Reply by Johnny B Good August 4, 20182018-08-04
On Tue, 17 Jul 2018 18:35:11 +0000, Johnny B Good wrote:

> On Tue, 17 Jul 2018 15:37:08 +0000, Johnny B Good wrote: > >> On Mon, 16 Jul 2018 20:29:23 +0000, Jasen Betts wrote: >> >>> On 2018-07-16, Johnny B Good <johnny-b-good@invalid.ntlworld.com> >>> wrote: >>> >>>> Well, it seems my 'guessed 1 ohm or so windings resistance figure' >>>> was >>>> way off the mark. When I finally got round to lifting the lid off the >>>> genset to take some resistance measurements, I was quite surprised to >>>> discover a phase to phase resistance value of 8.5&Omega;! Making the >>>> reasonable assumption that the alternator stator is wired in star >>>> (Wye) >>>> configuration, that's a phase coil resistance value of 4.25&Omega;. >>> >>> If you remove the spark plug and energise the windings with say 1A DC >>> you can get a feel for how much rope-force that's equivalent to and >>> compare that with how much rope force is needed for the compression >>> stroke. that should allow you to estimate how much run current it'll >>> take to strart the thing using the generator. >> >> Thanks Jasen. >> >> Funnily enough, that thought *had* crossed my mind at some point in my >> pondering this 'Great Unknown'. :-) >> >> Now that I've actually *measured* the phase to phase winding >> resistance >> (all three measuring 8.5&Omega; - 8.6&Omega; minus the 0,1&Omega; test lead resistance), >> I can try this test for the 1.5A value by simply connecting a 12v SLA >> across two of the phases. >> > ====snip=== >> >> I'll report back as soon as I've tried this test. >> >> > Well, the results were encouraging enough to make me think this project > might be even easier to accomplish than I had envisioned. I'm now > looking to concocting some sort of manually operated commutating switch > to get an even better feel for just how well the alternator might > perform as a BLDC starter motor (I've just got to take another look at > the basic principles to refresh myself of the details). > > Also, it has crossed my mind that I might be better off designing a > custom BLDC motor drive circuit that bypassess the need for a seperate > 12 to 48 to 60 volt dc-dc converter with a mains voltage rated BLDC > driver module.
Scratch that last thought. On reading it in the clear light of (the next) day, that final paragraph was just a load of bollocks. :-( I've come to the conclusion that I don't need any more than a 50vdc supply to drive the PMG as a BLDC starter motor (back emf versus a minimum cranking speed requirement plus enough voltage to drive some 5 to 6 amps through a phase to phase windings resistance of 8.5 ohms for starting torque). Given the above, I seem to have two choices, either use a 60v 10A rated electro-bike BLDC controller with some means of isolating it from the 400v once it's running as a generator or else use a high voltage bridge driver module with a separate controller that doesn't need a separate 3 pole isolating relay (just an 800PIV rated 8A diode to protect the 48v supply). Using a low voltage rated BLDC (or R/C BLDC ESC module) requires some form of 3 pole contactor relay. I'd initially considered a solid state switch for its fast response but this is likely to introduce an extra volt or two drop per phase connection (in effect, a bad 2 to 4 volt loss of drive voltage). A more conventional 3 or 4 pole contactor relay would eliminate such voltage loss but my worry is whether the release time would be fast enough to isolate the BLDC controller from the rapidly increasing voltage of the PMG once the engine springs into life. Remembering how swiftly that 54cc 4 stroke single cylinder engine accelerates from 400rpm or so cranking speed up to circa 4000rpm in little more than a second, that leaves very little time to detect and react to the rapidly rising PMG voltage before it exceeds the BLDC controller's drive transistors' maximum voltage ratings somewhere around the 500 to 600 rpm mark. I'm estimating that I'll have little more than 50ms to detect and isolate the BLDC before it gets fried by the resulting over-voltage. A fast release 3 or 4 pole contactor relay might disconnect in 10 to 15ms, maybe in as much as 20ms which leaves just 30ms to detect the initial rapid voltage increase and disconnect the drive current to the contactor relay's coil in response. Obviously, I'll need to add a lockout circuit to inhibit further startup sequences whilst the engine is running but that's a problem common to the other option of using a high voltage bridge driver module where isolation can be more readily achieved with a simple 800v 8A blocking diode. The low voltage module with isolating contactor has the charm of relative cheapness and the simplicity of a self learning module but it does have the addition of a fault critical component that could result in an expensive burn out. The other high voltage bridge driver is pretty well proof against this fate but does require a controller of some sort. Looking at the data sheets of these high voltage bridge driver modules, it's not clear whether it's possible to use a simple trapezoidal 6 step driving scheme or whether I'm obliged to use a PWM drive scheme. Quite frankly, a simple 6 step trapezoidal drive would be the ideal in this starter motor application - I just don't need the sophistication of PWM drive. I suppose I could fake a PWM signal (99% duty cycle) and use a modern cmos version of the Z80 cpu to drive the whole thing since, despite the lack of practice these past three decades, that's a processor I'm more than familiar with. However, I'm now back into the realm of "Re- inventing the feckin' wheel yet again!". :-( Another puzzling thing is there being a mention of a lower motor voltage supply limit way higher than my target 24 to 60 volt range mentioned in some data sheets along with undervolt lockout (logic or motor supply?). An exception to this appears in the specs shown here: <https://uk.rs-online.com/web/p/motor-driver-ics/1263478/> Where they quote a "Maximum Operating Supply Voltage of 400v" and a "Minimum Operating Supply Voltage of 0V" (presumably the motor supply voltage *magnitude* range which, if I've correctly interpreted it, means it would suit my intention to use it with motor supply voltages in the 24 to 60 volt range). What I can't pin down are example circuits demonstrating any flexibility of use between minimalistic simplicity and the OTT in complexity of sophisticated control circuit configurations typically implied by what example configurations do happen to appear in the data sheets. I suppose I'm possibly expecting too much by way of a sensorless controller setup under my specific and rather conflicting requirements. :-( Just as a footnote to this thread, I came across the following patent: US7180200B2 registered way back in 2002 by Black & Decker employees wherein they described a rather kakamaimee version of my idea where they seemed oblivious to the fact that you let the electronics overcome the impediment of high voltage windings instead of, as they suggested, the use of secondary low voltage high current windings to allow a 12v based BLDC motor controller to be used directly, forgetting that accommodating such extra 'copper' needlessly compromises the efficacy of the PMG's primary function as an alternator. More fool them for trying to pass off a piss poor variant of "The Bleedin' Obvious" as being something patentable, especially considering all the 'A Priori Art' involved (or whatever the exact Latin expression is for using well established existing technologies in a not particularly unique and novel form). -- Johnny B Good
Reply by Johnny B Good July 21, 20182018-07-21
On Thu, 12 Jul 2018 14:54:12 -0700, Lasse Langwadt Christensen wrote:

> torsdag den 12. juli 2018 kl. 23.10.05 UTC+2 skrev Johnny B Good: >> On Thu, 12 Jul 2018 07:48:07 +0000, Jasen Betts wrote: >> >> > On 2018-07-11, Johnny B Good <johnny-b-good@invalid.ntlworld.com> >> > wrote: >> >> I'm posting here in the hope that I can get some guidance on how to >> >> add >> >> an electric start feature to a cheap inverter generator cursed by >> >> the traditional recoil pull starter (the Devil's invention, imho). >> >> Apologies beforehand for what is a rather long initial post but I >> >> thought it best to appraise you of my thoughts and what I know of >> >> the problems involved in my proposed electronic add-on starter >> >> solution.
====snip====
>> >> Now that I have a rather neat solution to the problem of endowing this >> and, for that matter all such inverter gensets, with an electric >> starter option in mind, my question has been narrowed down from a >> general "How do I add an electric starter to this genset?" to the more >> specific "What is the best way to drive the 350v three phase PM >> alternator in my inverter genset as a BLDC starter motor?". >> >> It's a pity I hadn't framed my question that succinctly in the first >> place but it's only after receiving so much advice on every alternative >> but the one I'm after (all of which I'd already considered and >> discounted) that the need to restate my question becomes apparent. Not >> to put too fine a point on it, the question I'm really asking is this:- >> >> "What is the best way to drive the 350v three phase PM alternator in >> my >> inverter genset as a BLDC starter motor?" >> >> I can understand why such alternatives were being offered. A lot of >> similar pleas for help often involve needlessly complex electronic >> solutions to problems that can often be better addressed more >> pragmatically using a low tech option. Believe me, the complexity of a >> BLDC motor controller *is* the most pragmatic solution in this case. >> :-) >> >> > some kind of BLDC controller, sensored using three HALL elements rather > than sensorless is probably preferable with the low speed and uneven > load from the engine.
I know I said I'd prefer to avoid the complication of *retrofitting* the required Hall effect sensors but I'm beginning to think that this should considerably simplify the whole sensorless controller business down to a simple analogue of a brushed DC motor where the motor draws whatever stall current arises from slapping the starter voltage across its brush gear terminals dropping to whatever current is demanded by the cranking load once it reaches its limiting speed. IOW, can it be as simple as driving the motor by direct control from the Hall Effect sensor input to the high voltage driver module without using a fancy PWM scheme, just a simple application of the starter voltage supply to the windings in the classic 6 step sequence?[1]
> > 3x half-bridge drivers that can handle the starting current and the > voltage of the generator running, something like this might work > https://www.digikey.com/product-detail/en/sanken/SCM1243MF/SCM1243MF-
ND/4454409 That's the 15A rated driver chip. In view of the actual Ph2Ph resistance being 8.5 ohms as opposed to my initial one ohm guess, the 10A rated SCM1241MF would also suffice for the task.
> > when the motor starts, the diodes in the half-bridges will basically > form a 3-phase rectifier so a diode to isolate the battery.
Understood (and a damn good reason to have integrated this into the inverter module in the first place!). :-)
> > and some way to stop commutation when the voltage rises
That'll be trivial enough to arrange by detecting when that isolation diode becomes reversed biased. :-) [1] I've spent some considerable time studying what information I could extract from internet based sources since my initial response to your follow up reply. I have to say I've been getting sucked into the details of micro controller systems to the point where MEGO and I quit reading any more with the thought, "This is just details I *don't* really need to know about in quite so much depth, I just want a high voltage version of a self learning controller module I can slap into the genset, along with a 48v 7A 12v dc- dc converter module and an 8A 800v rectifier diode and a starter button and have done with the job. I don't want to re-invent this particular wheel all over again for feck's sake!". As best as I can understand it, a BLDC controller can be as simple as adding the three Hall effect sensors to a three phase synchronous motor and letting them directly control the commutation of DC pulses into the motor windings without any interference from a microcontroller to create in effect a motor that exhibits all the essential voltage/current characteristics of a simple PM DC brushed motor. Assuming I've interpreted this properly, such a simple motor setup would be just fine for the purposes of starting the engine. Torque ripple would be an asset rather than a liability in this case. Also, the unexpectedly high Ph to Ph windings resistance limits the maximum stall current anyway without any further need to invoke fancy PWM current limiting techniques. If that's the case, the use of Hall Effect sensors now looks a very attractive option indeed! Unfortunately, with the alternator likely to be either a 12 or 14 pole rotor design, trying to add some 12 or 14 magnets onto an accessible part of the crankshaft may be no easy task. There just might be sufficient space between the end of the crankshaft and the back of the inverter module right where the recoil starter mechanism lives to attach a sensor disk and detector pack (optical or Hall Effect) but at best, it won't be an easy modification even if it's at all possible. I'm not sure but I've seen the use of three Hall effect sensors embedded into the stator illustrated in a tutorial which implies that the need for 12 or 14 magnets can be neatly avoided in this case. However, retro fitting the Hall effect sensors into the stator assembly mightn't be quite so straightforward (or even practical). Although a sensorless commutating technique doesn't seem the best option here, it might be the only practical method even if starting from a standstill will be a less than elegant and problematical process as a consequence. Still, cranking an engine into life with electric start has never been a particularly elegant process anyway so I don't think this will be that much of a downside. A little bit of twitchiness persuading the motor to rotate is a small price to pay for the luxury of a working electric starter system. It doesn't have to function perfectly. Just functioning at all, "Warts and all", will suffice. :-) -- Johnny B Good
Reply by Johnny B Good July 20, 20182018-07-20
On Fri, 20 Jul 2018 08:02:27 +0100, Charlie+ wrote:

> On Thu, 19 Jul 2018 16:53:42 GMT, Johnny B Good > <johnny-b-good@invalid.ntlworld.com> wrote as underneath : > > snip >> It's just a shame that Parkside didn't in this case spring for the >> extra >>25 cents it would have added to the BOM costs to fit a priming button/ >>lever/squeeze bulb to manually operate said fuel lift pump to avoid >>tripling the wear and tear on an already "So cheap it'll fail before the >>first oil change" recoil rope starting mechanism. > > I found a solution against cranking forever (on one of the Tooltech > small non inverter gen kits, see Yamaha ET650 / ET950). > Drilled a 1.5mm hole in plastic panel cover directly in front of the > carb bell mouth, with a syringe I just squirt about 1.5 to 2 ml of gas > directly into the bell just before starting after a period uf unuse, > goes first pull every time doing this and saves arm and start mechanism! > C+
Seems a neat enough solution, a minor bit of 'faff' prepping it up with a worthwhile payback in reduced wear and tear on both yourself and the pull cord starter. If it's a two stroke prime mover, that'll be an exceptionally effective trick, assuming it's not gravity fed with a float bowl priming button which would be the usual way to achieve the same effect. I couldn't find any detailed enough pictures of the ET650/950 gensets to clarify your description of the 'carb bell mouth' but I presume you drilled through the plastic ducting from the air filter box right where it goes over the carburettor's intake (the ducting on these gensets typically makes absolutely no concession to the concept of 'gas flowing' to maximise volumetric efficiency). As useful a trick as that is, as far as starting on the pull cord, I'm hoping to render such starting measures (squirting a dose of fuel down the carb's throat or, better yet, adding a priming bulb to the fuel pump's engine vacuum port) a very low priority feature by relegating the pull cord from 'only available starting option' to 'emergency backup only, very last resort starting option' by pressing the alternator into BLDC starter motor service. When I look at the design of these pull cord start inverter gensets, I see an electric motor/generator mounted on the crankshaft where the motor/ generator is only half used. As an engineer, only making use of half of a motor/generator's capabilities is, quite frankly, an engineering obscenity. If a circuit component can have more than just one function, it's the designer's duty to ensure that it gets used for more than just one function whenever possible, especially if it reduces the component count. In this case, my correcting the original designer's oversight in letting the motor functionality of the motor/generator assembly go completely to waste is simply an exercise in good engineering practice to provide an electric start for no more than the cost of a cheap 4 or 7AH SLA battery and a sensorless BLDC motor controller module costing no more than 20 or 30 dollars and weighing in at a couple of ounces (a few extra cents and milligrammes if properly integrated within the inverter/controller module in the first place). So far, Lasse Langwadt Christensen has been the only contributor who has actually addressed my query with regard to the problems of adding a BLDC motor controller into this rich mix of 400vdc alternator come circa 50vdc driven BLDC starter motor system I'm trying to retrofit to this surprisingly good quality pull cord start inverter genset. Quite frankly, considering all the seemingly high quality electronic expertise in this news group, I'm rather surprised at the dearth of advice in regard of the electronics aspect of my little project and the almost boundless advice (good though it is) in regard of solutions for improving the efficacy of the existing pull cord system of starting that I'm trying to relegate to "Last chance saloon, emergency only, when all else fails, starting option". Thanks to Lasse, I now know that I require a 400v 10A rated BLDC motor controller using 600v rated drive transistors to cope with the 400v generating output from the BLDC/Alternator once the engine fires up. This shifts the problem of mixing a circa 50v BLDC motor drive voltage supply with the resulting 400vdc back emf to the simple matter of using an 8A 800v PIV anti-backfeed diode between the 50v supply and the generator output voltage. I don't need a sophisticated PWM drive, just a basic 'trapezoidal' six state switching 120 deg drive system (but one rated to cope with some 400 volts or more of back emf). Since the three phase alternator may well be optimised to its use, after rectification, as a DC generator with minimal ripple voltage, it may well be better to use a simple trapezoidal current drive setup anyway. In any case, 'torque ripple' won't be an issue - indeed it may well prove an asset in this usage case. Since I'm not entirely sure of the BLDC motor drive's starting voltage requirement in this specific case (all I'm fairly certain of at this stage, is that 24v is unlikely to suffice), I'm planning on using a 12 volt to 48 volt 7A rated DC-DC converter with galvanically isolated output so I can use the 12v battery to buck or boost the BLDC voltage between 36 and 60 volt to optimise starting performance for the minimum of stress on the starter battery. Now, all that remains of my project is the matter of sourcing the necessary components, sensorless BLDC and DC-DC converter modules (and an 8A 800v rectifier diode). Any advice in regard of such modules and sources will be greatly appreciated thank you very much. If you have anything further to add, please, can it *not* be anything to do with pull cord starting techniques this time? :-) -- Johnny B Good
Reply by Charlie+ July 20, 20182018-07-20
On Thu, 19 Jul 2018 16:53:42 GMT, Johnny B Good
<johnny-b-good@invalid.ntlworld.com> wrote as underneath :

snip
> It's just a shame that Parkside didn't in this case spring for the extra >25 cents it would have added to the BOM costs to fit a priming button/ >lever/squeeze bulb to manually operate said fuel lift pump to avoid >tripling the wear and tear on an already "So cheap it'll fail before the >first oil change" recoil rope starting mechanism.
I found a solution against cranking forever (on one of the Tooltech small non inverter gen kits, see Yamaha ET650 / ET950). Drilled a 1.5mm hole in plastic panel cover directly in front of the carb bell mouth, with a syringe I just squirt about 1.5 to 2 ml of gas directly into the bell just before starting after a period uf unuse, goes first pull every time doing this and saves arm and start mechanism! C+
Reply by Johnny B Good July 19, 20182018-07-19
On Wed, 18 Jul 2018 22:24:40 -0700, +++ATH0 wrote:

> On 2018-07-11 15:27, Johnny B Good wrote: >> From new, pretty well all of these engine vacuum driven fuel pumped >> carburettors needed several pulls on the starter cord to finally prime >> the float bowl so, since this was effectively an unboxing video (but >> with very little actual unboxing shots), 13 pulls doesn't seem >> excessive in the circumstances. > > You don't need a fuel pump on a small generator, you need an external > fuel tank with some head.
You do when you're making a compact suitcase inverter genset that you'd rather was no taller than it was long, if not for its aesthetics then at least in the interests of safety against it being accidentally knocked over onto its side. This design decision (not to go for that retro 80s look) often means the fuel tank can no longer be perched aloft of the engine's intake port just to create a gravity feed fuelling system. The one exception I'm aware of is Parkside's earlier PGI 1200 A1 with its classic pressed steel tank sat on top of a basic open frame fitted with cosmetic side panels.
> > Maybe if you bought some piece of cheap Chinese 2kVA shit on a > self-contained skid and threw it straight in your garage expecting to > plug it into an outlet when you need power later, it has a fuel tank > that's 2 inches above the carburator and you need to prime it.
The normal solution provided for gravity feed setups like you describe is the use of a carburettor float chamber that has a spring loaded priming button to let you flood the float bowl just prior to cranking it up. It's a neat and effective solution that also has the virtue of giving a positive indication that there is still fuel in the system and, more importantly, that it's reaching the carburettor float bowl.
> > Otherwise the float bowl ought to be full as soon as you turn the cutoff > valve on and it will start on the first pull.
Provided you allow a few second's wait for the float chamber to refill, that's generally true with most four stroke gensets, less so with two stroke units - they tend to play the startup game far less nicely. Anyway, all of the niceties of gravity feed fuel systems are rather beside the point in this thread. The inverter generator I'm trying to convert to electric start by adding a BLDC motor controller and starter battery to take advantage of the existing BLDC motor, currently passing itself off merely as a '400 volt three phase PM flywheel alternator' bolted onto the engine crankshaft, of necessity uses an engine vacuum driven fuel lift pump since the level of the fuel in the side panel tank is below the level required in the carburettor float bowl. I'm sure Parkside would have dearly loved to avoid the expense of an engine vacuum driven fuel lift pump, even as commoditised as such, made in their millions, components are cost wise - dirt cheap. The removal of the fuel tank from its 80s retro look location to a tank integrated into a side panel forced the use of just such a fuel lift pump. It's a solution but it does have its downside if the manufacturer short changes their customer out of some sort of fuel priming mechanism independently of having to hand crank the engine through half a dozen or more revolutions. It's just a shame that Parkside didn't in this case spring for the extra 25 cents it would have added to the BOM costs to fit a priming button/ lever/squeeze bulb to manually operate said fuel lift pump to avoid tripling the wear and tear on an already "So cheap it'll fail before the first oil change" recoil rope starting mechanism. Never mind. It won't matter so much when I'll be able to start it using battery power to drive its already built in BLDC starter motor. :-) -- Johnny B Good
Reply by +++ATH0 July 19, 20182018-07-19
On 2018-07-11 15:27, Johnny B Good wrote:
> From new, pretty well all of these engine vacuum driven fuel pumped > carburettors needed several pulls on the starter cord to finally prime > the float bowl so, since this was effectively an unboxing video (but with > very little actual unboxing shots), 13 pulls doesn't seem excessive in > the circumstances.
You don't need a fuel pump on a small generator, you need an external fuel tank with some head. Maybe if you bought some piece of cheap Chinese 2kVA shit on a self-contained skid and threw it straight in your garage expecting to plug it into an outlet when you need power later, it has a fuel tank that's 2 inches above the carburator and you need to prime it. Otherwise the float bowl ought to be full as soon as you turn the cutoff valve on and it will start on the first pull.
Reply by Johnny B Good July 19, 20182018-07-19
On Wed, 11 Jul 2018 17:57:00 -0700, Jeff Liebermann wrote:

> On Wed, 11 Jul 2018 22:27:07 GMT, Johnny B Good > <johnny-b-good@invalid.ntlworld.com> wrote: >
====snip==== Just reviewing the thread and spotted a couple of points I should have addressed sooner - I guess my t&ecirc;te-&agrave;-t&ecirc;te with tabbypurr must have distracted my attention. Sorry for that oversight and it's probably a bit late to address them now but I feel the need to do so anyway.
> >> I know some pull start inverter generators incorporate a 'Priming Bulb' >>to provide pulses of suck 'n' blow to operate the fuel pump >>independently of the engine but, like the piece of shit Workzone >>1800/2000W inverter genset sold by Aldi a week ago that I sampled three >>of before concluding they were utter shite, this one isn't so blessed. > > Ummm... I take it that you don't like this inverter generator, and by > implication also don't like the primer bulb concept? I don't want to go > into the pros and cons of primer bulbs. Let's just say they're a > necessary evil without which you would be draining the tank and carb > after every use and possibly removing crud from the carb.
I think I tried a little too hard for 'conciseness' when I was comparing the shite workzone's lack of a priming bulb to the same lack in the Parkside unit and bemoaning this act of utter stupidity (no doubt the result of short sighted "Beancounteritis"(tm) at Parkside/Lidl HQ). Despite my remarks on the benefit of a fuel priming bulb, you seem to have the idea that I don't like the primer bulb concept when in fact I view it as a necessary *good* which would not only have reduced the effort of starting it from 'dry', but would also have tripled the service life of a recoil pull cord starter mechanism seemingly designed to fail long before the very first oil change service becomes due. If as a manufacturer, you must curse your ICE powered product with only a recoil starter option, then it behoves you to at the very least, spend an extra few pennies fitting the necessary mitigation device(s) to prevent its use becoming a danger to your customers' health and wellbeing. In the case of a generator, a fuel priming bulb is the absolute minimum you can provide, even better would be an auto-decompression release as well as the fuel priming bulb. I suppose in the case of a chainsaw, an auto decompression release mechanism would take priority over a priming bulb.
> >>A fuel priming lever would have removed a lot of the effort required to >>start it after its been retrieved from 'dry' storage. I could probably >>add a priming squeeze bulb without compromising the warranty but I'll >>wait and see how my electric starter project goes before I look to >>adding that enhancement (it won't matter so much with electric start but >>it's one hell of a bonus when your only option is that accursed recoil >>starter mechanism). > > I think a fuel shutoff petcock and carb drain would be more useful.
That's why *all* ICE powered kit includes them as standard. Unfortunately, some genset manufacturers think they're doing their users a favour by the unforgivable sin of combing the fuel shut off lever with the engine kill switch. Workzone committed this very sin, compounded by saving a few cents on an eco-fast option switch, an especially egregious omission in view of the incredibly bad response it has to adjusting to just half of its rated load being switched on when it bogs down so severely, you're left wondering "How in Gawd's name did it not stall?" The Parkside genset is blessed with both a seperate engine kill switch *and* an eco-fast option switch, the value of which I never got the chance to appreciate until this Tuesday when I managed to stall it by switching on an 86% load before it had any chance to warm up. I restarted it easy enough and set it to fast idle before testing with my 900W toaster load to make sure it wasn't due to my having had the alternator disconnected to do some tests a little earlier. A few minutes later, I switched back to eco-idle mode and tried the 900W test again, proving it had returned to its impeccable response to sudden loadings where you can't detect any bogging down, just an immediate response to the throttle and stable power, so unlike that piece of shit Workzone sorry excuse for an inverter genset that I'd tried three examples of from Aldi a fortnight or so back.
> You would have the option of leaving gas in the tank as long as > necessary. If it's been sitting for months, drain the carb, or flush it > with carb cleaner. Most larger gas generators, motorcycles, and paint > sprayers have this feature.
I know all this, that's why I really appreciate the seperate from the fuel stopcock engine kill switch of the Parkside inverter genset, which allows me to run the carb float bowl and fuel line dry before I put it into extended storage.
> > You're not going to preserve the warranty after installing a starter > motor unless you're VERY careful. Your decision. Electric start or > warranty. Pick one.
I'm only connecting a plug-in 400vdc rated BLDC motor control module and a 48 or 60 volt DC-DC converter to run the alternator as a start motor. There'll be no gross mechanical modification involved whatsoever (I thought I'd made that totally clear in the first place).
> > How many pulls does it take to cold start your generator? If 13, then > maybe you have a point. If 5, don't bother, and just get some more > exercise.
The exercise I could do with but risking a premature failure of a cheap ropey pull start mechanism is something I'd prefer not to have to deal with, especially when I can convert to electric start without the use of excess baggage in the form of an add on starter motor whilst there's a perfectly capable BLDC brushless starter motor already fitted to the crankshaft in the guise of a 400vdc alternator come flywheel. A rather basic test recently with a 12v SLA to get some idea of the torque that can be generated by the alternator, suggests this pet project of mine is a lot more do-able than I'd hoped for. So much so that I think it's high time for me to track down a 400vdc 10A rated sensorless BLDC motor controller and a 5A 48v DC-DC converter to complete the project. :-) -- Johnny B Good
Reply by Johnny B Good July 17, 20182018-07-17
On Tue, 17 Jul 2018 15:37:08 +0000, Johnny B Good wrote:

> On Mon, 16 Jul 2018 20:29:23 +0000, Jasen Betts wrote: > >> On 2018-07-16, Johnny B Good <johnny-b-good@invalid.ntlworld.com> >> wrote: >> >>> Well, it seems my 'guessed 1 ohm or so windings resistance figure' >>> was >>> way off the mark. When I finally got round to lifting the lid off the >>> genset to take some resistance measurements, I was quite surprised to >>> discover a phase to phase resistance value of 8.5&Omega;! Making the >>> reasonable assumption that the alternator stator is wired in star >>> (Wye) >>> configuration, that's a phase coil resistance value of 4.25&Omega;. >> >> If you remove the spark plug and energise the windings with say 1A DC >> you can get a feel for how much rope-force that's equivalent to and >> compare that with how much rope force is needed for the compression >> stroke. that should allow you to estimate how much run current it'll >> take to strart the thing using the generator. > > Thanks Jasen. > > Funnily enough, that thought *had* crossed my mind at some point in my > pondering this 'Great Unknown'. :-) > > Now that I've actually *measured* the phase to phase winding resistance > (all three measuring 8.5&Omega; - 8.6&Omega; minus the 0,1&Omega; test lead resistance), I > can try this test for the 1.5A value by simply connecting a 12v SLA > across two of the phases. >
====snip===
> > I'll report back as soon as I've tried this test. >
Well, the results were encouraging enough to make me think this project might be even easier to accomplish than I had envisioned. I'm now looking to concocting some sort of manually operated commutating switch to get an even better feel for just how well the alternator might perform as a BLDC starter motor (I've just got to take another look at the basic principles to refresh myself of the details). Also, it has crossed my mind that I might be better off designing a custom BLDC motor drive circuit that bypassess the need for a seperate 12 to 48 to 60 volt dc-dc converter with a mains voltage rated BLDC driver module. -- Johnny B Good
Reply by Johnny B Good July 17, 20182018-07-17
On Mon, 16 Jul 2018 20:29:23 +0000, Jasen Betts wrote:

> On 2018-07-16, Johnny B Good <johnny-b-good@invalid.ntlworld.com> wrote: > >> Well, it seems my 'guessed 1 ohm or so windings resistance figure' was >> way off the mark. When I finally got round to lifting the lid off the >> genset to take some resistance measurements, I was quite surprised to >> discover a phase to phase resistance value of 8.5&Omega;! Making the >> reasonable assumption that the alternator stator is wired in star (Wye) >> configuration, that's a phase coil resistance value of 4.25&Omega;. > > If you remove the spark plug and energise the windings with say 1A DC > you can get a feel for how much rope-force that's equivalent to and > compare that with how much rope force is needed for the compression > stroke. that should allow you to estimate how much run current it'll > take to strart the thing using the generator.
Thanks Jasen. Funnily enough, that thought *had* crossed my mind at some point in my pondering this 'Great Unknown'. :-) Now that I've actually *measured* the phase to phase winding resistance (all three measuring 8.5&Omega; - 8.6&Omega; minus the 0,1&Omega; test lead resistance), I can try this test for the 1.5A value by simply connecting a 12v SLA across two of the phases. My thought (and I presume yours too) is that it'll lock the rotor into the nearest 'cogging point', allowing me to get a feel for or even measure how much pull is needed to drag it out of lock. Obviously, I'd want to rotate the crankshaft to the point of least resistance from the camshaft which happens to be when both valves are shut. This of course, is when the engine is on its compression/power stroke. Removing the spark plug in this case is vital to minimising the loading to just the minimum of frictional forces to get a sense (whether measured or by 'feel') of the net torque available from circa 1.5A of drive current. Taking a closer look at the pull cord mechanism, the rope unwinding diameter seems to start off at about 6 to 7 cm, reducing as the rope is (rapidly, one hopes!) unwound. This seems a remarkably high gearing ratio, requiring a lot of rope tension to achieve the required starting torque. With regard to the 'pull' required at the outer rotor/flywheel of the alternator to achieve the same torque figure, this looks to have about a 3:1 advantage from its circa 20cm rotor diameter. However, that of course means it'll need three times the driving voltage to attain the required cranking speed which, in view of the surprisingly high winding resistance [2], is less likely to be a mere 24v and more likely a 48 to 60vdc requirement. Of course, given a reasonable level of fitness and upper body strength (and cunning[1]), this high gearing ratio (which neatly gets even higher as the rope is unwound) has obviously been chosen as an optimised match to the dynamics of 'throwing a spear with the aid of a throwing stick' technique which extends the hand's reach to multiply the velocity being imparted to a spear that's just a fraction of the mass of the hand and forearm to improve kinetic energy transfer from the arm muscles into said spear (in this case, the cranking of a single cylinder 4 stroke engine with sufficient speed to start it up). The problem of course is that you need to be reasonably fit with, given the lack of a fuel priming bulb or automatic decompression release mechanism, a minimum level of stamina to boot. In my case, fitness and stamina are best described as being on the 'marginal side', hence my interest in taking advantage of what I'm pretty certain is a 'built in BLDC starter motor' by the trivial matter of connecting a small electronic module and battery with no mechanical modifications whatsoever. I'll report back as soon as I've tried this test. ======================================================================== NOTES: [1] You need to feel for the compression point and drag through it just far enough to avoid its impact (a complete stalling of your efforts and often, most horrible of all, a nasty 'Kickback') so as to give you the required run up to build sufficient kinetic energy in the flywheel (PM alternator rotor) to assist your efforts in rapidly completing the next suck, squeeze, bang part of the cycle - a successful 'bang' will take of the 'blow' part and all subsequent 'blow's. :-) [2] My gut feeling about the actual measured resistance of the stator windings being unexpectedly higher than my initial guestimate has been somewhat validated by the published values for the alternator used by the Powerhouse PH2700PRi 2.6KW cont. rated inverter genset where they quote phase to phase values of just 0.250 to 0.350&Omega;. Since this is to power a 120v inverter module of just under three times the power rating of the Parkside unit, you need to multiply by a factor of four to account for the doubling up of voltage required for a 240v version and then by 2.6 to account for the lower power output of a 1000W cont. versus the 2600W cont. of the Powerhouse example. This gives a suggested value of resistance around the 2.6 to 3.64 ohms mark, a median of 3.12&Omega; which makes the Parkside alternator winding resistance values some 2.7 times higher than expected by extrapolation from the Workhorse example. Even so, this would still have been about twice what I'd initially guestimated before I'd clapped eyes on the Powerhouse data. Comparing the values given for the DC charging winding (Blue-Blue in a seperate connector) in the Powerhorse manual of 0.045-0.070 &Omega; versus the 0.4&Omega; measurement of the Blue-Blue wires in the 6 station connector used for the phase winding connections in the Parkside unit, I'm beginning to wonder whether this dichotomy is the result of Powerhouse utilising Neodymium magnets where Parkside have settled for the old fashioned cobalt nickel cheapies. Incidentally, what takes the place of the blue battery charging connections in the Parkside connector, is shown in the Powerhouse diagram as being a "Sub winding", white wires with a resistance value of 0.100-0.160 &Omega;. Presumably, if such a sub winding exists in the Parkside unit (I don't have a diagram to check this out and it's hard to sort out from the mass of wires visible between the inverter module and the engine/ generator assembly just what else may be wired into the alternator stator plate), it's presumably wired to its own seperate 2 pin connector hidden somewhere in the wiring harness. One can only guess the purpose of this "Sub winding" in the case of the Powerhorse since the ignition coil seems to be a magneto coil powered from a magnet (flanked by similar sized lumps of soft iron) all riveted to the outside of the flywheel rotor (with a larger chunk of iron riveted onto the opposite side as a counterbalance). It clearly isn't anything to do with powering an electronic ignition module as might be the case with the Parkside unit, so whatever its purpose (it connects to the inverter module via a flylead connected 6 pin male connector in the Powerhorse in place of the blue DC generator connections in the parkside's embedded 6 pin male connector), it remains a mystery. In the case of the Parkside, there may well be a similar "Sub winding" which could well be providing power to the electronic ignition module for all I know. However, what I do know is that the Powerhorse has split the blue DC battery charging coil connections off to feed a totally seperate from the inverter DC generator module for battery charging, including its own starter battery. Whilst, otoh, this function appears to have been integrated into the inverter module used in the Parkside unit which appears to have not been wired up to a battery charging socket and hence an unused feature which might ultimately prove useful if I do succeed in my electric starter project as the means to keep the starter battery charged up - not necessarily a 12v LA (flooded cell, GM or Gel type). I can keep my starter battery options open on that one in view of the availability of commodity DC-DC converter and battery management modules. A long life Li-ion battery ime, might prove a less troublesome option. :-) -- Johnny B Good
Reply by Jasen Betts July 16, 20182018-07-16
On 2018-07-16, Johnny B Good <johnny-b-good@invalid.ntlworld.com> wrote:

> Well, it seems my 'guessed 1 ohm or so windings resistance figure' was > way off the mark. When I finally got round to lifting the lid off the > genset to take some resistance measurements, I was quite surprised to > discover a phase to phase resistance value of 8.5&Omega;! Making the reasonable > assumption that the alternator stator is wired in star (Wye) > configuration, that's a phase coil resistance value of 4.25&Omega;.
If you remove the spark plug and energise the windings with say 1A DC you can get a feel for how much rope-force that's equivalent to and compare that with how much rope force is needed for the compression stroke. that should allow you to estimate how much run current it'll take to strart the thing using the generator. -- &#1578;