Sending PWM through a relay

[clearchris]

Re: Sending PWM through a relay

You have very reasonable prices. My guy, the one I like, is $100 to show up, $75 an hour after the first and AFAIK, 100% markup on parts. He's on the cheap side too. The funny thing is, he stocks virtually nothing except refrigerant. No fan motors, no caps, nothing. So everything is a trip to the supply house. But he actually knows what he's doing, though I will say, he's old school as hell. He looked at me like I had two heads when I pulled out an appion core remover.

$200 for a cap change is pretty normal in this area. My guy would come in probably around $125-150 on that.

The big companies around here with the fancy vans pull out a binder with glossy pages with prices that will make your eyes pop. $250 for a cap, another $250 to install a kickstart, $1600 for a condenser fan, etc, no warranty on that kickstart mind you. That's when I learned to do it myself.

[clearchris]

Re: Sending PWM through a relay

I wrote some python scripts to manage the gcode for holding temp, ramping up and down and made a go at a test trapezoidal reflow (at lowered temps). Pretty good considering I left the door open for the first ramp up, which also probably explains the overshoot at the top. The ramp down diversion was expected, that's a maximum decrease slope, i.e. if the oven cools any quicker than that, it needs to add heat to slow the cooling. Also at higher temps, it would ramp down quicker.

It's also a very smooth ramp considering it changes the temp every five seconds, so the target isn't a straight line up as octoprint shows it, but a stair step shape. But that's pretty normal for CNC machines/3d printers, any diagonal line is, at a micro level, a lot of very small steps.

[sam_sam_sam]

Re: Sending PWM through a relay

That's when I learned to do it myself.

I can understand that and I done this myself because a lot of these trades people and trades company’s like air conditioning plumbing and other trade like this it seems all they want to do is take as money as they can do and if you are very lucky you might get some one that does a good job and least you get some value for the money that you are spending

[Curious.George]

Re: Sending PWM through a relay

The price that is charged can vary greatly depending who you call I know a case where someone got bill for changing a capacitor of $200.00

"Price" has nothing to do with "cost".

You should PRICE your goods/services at whatever the market deems appropriate. The market sets the VALUE of your goods/services. If you can't make a profit at that price, then avoid that market.

There is some psychology in NOT letting the customer know just how large the gap between your cost and price. Ideally, you don't want him to even begin to be able to assess your "cost"! This can breed resentment and bias further transactions with that customer (or, customers that are "recommended" to you).

E.g., if I quote a client $20K for a little project, I surely DON'T want to DELIVER the finished product to him a day or two later! Even if he was HAPPY with that price, he will subconsciously "do the math" and figure I charged him $1000+/hour! So, I'll "waste" some time and deliver the item to him at a later date where he can think that I "was really efficient" at delivering the product so quickly WITHOUT thinking he "got raped" in the process!

[Curious.George]

Re: Sending PWM through a relay

I can understand that and I done this myself because a lot of these trades people and trades company's like air conditioning plumbing and other trade like this it seems all they want to do is take as money as they can do and if you are very lucky you might get some one that does a good job and least you get some value for the money that you are spending

We had some guys out for quotes on painting the exterior of the house, many years ago. We were new to town so had to get a feel for the "going rate" (surely don't want to UNNECESSARILY pay at the HIGHER rate that was the norm where we'd come from!)

My first experience with exterior stucco (instead of cedar shingle) so I had no idea as to how porous it was and how much paint it would require. So, that was one of the questions I'd ask the contractors.

When I posed this to one guy, he promptly answered "300 gallons". (Remember, this is a single family detached home). I was so taken aback by his (obviously severely inflated) answer that I replied, "It's a house, not a frigging BATTLESHIP!!"

Most folks haven't a clue as to the types of questions to ask nor the range of "normal" replies to expect. These people are called "suckers".

[clearchris]

Re: Sending PWM through a relay

"Price" has nothing to do with "cost".

That's true to some extent. There is a price to be paid for knowledge and craftsmanship. I was far more butthurt paying the kid $500 for a cap and a kickstart because it was clearly basic crap that I would have figured out had I bothered to look at it first. And also because they thought I was clearly a sucker and might spend $1600 on a fan motor.

Alternately, I was happy to pay my hvac tech $75 to literally turn a screw, because I couldn't figure it out. My ice machine wasn't working, and I couldn't for the life of me figure it out. I spent far too long on it, called by guy, he shows up, takes a look, starts laughing, you are going to be pissed when you see this. He says I'm not going to charge you for this call, let me show you this. He takes out his screwdriver, turns an adjustment screw on the ice plate, says, your ice machine will work now. I insisted on paying him and gave him the check. I was glad to pay it though and I valued the help, I was glad he knew, because I wasn't going to figure that one out.

So really it depends, and there's a lot in the delivery too. If you are fresh out of a 2 week training course, probably best not to mention that

When I posed this to one guy, he promptly answered "300 gallons".

That's a wierd one. He could have been sarcastic, or he could just be an old painter. I started painting, with a respirator, because I met one too many old painters who were clearly driven to a very off mental state from inhaling too much paint fumes over the years. The fumes are real, especially with the oil paints.

[Curious.George]

Re: Sending PWM through a relay

There is a price to be paid for knowledge and craftsmanship. I was far more butthurt paying the kid $500 for a cap and a kickstart because it was clearly basic crap that I would have figured out had I bothered to look at it first. And also because they thought I was clearly a sucker and might spend $1600 on a fan motor.

I figure I'm at least as "smart" as 99.724% of the folks that I might encounter so it's a question of experience and tools. As an EE, when the condensor fan died, it was relatively easy to GUESS at the likely problem, verify it to be the case and then grimace at the $8 the local appliance place dinged me for the cap. But, I wanted it TODAY, not "3 days hence" and there's value to my time -- not having to hunt for a cheaper supply.

Alternately, I was happy to pay my hvac tech $75 to literally turn a screw, because I couldn't figure it out. My ice machine wasn't working, and I couldn't for the life of me figure it out. I spent far too long on it, called by guy, he shows up, takes a look, starts laughing, you are going to be pissed when you see this. He says I'm not going to charge you for this call, let me show you this. He takes out his screwdriver, turns an adjustment screw on the ice plate, says, your ice machine will work now. I insisted on paying him and gave him the check. I was glad to pay it though and I valued the help, I was glad he knew, because I wasn't going to figure that one out.

On my first visit to meet my future wife's folks, I accompanied my future father-in-law on his vending route. Electronic pintables had just come out and he (no "formal education") was really intimidated as he saw his livelihood in jeopardy -- HE wouldn't be able to repair these things! (at least, not the electronic parts).

He had just received one machine that wasn't working and asked me to look at it. (WTF? I don't have any equipment with me! I'm thousands of miles from home! What the hell can I do?)

I opened the head (the top part where the score displays are) to access the electronics (od machines had most of the "logic" in the belly of the machine and very little in the head). The displays were mounted on a swing-out panel -- like a door, of sorts. (older machines the electromechanical displays were accessed from behind).

The displays were PGDs (Planar Gas Discharge... sort of like neon bulbs except 7 segments). When the "door" was opened to the point that my vision was sighting down along the faces of the displays, I could see the plasma clouds "pulsing" as the displays were being refreshed. That shouldn't have been visible. The refresh frequency is tied to the crystal oscillator -- the only component that has a sense of "time" on the board.

Crystal was mounted tall/upright. I gave it a gentle "flick" with my fingernail and the game booted.

I was "surprised" (cold solder joint).

My father-in-law was AWESTRUCK! (I could have probably "had" BOTH his daughters, at that point! <grin>)

That's a wierd one. He could have been sarcastic, or he could just be an old painter. I started painting, with a respirator, because I met one too many old painters who were clearly driven to a very off mental state from inhaling too much paint fumes over the years. The fumes are real, especially with the oil paints.

No, he was just hoping to charge for 300G of materials and pocket the difference (cuz you KNOW he wasn't going to APPLY that much paint to the building!) as "extra profit". We regularly hear of folks who hire blindly and get taken to the cleaners. Or, get poor results.

We have several citrus trees and do a great harvest, each year (almost half a ton, total). Neighbors are envious -- what they don't see is all the effort that goes into ENSURING that harvest!

Neighbor landscaped their yard and wanted a Blood (Cara) Orange tree. You'd think he would ask for our advice BEFORE hiring someone to arrange the plantings? Nope. So, he's got this tree tucked into a corner -- two feet from 5 ft walls on two sides. Does he really think its ever going to grow/bear fruit?

By contrast, our trees are ~20 ft diameter and ~20 ft tall. Even the new plantings are 6x6x6 "balls".

He got ONE orange after three years. Why bother? We give him ~20 pounds of fruit each season, out of sympathy... (and, so he can see how sweet ours are!)

Same neighbor hired a guy to paint his roof -- instead of asking me how WE maintain ours. $1600 later... (we spend ~$120/year on roof maintenance; he'll get < 5 years out of that $1600 -- and likely STILL not ask our advice! (I've kept our roof operational for > 25 years; most neighbors have had theirs replaced, at least once, in that time)

Vanity? Pride? Stupidity? <shrug>

[clearchris]

Re: Sending PWM through a relay

By contrast, our trees are ~20 ft diameter and ~20 ft tall. Even the new plantings are 6x6x6 "balls". He got ONE orange after three years. Why bother? We give him ~20 pounds of fruit each season, out of sympathy... (and, so he can see how sweet ours are!)

Heh, I have one small orange tree, maybe 10x10ft and that's more than enough to get absolutely sick of oranges every other year.

Same neighbor hired a guy to paint his roof -- instead of asking me how WE maintain ours. $1600 later... (we spend ~$120/year on roof maintenance; he'll get < 5 years out of that $1600 -- and likely STILL not ask our advice! (I've kept our roof operational for > 25 years; most neighbors have had theirs replaced, at least once, in that time)

Is this for a metal roof? What do you do?

[Curious.George]

Re: Sending PWM through a relay

Heh, I have one small orange tree, maybe 10x10ft and that's more than enough to get absolutely sick of oranges every other year.

We juice much of it and freeze it. So "fresh squeezed" OJ for the entire year. Ditto with the lemon and lime juices.

Is this for a metal roof? What do you do?

We have a "membrane" roof -- similar to a "built up" (felt+tar) roof -- only different in construction. For felt roofs, one typically paints every 5-7 years as the sun just vaporizes the paint (the paint is formulated with lots of "solids" so it goes on VERY thick). You get about 60-100 sq ft of coverage per gallon so a modest home is ~25G of paint -- about $600.

I've learned that the REAL secret to keeping the roof intact is to inspect it annually. There are lots of protrusions that violate its integrity -- sewer vents, "attic vents", skylights, some folks have HVAC plants up there, etc. All are sealed with a tar-like compound. This dries out, over time. And, the building/roof "moves" (expansion/contraction as well as subsidence) which puts mechanical strain on that compound.

Little cracks form in these areas that are only apparent on close visual inspection. Easy to patch -- cut some of the old compound out and slop some new on.

Do this, religiously. Then, sweep all the accumulated plant detritus off of the roof (neighbors with tall pine trees, etc.) And, finally, paint the next ~20% of the roof (keeping track of where you left off, last year) with 5G of paint.

In this way, you paint the entire roof every ~5 years and manage to inspect it EVERY year before problems manifest. E.g., I replaced the lenses in the skylights last summer as it was obvious they were compromised.

So, while neighbors end up having leaks, YOU don't!

[The roof painting industry likes to foster the belief that the PAINT is what keeps the roof from leaking. No, the paint just protects the felt or, in our case, membrane. If you are relying on PAINT to "seal" your roof, you've got some big problems!!! As a result, a roof leak suggests a repaint -- after the patch -- even if you just had it painted LAST YEAR! Fools!!]

[clearchris]

Re: Sending PWM through a relay

It's ugly, but I'm calling it a minor success. Yeah, this board didn't look good before this, the chip was already fried, were it not already fried it would have been needed to dry out first (moisture sensitive part), and yes, I didn't remove the plastic parts before reflowing, but I just wanted to see if I could get the solder melted. No problems there. In retrospect, there were some issues though.

1) I read the wrong lead solder paste profile. Most are a bit colder than the one I used.
2) I programmed it on the faster end of things. A more gradual curve wouldn't have hurt anything and probably would have allowed the oven to heat up a bit and the final ramp to go quicker.
3) The way I programmed it, I think it was working against the pid control. I had set it up to update the setpoint every two seconds, but I'm not sure that took advantage of the Integral part of PID, i.e. the algorithm always thought the goal was close, so I don't think it ramped as fast as it could have. My testing showed that the oven could achieve 1.3C/sec up to 225 (IIRC, maybe 200C), and the oven was not hitting the numbers when I was ramping at 1C/sec. Granted, this time the oven had some load, but this was a puny bluepill board, with a negligible thermal mass.
4) I was hoping I'd not have to insulate the back, but I think I'll be bending some flashing tomorrow to make a back that I can stuff some insulation in. Once that is done, only the front door remains uninsulated. Hope that can give me a touch more heat. I'll probably have to fashion a gasket for the door, but it seals decently well right now.
5) I slapped on the paste too thick. And here I thought I was going thin. A little really goes a long way. I counted eight solder bridges.
6) A door opener might be required, or at least a buzzer to tell me to open the door.

What went right?

1) Well, the software did exactly what I asked it to do, though the end product wasn't right. That's always nice.
2) The chip ended up aligned pretty nicely on the pads, I was concerned about the fan possibly knocking it out of alignment, or possibly drift when the solder melted.
3) The solder melted pretty nice even though I applied too much. I didn't see any solder balls. No solder balls behind the pins afaik either.

I have some things to think about, and some tasks, but nothing insurmountable. In the meantime I have been getting good use out of the oven at lower temps, this may become one of my favorite tools.

So here's the result, warts and all. Truth be told, the board looked pretty bad going into the oven, so I can't pin it all on the reflow.

[Curious.George]

Re: Sending PWM through a relay

Digging through my literature archive and noticed this series of articles -- which can be skimmed for some of its less "technical" content to give a backgrounder.

[clearchris]

Re: Sending PWM through a relay

I added insulation around the oven door to seal the crack.
Made a back panel out of flashing to put insulation on the back outer wall.

Ran a test. Still 1.3C/sec. Grumble.

Double checked my halogens, one was 500w, the other was 350w. So I upgraded to 500w/500w. Now I'm at 1.4C/sec. I'd have liked to hit at least 2C/sec, but at that rate, I'd need another 900w to get to 2C/sec. Or more insulation. I don't have much room between the oven chamber and the outer metal case, it's packed pretty full of the ceramic insulation. The other option would be to add some reflective matting that seems to be the rage among reflow oven builders.

https://www.amazon.com/Design-Engine...000E243AW?th=1

In related news, I reprogrammed the gcode to get a better controlled slope where required and max output for when the required temp gain is closer to 1.4C/sec. Pretty happy with the slope. Near term I'm going to have to add at least a buzzer, longer term a stepper to open the door.



GC, thanks for the docs, I'll read up, though my latest test I'm getting only 5C overshoot with all the heaters on, which isn't that bad considering.

[clearchris]

Re: Sending PWM through a relay

Grumble, I usually like to set up external devices on a optocoupler (sharp pc123). My buzzer is a 3v3 buzzer, didn't think it was a problem, but apparently, it will not work with the optocoupler. I didn't know optocouplers had a minimum voltage. I guess I'll have to hook it up direct, but I generally prefer to have external devices isolated. Oh well. Going to have to read up on how to use calculate a resistor to do a voltage drop from 5v to 3v3. You would think I'd have run into this before.

[Curious.George]

Re: Sending PWM through a relay

My buzzer is a 3v3 buzzer, didn't think it was a problem, but apparently, it will not work with the optocoupler. Going to have to read up on how to use calculate a resistor to do a voltage drop from 5v to 3v3. You would think I'd have run into this before.

Assuming your "load" is largely resistive (when I think "buzzer" I think electromechanical device -- far from resistive!), then the isolated side of the isooptilator looks like 0.2V across the transistor (assuming there's enough gain to drive it into saturation). Then, the 3.3V that your load requires.

This leaves a difference of 1.5V [5.0-(3.3+.2)] -- neglecting tolerances. You have to "drop" this much across the resistor.

Now you need to know the current draw of the buzzer. You can measure that with a DMM/VOM by exciting it with 3.3V (it's rated operating voltage) and measuring the current through it.

Knowing that current -- I -- you solve for I * R = 1.5V.

Then, you also check (1.5V * 1.5V / R) to determine the wattage required of the resistor, adding sufficient margin to ensure it doesn't release any blue smoke.

Note that I must be within the current carrying capacity of the iso-optilator.

If your buzzer is not a pure resistive load, then the initial current may be higher than the "steady state" current that you measure (because your meter is slow and likely won't see the inrush current, if there is any). In that case, the buzzer might not start to "ring".

But, you can largely test this without the iso-optilator being involved: put R in series with buzzer and apply 5V (4.8V = 5.0-0.2) if you want to be pedantic) across the network and see if it "rings".

Note that R is in ohms, I in amps.

[clearchris]

Re: Sending PWM through a relay

Thanks! It's always good to see this written another way. Sometimes I regret skipping that year of HS and not taking physics.

[Curious.George]

Re: Sending PWM through a relay

Thanks! It's always good to see this written another way. Sometimes I regret skipping that year of HS and not taking physics.

I skipped the freshman "Earth Sciences" and just moved the other three years up by one. Rocks? Geology? Sheesh!

Biology was boring.

Chemistry was cool -- until the instructor put the kibosh on the compounds I was throwing together... :> (gee, no sense of adventure!) But, she was probably the hottest instructor in my "primary education".

The physics class was almost everything BUT physics! I recall spending a good bit of time on queuing theory (which I never was exposed to later in my education so found it very worthwhile). And, spent a lot of time playing with analog computers (no digital computers, at the time).