|
Post by Roger on Dec 11, 2020 16:21:07 GMT
Is the element switched on and off to control the temperature or voltage controlled (via PWM or similar)? By the first option I mean like a conventional immersion heater. No, It's a Phase Angle control, so it's proportional control at the switching point. I'm just using a thermistor in the tube where the thermostat would normally sit.
|
|
stevep
Elder Statesman
![*](//storage.proboards.com/forum/images/stars/star_red.png) ![*](//storage.proboards.com/forum/images/stars/star_red.png)
Posts: 1,071
|
Post by stevep on Dec 11, 2020 17:15:24 GMT
At the risk of teaching my grandmother to suck eggs, did you put any appropriate grease in the tube to ensure thermal contact? (I'm sure you did).
|
|
JonL
Elder Statesman
![*](//storage.proboards.com/forum/images/stars/star_red.png) ![*](//storage.proboards.com/forum/images/stars/star_red.png)
WWSME (Wiltshire)
Posts: 2,924
|
Post by JonL on Dec 11, 2020 18:45:57 GMT
Is the element switched on and off to control the temperature or voltage controlled (via PWM or similar)? By the first option I mean like a conventional immersion heater. No, It's a Phase Angle control, so it's proportional control at the switching point. I'm just using a thermistor in the tube where the thermostat would normally sit. If I've got this right (I've no doubt you know them better than me) phase angle control can be a little finicky when it comes to hysterisis, depending on if the characteristic has been designed out they can sometimes be a bit laggy when it comes to response. They can also give some interesting issues with Power Factors, although I don't know if thats going to be an issue at this size of element. I just don't know if it will compromise your effective power available for responding to heat demand, meaning it might be a bit slow to kick back in. I know what I'm trying to say, I'm not very good at saying it.
|
|
|
Post by Roger on Dec 11, 2020 23:09:38 GMT
This is one of the lamp loads that I use for testing equipment. Here I'm using the heat of the lamp to feed back into the thermistor. (This particular load is usually fitted in place of a fuse on small equipment that keeps blowing fuses before you can figure out what's going on. Lamp filaments have a very low resistance when cold, but that drops dramatically when they get hot. It's the ultimate resettable fuse and current limit. It's a very old idea, but rather useful.) I managed to get it to control the temperature very nicely, but not at a high enough level. It's taken a while for the penny to drop that I'd chosen too low a resistor value in the thermistor voltage divider circuit in an attempt to make it more sensitive. That ended up drawing more current than the voltage reference in the chip could supply. That meant the voltage reference was dropping as the temperature neared the level I need. Anyway, that's been rectified now, so hopefully it will now behave on the boiler. 20201211_145130 by Roger Froud, on Flickr
|
|
|
Post by Roger on Dec 11, 2020 23:10:24 GMT
At the risk of teaching my grandmother to suck eggs, did you put any appropriate grease in the tube to ensure thermal contact? (I'm sure you did). Yes, I didn't mention it, but I've filled a syringe with Silicone Grease for that purpose.
|
|
|
Post by ettingtonliam on Dec 11, 2020 23:24:49 GMT
Just to nit pick, in the photo of the boiler under pressure, the gauge is nearly at the top end of the scale. I think its normal for maximum gauge reading to be 50% or even 100% more than working pressure.
Well done to have got the boiler to work, anyway.
|
|
|
Post by Roger on Dec 11, 2020 23:29:48 GMT
No, It's a Phase Angle control, so it's proportional control at the switching point. I'm just using a thermistor in the tube where the thermostat would normally sit. If I've got this right (I've no doubt you know them better than me) phase angle control can be a little finicky when it comes to hysterisis, depending on if the characteristic has been designed out they can sometimes be a bit laggy when it comes to response. They can also give some interesting issues with Power Factors, although I don't know if thats going to be an issue at this size of element. I just don't know if it will compromise your effective power available for responding to heat demand, meaning it might be a bit slow to kick back in. I know what I'm trying to say, I'm not very good at saying it. It's not a very subtle control setup, but the element is close to the thermistor in the tube, so the response shouldn't be too long. It's not one of these clever PID controls where you're in control of all three terms. This is a simple proportional control close to the setpoint, so overshooting is quite likely in a step response. Once it's settled down though, I think it will be ok, but I won't know until I've tried it tomorrow. I've chosen a 2K2 resistor in the thermistor voltage divider. The thermistor is about 10K at room temperature but goes down to about 550R at 166C. That makes it more sensitive than the original 10K resistor I used. I'd changed it to 1K but it drew too much current from the -5V regulated supply so I can't go that far. We'll see.
|
|
|
Post by Roger on Dec 11, 2020 23:31:47 GMT
Just to nit pick, in the photo of the boiler under pressure, the gauge is nearly at the top end of the scale. I think its normal for maximum gauge reading to be 50% or even 100% more than working pressure. Well done to have got the boiler to work, anyway. I'm sure that's true for a locomotive, but this is just a test boiler. It's only going up to 90PSI, so there's a 10% overhead.
|
|
|
Post by Roger on Dec 12, 2020 12:21:13 GMT
This is what I hadn't shown before, but it important to get good feedback from the Thermistor. The syringe is filled with Silicone grease and I'm just using a piece of heat shrink tubing to inject a decent amount of this deep into the tube. The Thermistor has a single layer of high temperature heat shrink sleeving on it because it does need to be electrically isolated from the tube. You can also see the high temperature boots on the crimps which haven't melted. The 3D printed piece is a bit soft, so you couldn't clamp it up tight. It's just an insulating cover so it's probably going to be good enough... just! 20201212_111145 by Roger Froud, on Flickr I've also added a Neon indicator because it's impossible to see what's going on with the heater without some kind of indicator. The feedback is pretty slow, so it does tend to hunt around the set pressure by around 5PSI when the temperature is set, and of course it will vary a bit when I start to use it in anger. It has settled down a bit now it's all achieved a steady temperature though. There's a lot of mass there to stabilise. 20201212_112159 by Roger Froud, on Flickr Still, I've left it running at 90PSI for about 20 minutes and it hasn't blown off. I can see it has finally arrived at a point where the Neon is glowing all the time and the pressure is pretty stable. Anyway, I'm calling that done for the moment, I'll crack on with the rest of the setup and get it to the point where I can actually do some testing. 20201212_115555 by Roger Froud, on Flickr I've just ordered a cheap 10 turn potentiometer with a counter from China which I'll use in place of the single turn pot. That will allow me to set the temperature more accurately.
|
|
|
Post by keith1500 on Dec 12, 2020 19:50:53 GMT
I think that temperature aligns with the steam tables for the pressure.
Nice bit of work Roger.
|
|
|
Post by Roger on Dec 12, 2020 22:40:06 GMT
I need to connect the Hand Pump up to the boiler, so this is a little adaptor that I can Silver Solder to a Copper extension pipe. The idea is to keep the heat away from this connection because it's going to have an M5 thread and a push fitting for a 4mm Pneumatic pipe. It's getting a hex added so I can put a spanner on this and the connector. 20201212_172744 by Roger Froud, on Flickr 20201212_175353 by Roger Froud, on Flickr 20201212_180227 by Roger Froud, on Flickr This is the only connector I've got in my box of parts, but it will suffice until the ones I've just ordered arrive. 20201212_223429 by Roger Froud, on Flickr This is the Steam feed for the injector which has a very thin flange with a small recess to locate the Copper pipe. I've placed a thick ring of Silver Solder on the tube so that it forms a substantial fillet. I've used plenty of Tippex everywhere except for the fillet area, I don't want it all over the bolt holes. 20201212_222824 by Roger Froud, on Flickr That looks good from this side... 20201212_224349 by Roger Froud, on Flickr ... but I clearly should have put Tippex inside the pipe too! 20201212_224236 by Roger Froud, on Flickr There was a surprising amount of Silver Solder in there, which took a fair bit of removing. I'll watch out for that on the other ones! 20201212_225303 by Roger Froud, on Flickr
|
|
|
Post by 92220 on Dec 13, 2020 9:27:40 GMT
Hi Roger.
At least you know the silversolder has filled the joint. A very interesting and handy project, this boiler!!
Bob.
|
|
|
Post by Roger on Dec 13, 2020 23:06:46 GMT
I've taken on board Adam's suggestion that a pressure gauge would be useful to check the delivery pressure, and I also need to arrange for the output of the injector to either go to the balanced clack valve or into the boiler. I could see it being a pain to have to keep using the hand pump. So I've spent the day designing and making some of the parts for this arrangement. The water from the injector comes in on the LH union fitting, and the Tee piece has an 1/8" BSP thread for a pressure gauge to sit on top. I'm not bothering with a syphon on this since most of the time it's going to be seeing water. I've borrowed from my Bypass valve design for the 3 port valve which has a 'T' arrangement for the porting and the valve element. Three way valve by Roger Froud, on Flickr The valve element can be in four positions. 1) Feed left to right for feeding the Balanced Clack valve 2) Feed left to bottom for feeding the boiler. 3) Off, so the output from the injector is blocked 4) Feeding to both the boiler and Balanced Check valve, although I can't imagine that being very useful Sectioned three way valve by Roger Froud, on Flickr This is the valve body in Brass getting the locations for the unions and the port holes... 20201213_163317 by Roger Froud, on Flickr ... before standing up to add the M2 fixing holes. You can see how tall the 4th axis is and why I wanted the slim 4-jaw chuck on there. Without the neck extension, it wouldn't be feasible. 20201213_164637 by Roger Froud, on Flickr 20201213_164654 by Roger Froud, on Flickr The Banggood boring bar was ideal for creating the 10mm bore. 20201213_173237 by Roger Froud, on Flickr This is the valve cover which has an O-ring seal on the OD and on the shaft ID 20201213_195952 by Roger Froud, on Flickr 20201213_200700 by Roger Froud, on Flickr The valve element is made from Fluorosint. Here I'm checking that the fit is about right. It needs to be water tight, but not too tight to turn. 20201213_204134 by Roger Froud, on Flickr The 'T' port holes were then drilled. 20201213_213152 by Roger Froud, on Flickr The valve element has a hex pocket for the drive. 20201213_214854 by Roger Froud, on Flickr
|
|
|
Post by Roger on Dec 14, 2020 23:01:44 GMT
This is one of the two outlet unions which has just had the M8 x 0.75 (fine) thread and conical seating machined. I'm just checking it fits a union nut and nipple prior to adding the hex on the mill. 20201214_124353 by Roger Froud, on Flickr ... like this. (Wobbly video) 20201214_125231 by Roger Froud, on Flickr This is the extension piece for the Pressure gauge receiving the location for the threaded boss. 20201214_111335 by Roger Froud, on Flickr That was parted off and the register turned to fit the valve body. 20201214_112233 by Roger Froud, on Flickr The 1/8" BSP threaded boss was then turned and here it's getting a 'U' shaped cutout so it doesn't block the through hole. 20201214_113917 by Roger Froud, on Flickr 20201214_114657 by Roger Froud, on Flickr So all of those pieces were assembled, with Tippex everywhere I don't want Silver Solder and flux where I do. 20201214_153558 by Roger Froud, on Flickr I did the usual trick of a ring of thin Silver Solder around each joint, and a shorter thicker piece on the body joint. 20201214_153809 by Roger Froud, on Flickr This is how it looked after 15 minutes in the ultrasonic tank... 20201214_161455 by Roger Froud, on Flickr ... then a quick clean up with a wire brush and emery in difficult to reach places. 20201214_162203 by Roger Froud, on Flickr I'd used the last of the batch of Union nuts and nipples, so I made three more. I'd written down the procedure and had the programs already so this was a really quick and simple job. I've shown it before, but that was years ago. Here I'm using a 5mm cutter to create the hex, the tapping size and the through hole. 20201214_171246 by Roger Froud, on Flickr The hole was tapped with a collar stop to make sure I didn't hit the bottom and damage the seat. 20201214_172710 by Roger Froud, on Flickr Then I used the 1.75 pitch thread insert to create the reduced diameter and the chamfer... 20201214_172817 by Roger Froud, on Flickr ... and the process repeated on the other end after parting off. The end gets a small chamfer inside and out. 20201214_173552 by Roger Froud, on Flickr The nipples are done in a similar way, with all the usual turning operations done on the mill to save time. It's all done with a 3mm cutter, and was very quick since I already had the programs. 20201214_221544 by Roger Froud, on Flickr This is the parallel blank parted off... 20201214_220944 by Roger Froud, on Flickr ... and then clocked up in the split collet which I already had... 20201214_221220 by Roger Froud, on Flickr ... ready for turning the nose. Very quick and simple once you've done it once. 20201214_221355 by Roger Froud, on Flickr So it's been a pretty productive day, even though it doesn't seem like I've spent much time in the workshop. 20201214_222208 by Roger Froud, on Flickr I just need to make the valve shaft and the handle, then make the pipes. Hopefully I can then look towards putting the Injector together for a test.
|
|
jma1009
Elder Statesman
![*](//storage.proboards.com/forum/images/stars/star_red.png) ![*](//storage.proboards.com/forum/images/stars/star_red.png)
Posts: 5,903
|
Post by jma1009 on Dec 15, 2020 0:00:26 GMT
Hi Roger,
You make reference today to Adam suggesting measuring the pressure of the injector delivery, but I believe I mentioned this in my post on here of 28th October?
Cheers, Julian
|
|
|
Post by Roger on Dec 15, 2020 7:35:35 GMT
Hi Roger, You make reference today to Adam suggesting measuring the pressure of the injector delivery, but I believe I mentioned this in my post on here of 28th October? Cheers, Julian Hi Julian, Sorry about that, I'd forgotten that you'd mentioned it before. I'll blame old age creeping on!
|
|
jem
Elder Statesman
![*](//storage.proboards.com/forum/images/stars/star_red.png) ![*](//storage.proboards.com/forum/images/stars/star_red.png)
Posts: 1,069
|
Post by jem on Dec 15, 2020 18:01:05 GMT
Roger , why bsp when you say that you always use metric for preference? M10 10 is very very close to 1/8 bsp. I am always very interested in your explanations of how to do things, so a big thank you
best wishes
Jem
|
|
|
Post by Roger on Dec 15, 2020 19:21:00 GMT
Roger , why bsp when you say that you always use metric for preference? M10 10 is very very close to 1/8 bsp. I am always very interested in your explanations of how to do things, so a big thank you best wishes Jem Hi Jem, I want to screw the gauge straight into the fitting, and all the gauges I've got have 1/8" BSP threads. That seems to be the standard for pneumatic gauges. This sort of gauge I'd plenty good enough for this application.
|
|
|
Post by David on Dec 15, 2020 22:08:31 GMT
It would never occur to me to use a threading insert for a chamfer tool so there's something learned.
How do you clock something up in a 3 jaw chuck? Can you bump yours around on the backplate? I saw ThreadExpress explain that on YouTube.
|
|
|
Post by Roger on Dec 15, 2020 22:55:41 GMT
It would never occur to me to use a threading insert for a chamfer tool so there's something learned. How do you clock something up in a 3 jaw chuck? Can you bump yours around on the backplate? I saw ThreadExpress explain that on YouTube. Hi David, I've got a GripTru chuck which has three tapered screw adjusters that give about 0.15mm travel in each direction. It's a huge time saver because you're always close to it being true before you start clocking. It usually takes about a minute to get something within a couple of microns, so I tend to clock up everything. I find it saves more time than it takes because you know that everything is as close to true as you can get.
|
|