Member postings for Clive Steer

Having designed gas boiler ignition controllers the flame failure detector uses the ionisation effect of a flame which acts like a diode. So an AC voltage is applied to the flame probe and the resultant DC voltage operates a transistor to allow the main burner gas valve to open. The controllers have to be designed so that a single circuit component failure will either shut down or prevent start up of a boiler.

Most modern domestic boilers are fan assisted so the controller first runs the fan for a time in order to clear any combustion products or gas from a previous start attempt and then tests for a change of state of the microswitch on a diaphragm air pressure sensor. If this proves OK the controller opens the pilot flame valve and begins spark ignition whilst detecting for pilot flame presence. If this is proved the controller opens the main burner gas valve and the main burner is ignited by the pilot flame. If pilot flame ignition isn't proved within something like 5 seconds the controller may immediately go to lockout and display a warning, usually a red LED, or repeat the sequence, usually no more than two times before going to lockout.

Older boiler designs used a continuous pilot flame which used a thermocouple to detect flame and hold the pilot gas valve open. Since this type of boiler contained no electronics the thermocouple although only producing a few millivolts could generate a reasonable amount of current. The coil of the pilot valve was just a few turns of thick wire and the valve would be opened manually to close the magnetic circuit. So the current from the thermocouple was insufficient to pull the valve in but sufficient to hold the valve open so providing fail safe control. If the thermocouple went open circuit, which was the most common failure mode, the pilot gas valve would close. An interlock within the valve block would prevent the main burner valve being opened if the pilot wasn't operated.

CS

For a lasting repair soft solder is the best as the thickness of the solder film can be quite thin and maintaining pallet dimension for correct escapement action is essential. Having said that there are excellent epoxy glues, such as those used to attach ultrasonic transducers to the tanks, but these need a glue film thickness to work correctly.

Soft soldering at about 250 C shouldn't reduce the hardness of the spring material being applied and even if it did the clock would probably work for another 100 years before the needing the "veneer" replacing.

CS

Chris

I also think the fault is in the inverter and most likely in the power transistor block that generates the 3 phase output from the DC link. The inverter would have initially caught the over current condition as a transistor went short circuit and later the blown transistor would produce an open circuit fault.

CS

If a VFD hasn't been used for some time as has been mentioned the electrolytic capacitors may need reforming. However this can be done at a fairly low voltage, say 12v, rather than the full 440v.

I have several inverters (8) , that have 3 phase inputs and are all working on single phase either at 240v or via a autotransformer at 440v for motors that can't be converted to delta. They are now all over 20 years old and show not signs of any issues.

The down side of working single phase is that, as mentioned, the DC link voltage will droop on load as the ripple will be higher than when working on 3 phase so at probable less than 50% loading the inverter will run out of steam and possibly flag an under voltage warning neither of which I've found to be damaging.

CS

I think the bonding wire that connects the LED terminal to the top side of the diode die may be the feature that limits the max pulse current that can be achieved.

CS

I'm not sure I understand how the nut locking feature caused the failure. Most of the self locking nuts I've come across are a conventional nut with a feature on the top that mimics a lock nut such as a nylon, fibre or crushed threaded tube friction feature. These nut would be wound down to the required torque and checked visually to ensure the thread of the stud or bolt had engaged fully with the locking feature.

I wouldn't have thought that the locking feature would need to extend down into the normal load transfer region of the nut and stud. However if the stud was too short to provide a locking feature on top of the nut or by doing so the nut would become thinner reducing the load transfer area.

CS

Nick

The thread doesn't necessarily take a working load if a fastener such as a bolt is taking its design load in shear ie orthogonal to the bolt axis in which case the nut is just tightened enough to stop the nut coming undone and if there are cyclic or torsional loads acting on the bolt a split pin is added

In the case of the cylinder holding down studs there is a load on the thread due to tightening plus cyclic tension loads during compression and ignition cycles plus complex asymmetric loads due to side thrust of the piston against the cylinder wall.

Most bolts in airframe applications including holding the wings on are working in shear mode. Having said that I had an aerobatic aircraft where the wing was held on by two 1/2 inch bolts in tension when the aircraft was pulling positive G which could be up to +6G !!! The weak point is the transfer of load between the bolt and nut and for this application there were two nuts per bolt.

CS

Interesting reading and worrying where threaded fasteners are being used in tension.

One would have thought that for this application the stud would have had the highest yield strength as replacing a stripped nut is far easier than replacing a stripped stud.

However the thinking here may have been that it is better to strip the thread on the stud than pull the stud out of the Ali crankcase if over tensioned. Unfortunately the designers thinking has probably been lost and the only info is the specification of the stud and nut.

CS

Thanks Robert for the update about the rise and fall rates for light emitted from a white LED.

CS

White LEDs are usually constructed using a blue/UV led covered with a white fluorescent material and the shade of white depends on the materials recipe. I suspect that the material may have a degree of persistence that would moderate 10Mhz operating.

I used a similar circuit to make an LED strobe light to visually check the operation and timing of watch escapements but these operate at relatively low frequency.

CS

There is a reaction at the pendulum pivot point which is used in dual pendulum clocks to provide coupling between the pendulums that bring them into mutual resonance. It would be interesting to know if a dual pendulum clock has more stable amplitude and reacts less to noise and vibration.

CS

JA

I agree with your frustration of using the shorthand notation of say Newtons and Pascal. It is bad enough being asked to set my altimeter to 1013 hectopascal rather than the earlier mbar so heaven forbid if they used the fundamental units.

CS

The maths that links the various measurement system is still the same but working out the units becomes the main problem. Pressure is force per unit area whether it is Newtons per square metre or tons per acre! However the metric system is still complicated by using a short hand that separates the user from the units such as using the term Pascal to mean Newtons per square metre.

The main problem I have is that the Imperial system, which probably has its origins in France due to 1066 and all that, has a variable number base. So a ton is 20 hundred weight and a hundred weight is 112 pounds which is 16 ounces. However a tonne is simply a 1000kg or 1,000,000 grams. So one can simply move the decimal point depending on what size of engineering one is trying to compute.

I believe Napoleon instigated the adoption of the metric system because of lack of standardisation of weights and measures in France and the opportunity this gave traders with less scruples to confuse or even swindle their customers.

This may still be the case in the UK where my gas bill states the units I've used, converts this to BTU's and then kW/hrs. I'm OK with this but there will be some who will be confused.

CS

No. A scruple is equal to 20 grains a measurement of weight/mass used by Apothecaries.

CS

The American gallon is smaller than the Imperial so does that mean their pints are as well.

I've got some scruples if anyone still uses this unit of mass.

CS

I think modelling will remain largely Imperial for some yet not just because the drawings are Imperial but because of legacy tooling that often gets passed on to newbies as well as part completed models. Since modelling is rarely profit making using what is easily and cheaply available is prudent. The more likely scenario will be the gradual introduction of metric parts such as fasteners and gears etc. One may build a model to Imperial measurements but substitute commercial of the shelf metric screws rather spend time finding a source of ME screws.

In fact someone starting out in the hobby may find the complexity of different measurement systems and component availability a serious negative and find something easier to do.

CS

I believe that some "metric" ball races use Imperial sized balls and don't mention the issue of screw threads!

Most weighing devices are force balances rather than mass balances as they are relatively easy to make and more compact than a true balance. Unfortunately the wrong terms are often used by lay people so the "weight" used on the old style grocers "Scales" which is actually a balance is actually a reference mass. The later Grocers "scales" were a force balance where a spring extension measured on a scale indicated the weight being measured. Most precision laboratory scales are force balances where a current flowing in a coil of wire within a magnetic field produces a balancing force to the item being "weighed". These devices use a reference mass to periodically calibrate the "balance" to account for the local gravity and variations due to the Moon which is noticeable if you are measuring to 1 part in 10 million.

Other weighing devices such as kitchen and bathroom scales use a spring in the form of a strain gauge.

CS

I'm rather late to this subject but it is not only model engineering that is Imperial as electronic components are often on Imperial 0.1 inch pitch. I'm happy to work in either measurement system but the whole concept of fractions makes arithmetic excruciatingly difficult and rather like doing sums using Roman numerals and a recipe for error.

CS

For a small compact vertical milling machine I purchased a Boxford VM30. It is simple and well constructed with mechanical variable speed head. The only downside is that the hi/lo spindle speed range selector uses an idler gear arrangement that is always engaged for either range so the head is very noisy. My workshop is in the garage which is inside the house so this was a problem. So I grafted a Bridgeport M head on to it and an VFD to reduce the need for belt swapping. The M head has a quill which the VM30 doesn't have which makes drilling more difficult. The more common J head might work but being larger looks out of proportion.

All of this came in for less than £1k.

CS