A and on a 3/8″ dia tool a cutting rate of 1mm in 2 mins in stainless. The 5v and 12v gave the control circuit supply, It was all fairly basic but worked well. If what I made is of interest I will pass on the details .  Noel.

[martin haysomParticipant @martinhaysom48469]

thanks Noel. i dont want to rebuild the power supply as it not used much.

Julie. information you asked for 60v dc  with no load. capacitor 2200 x 2 in parallel. they are cheapy ones from China

no other information on them

Bazyle.  shorting it out is what i do now, i want it to shut down when its switched off.

martin

 

[Clive FosterParticipant @clivefoster55965]

martin

Mostly I used to slap 10 kΩ on and call it good! Difference between 1 minute and 5 minutes was rarely of great enough import to be worth doing the maths.

With 4,400 µF capacitance 10 kΩ will do it in around 2 or 3 minutes depending on what voltage you choose to define full discharge. Generally somewhere between 3 and 5 time constants, equivalent to 5% and 1% of voltage remaining, is considered discharged for practical purposes.

Clive

[martin haysomParticipant @martinhaysom48469]

thanks Clive, i guessing the resistor needs to be a high wattage one

Martin

[Robert Atkinson 2Participant @robertatkinson2]

Not too high a wattage for 10K and 60V. 0.36W (Volts squared divided by resistance). So a 1/2 watt would do but will run quite hot. It’s dissipating full power all the time it the unit is powered. I’d use a 3W resistor. Will take a few minutes to discharge to the point of no spark but would be below dangerous shock level in a minute.

Robert.

[Julie AnnParticipant @julieann]

I was going to suggest a 10k resistor but have been beaten to it. The resistor doesn’t need to be anything fancy; 5% and 1W will be fine.

Julie

[BazyleParticipant @bazyle]

If you are worried enough to bother with a resistor then make it with two in parallel so that if one does fail you still have a safety margin. You can test them both each year when you PAT test your PSU.

[Paul Mills 4Participant @paulmills4]

Not really an electronics anything but could you use a relay operated by the main power supply to short it out when you turn of the power input, I know its probably a stupid idea…

[John HaineParticipant @johnhaine32865]

May spark erode the relay contact!

[martin haysomParticipant @martinhaysom48469]

thank you every body for your help. i will get a resistor of 10k ohms from that well known online auction site. one of those gold coloured ones with two fixing holes.

[FulmenParticipant @fulmen]

@John: You would need a resistor to dissipate the power, but you could use a smaller one to bleed it out faster without loosing power while running.

[martin haysom]

On martin haysom Said:

thank you every body for your help. i will get a resistor of 10k ohms from that well known online auction site. one of those gold coloured ones with two fixing holes.

For your application one of the power resistors in an extruded aluminium case typified by the Welwyn WH series will be fine if a bit overkill. But for uses at higher power they have a trap for the unwary.
While they range from 10 to 300 watts you need to read the datasheet:

https://www.farnell.com/datasheets/423255.pdf

carefully. The rated power is for the resistor mounted on a substantial heatsink.  For example the common 25W “WH-25” requires a 544mm2 1mm thick aluminium heat sink to run at 25W With no heatsink the rating falls to only 5W and it will be over 100 deg. C at that.
The smallest WH5 (rated 10W) will do 2W without a heatsink.

Given you have two capacitors I’d use one 2W wire ended resistor in the 10K to 22k range connected directly across each capacitor.

For example a 15K from this range:
https://www.ebay.co.uk/itm/255410514029?
£3.19 for 10 including post. They are a seller in the UK who I trust.

Robert.

[Robert Atkinson 2]

On Robert Atkinson 2 Said:

On martin haysom Said:

thank you every body for your help. i will get a resistor of 10k ohms from that well known online auction site. one of those gold coloured ones with two fixing holes.

..The rated power is for the resistor mounted on a substantial heatsink.  For example the common 25W “WH-25” requires a 544mm2 1mm thick aluminium heat sink to run at 25W With no heatsink the rating falls to only 5W and it will be over 100 deg. C at that…

Robert.

On a pure tangent (everyone else does it, so why shouldn’t I?), I found out the hard way about the lack of disspiation.
At my first job I was tasked with adding a PSU test unit to our commercial Automatic Test Equipment. I needed to test the set of 8 variable PSUs, rated at 25V, 5A. Most of the equipment was basic, ensuring that each PSU functioned in the PSU tray and all of the combinations of series and parallel connectivity functioned.
Last test was going to be a load test – for which I built up a massive set of low ohmage WH style 100W resistors and wired them with solid copper wire in parallel.
Testing took place around my desk partly on the floor which was typical office stiff nylon.

Fired everything up and eventually got to the full power point – and was then distracted…..The smell of very hot plastic woke me up and I saw the arrar steadily de-soldering itself to allow individual resistors to fall into the carpet where they buried themselves in their own molten hole…..

Two years later when I resigned, several were still there hidden underneath my desk…..

First job, wonderful place but manic!

 

[Clive FosterParticipant @clivefoster55965]

Having a relatively high current discharge resistor permanently in circuit is inefficient and significantly increases the power consumption. Connecting the capacitor to the discharge resistor via the normally closed contacts of a relay connected to the power on-off switch so as to open when power is applied removes the parasitic drain. It also significantly reduces thermal stress on the discharge resistor as it only carries current for a short period immediately after things turn off. No heatsink needed if one of the aluminium finned case variety is used.

I’m partial to the ISO “inch cube” automotive relays for relatively high current duties that don’t involve switching mains and similar AC voltages. Readily available and relatively inexpensive up to 20 amp or so current capabilities, more if you hunt around, and generally specified up to 60 volts DC. Nominal isolation is 500 volts so fine for moderate voltages. In my hands AC mains level voltages get a proper contactor specified for the job.

Clive

[Robert Atkinson 2Participant @robertatkinson2]

With 10k or 15 k giving 0.36 or 0.24W per resistor less than 1/2W is not a significant increase in this case.

Even a “mini” ISO automotive relay has a 1.4W rated coil so would use 3 times as much power keeping resistors out of circuit. And that assumes the spark eroder  has a suitable low voltage available for the coil….

Robert.

[Robert Atkinson 2Participant @robertatkinson2]

Simon,
I can’t remember the first time I used these resistors but as a avid reader of component data I’d already spotted it. Probably in the RS Components catalogue. I’s seen plenty of examples of them un-soldering themselves of if the overload is a bit higher blowing the end plug out leaving a tail of resistance wire shorting out to the chassis. Somewhat worryingly more than one of those was on an aircraft. One got me a move from the hangar to the design office. I’d told them it would overheat and got “shot down”. So when it un-soldered itself on test I hooked it up on a bit of locking wire  walked through two hangars, up some stars into the design office and “handed” it to the designer. He promptly swore and dropped it. During the ensuing discussion I mentioned that my previous jobs included design. So the head of design said “why are you working in the hangar? Well I still have my CAA engineers licences and you weren’t advertising for designers. I managed to continue doing some work in the hangars after the move.

Robert.

[simondavies3Participant @simondavies3]

Robert,

I have had my own fair share of “it’s not working and here are the broken parts to prove it” – both on the receiving and transmitting sides. I try to avoid causing physical pain though!

Simon

[Author]

Posts