- Joined
- Oct 30, 2019
- Messages
- 225
I'm working on a project for repurposing captured water around the greenhouse+garden, have collected most of the parts from local auctions over the last year:
- A couple of IBC totes for bulk storage
- Various rolls of 1" and 3/4 poly tubing
- Recently brought home a 60 gallon pressure tank and old piston pump for something like $20 together
- The greenhouse already has a used sump pump since it sits slightly below ground
The pump looks to be a Demings Marquette, possibly 1949 based on "-J49" at the end of the serial number, which aligns with finding it in trade/marketing material from the early 50's:
It's about time to start putting everything together as proof of concept, before sketching out how the final system will get laid out, where spigots will go, etc
The pump and its motor both gave me a bit of false hope then grief:
- The motor started once, then wouldn't. It had the original cardboard-encased start capacitor, no markings. I shotgunned a couple guesses and found that a 40uf cap works fine, though I may try a larger one to see if it starts better under some load
- The pump primed and built pressure... once. then it wouldn't.
I figured it was worth getting into the pump to see how the valves are set up, whether they're in good shape, springs are broken, debris needs to be cleaned out, etc.
Pulled the pressure cavity apart, found the valve assemblies - basically 2 sets of check valves in parallel, on opposite sides of the double-cupped piston in the bore below.
You can kinda see that the back-right bolt isn't the same as the others - someone has been in here before and replaced that bolt with a brass t-bolt with a very shallow slot in the head. I'm guessing the rubber disks used for the valves are not original, maybe the original were leather or leather-adjacent.
Pulled the valves out. The two intake valves on the were tight but came out fine. I wrestled the t-bolt out of the upper-right by grabbing the flats with an open-end wrench since the slot was too shallow to use a screwdriver. I'm guesisng the prior person to put hands on this broke it, because it looked like the hole had been re-tapped and had a crescent of the original bronze/brass fastener left behind.
Also because I promptly did the same thing for the other outlet valve. I guess I'll make two new shoulder bolts
The failure-to-pressurize before didn't seem to be due to any broken springs or anything, but rather due to the inside being rusty enough that it was flaking off and getting caught in-between the valves and the plates they seal against.
Also the rust flakes were getting into the main bore, and sticking to the u-cups, asking to accelerate wear.
Once I had everything off, it looked like this:
I scraped at it with some picks, then took it outside and pressure-washed it to blast all the loose debris and rust off/out of every hole in the pump casting. The drain plugs' passages were all completely fully of rust/sludge debris, so I picked those out and got them pressure-washed out clean.
I have some 1/2" silicon bronze on hand, so I used that. 1/4-20 threads halfway up the reduced diameter, then .330" for just over .5" length, and 1/2" head deep enough to take a good slot for a big screwdriver.
I ended up using a very sharp HSS tool instead of carbide to keep tool pressure down, but also ended up using tail support as well.
I power-threaded it with a die in a pipe threading stock - I have a secondhand Ridgid 00-R set with not only the usual NPT die heads but also heads holding UNC from like 1/4 to 1/2" or so. A bit sketchy but I'm lazy and it went well, putting some pressure on with the back of the tailstock to get started then just moving the carriage along with the handle. If I was doing more than 1-2 of these maybe I would have even set the carriage up to move in time to follow along.
I got distracted by my HSS tool getting dull and overshot the shoulder diameter on one - going to .300 instead of .330. Rather than scrap the part, I threaded it then made a sleeve
At the bridgeport I visually aligned a "looks right" slitting saw with the pip left by parting the heads off - being that this was a part for another project and not a machining-for-the-sake-of-machining project, eyeballing things seems reasonable.
Came out pretty nice, camera really didn't want to focus without getting dark so I had to tweak these up in brightness after the fact --
Wirewheeled all the old parts and used sandpaper+glass to flatten the surfaces of some parts, put the 4 valve assemblies back together. The rubber is in fine shape so I just washed it with soap+water to remove the rust build-up and inspect the sealing surfaces
Put it back together, with silicone grease/paste wiped on everything that needed to seal or slide on one another (all the valve parts "because why not", and the piston's cup seals, the packing material that goes on the crankshaft, etc)
I did a quick test on the bench, pouring water in the cavity on top of the valves and confirming it did not leak down -- looked surprisingly good!
It runs and builds pressure great! The video below sounds noisy, there's a higher-pitched "pfft" which is the seal plug at the end of the bore leaking and hitting a plastic bag I put over it temporarily. I ended up taking that back off and cleaning up the sealing surfaces with a carbide scraper and sandpaper+block of steel and it sealed fine with a bit of extra torque.
www.youtube.com
Other video disclaimers:
- I mention the capacitor value with regard to the running current - I mis-spoke, this only has a start capacitor and I wasn't thinking straight
- The pressure gauge shown is new old stock from my drawer-o-pressure-gauges and sadly seems wildly inaccurate, reading 20-30psi high
The pump is really quite quiet after I got that leak noise down, and started running it at 20-40psi instead of shooting for 30-50. I think this will be better for the longevity of the pump and motor. It doesn't like starting under pressure, but if it sits (like the time it takes you to use 20-30 gallons of water at a reasonable pace), the pressure behind the tank's check valve does bleed out through the pump and it does start fine. It's only at the highest pressure with the piston in certain positions that it doesn't reliably start - I'm thinking about trying some slightly-higher-and-lower capacitor values in case my first guess wasn't tuned quite right. Eventually I'll trust it to start unattended, though I might look for some kind of in-line ~5amp breaker I can add so that it trips if it sits at its locked rotor current (~11.5a) for more than a couple seconds. If anyone has a suggestion on a product that might be good for that, I'm all ears
Thanks for reading
- A couple of IBC totes for bulk storage
- Various rolls of 1" and 3/4 poly tubing
- Recently brought home a 60 gallon pressure tank and old piston pump for something like $20 together
- The greenhouse already has a used sump pump since it sits slightly below ground
The pump looks to be a Demings Marquette, possibly 1949 based on "-J49" at the end of the serial number, which aligns with finding it in trade/marketing material from the early 50's:
It's about time to start putting everything together as proof of concept, before sketching out how the final system will get laid out, where spigots will go, etc
The pump and its motor both gave me a bit of false hope then grief:
- The motor started once, then wouldn't. It had the original cardboard-encased start capacitor, no markings. I shotgunned a couple guesses and found that a 40uf cap works fine, though I may try a larger one to see if it starts better under some load
- The pump primed and built pressure... once. then it wouldn't.
I figured it was worth getting into the pump to see how the valves are set up, whether they're in good shape, springs are broken, debris needs to be cleaned out, etc.
Pulled the pressure cavity apart, found the valve assemblies - basically 2 sets of check valves in parallel, on opposite sides of the double-cupped piston in the bore below.
You can kinda see that the back-right bolt isn't the same as the others - someone has been in here before and replaced that bolt with a brass t-bolt with a very shallow slot in the head. I'm guessing the rubber disks used for the valves are not original, maybe the original were leather or leather-adjacent.
Pulled the valves out. The two intake valves on the were tight but came out fine. I wrestled the t-bolt out of the upper-right by grabbing the flats with an open-end wrench since the slot was too shallow to use a screwdriver. I'm guesisng the prior person to put hands on this broke it, because it looked like the hole had been re-tapped and had a crescent of the original bronze/brass fastener left behind.
Also because I promptly did the same thing for the other outlet valve. I guess I'll make two new shoulder bolts
The failure-to-pressurize before didn't seem to be due to any broken springs or anything, but rather due to the inside being rusty enough that it was flaking off and getting caught in-between the valves and the plates they seal against.
Also the rust flakes were getting into the main bore, and sticking to the u-cups, asking to accelerate wear.
Once I had everything off, it looked like this:
I scraped at it with some picks, then took it outside and pressure-washed it to blast all the loose debris and rust off/out of every hole in the pump casting. The drain plugs' passages were all completely fully of rust/sludge debris, so I picked those out and got them pressure-washed out clean.
I have some 1/2" silicon bronze on hand, so I used that. 1/4-20 threads halfway up the reduced diameter, then .330" for just over .5" length, and 1/2" head deep enough to take a good slot for a big screwdriver.
I ended up using a very sharp HSS tool instead of carbide to keep tool pressure down, but also ended up using tail support as well.
I power-threaded it with a die in a pipe threading stock - I have a secondhand Ridgid 00-R set with not only the usual NPT die heads but also heads holding UNC from like 1/4 to 1/2" or so. A bit sketchy but I'm lazy and it went well, putting some pressure on with the back of the tailstock to get started then just moving the carriage along with the handle. If I was doing more than 1-2 of these maybe I would have even set the carriage up to move in time to follow along.
I got distracted by my HSS tool getting dull and overshot the shoulder diameter on one - going to .300 instead of .330. Rather than scrap the part, I threaded it then made a sleeve
At the bridgeport I visually aligned a "looks right" slitting saw with the pip left by parting the heads off - being that this was a part for another project and not a machining-for-the-sake-of-machining project, eyeballing things seems reasonable.
Came out pretty nice, camera really didn't want to focus without getting dark so I had to tweak these up in brightness after the fact --
Wirewheeled all the old parts and used sandpaper+glass to flatten the surfaces of some parts, put the 4 valve assemblies back together. The rubber is in fine shape so I just washed it with soap+water to remove the rust build-up and inspect the sealing surfaces
Put it back together, with silicone grease/paste wiped on everything that needed to seal or slide on one another (all the valve parts "because why not", and the piston's cup seals, the packing material that goes on the crankshaft, etc)
I did a quick test on the bench, pouring water in the cavity on top of the valves and confirming it did not leak down -- looked surprisingly good!
It runs and builds pressure great! The video below sounds noisy, there's a higher-pitched "pfft" which is the seal plug at the end of the bore leaking and hitting a plastic bag I put over it temporarily. I ended up taking that back off and cleaning up the sealing surfaces with a carbide scraper and sandpaper+block of steel and it sealed fine with a bit of extra torque.
1949 shallow well water pump running after making new parts
Deming Marvalette shallow well piston pump first test run after making/repairing the brass and bronze valve assemblies. For the diagnosis, lathe work and oth...
www.youtube.com
Other video disclaimers:
- I mention the capacitor value with regard to the running current - I mis-spoke, this only has a start capacitor and I wasn't thinking straight
- The pressure gauge shown is new old stock from my drawer-o-pressure-gauges and sadly seems wildly inaccurate, reading 20-30psi high
The pump is really quite quiet after I got that leak noise down, and started running it at 20-40psi instead of shooting for 30-50. I think this will be better for the longevity of the pump and motor. It doesn't like starting under pressure, but if it sits (like the time it takes you to use 20-30 gallons of water at a reasonable pace), the pressure behind the tank's check valve does bleed out through the pump and it does start fine. It's only at the highest pressure with the piston in certain positions that it doesn't reliably start - I'm thinking about trying some slightly-higher-and-lower capacitor values in case my first guess wasn't tuned quite right. Eventually I'll trust it to start unattended, though I might look for some kind of in-line ~5amp breaker I can add so that it trips if it sits at its locked rotor current (~11.5a) for more than a couple seconds. If anyone has a suggestion on a product that might be good for that, I'm all ears
Thanks for reading
Last edited:
