Accurate Digital Scales

[Howard]

I agree most heartedly with AlanF
I've paid a lot of money for only a couple of scales, and always go back to the Redding.
Clean up the Redding knife edge vees with 1200 grip paper on a snap blade, and it will move with one granule.
Pulled an electonic one apart recently, and there's just too much to vary. Analogue Voltage Amplifers for a start.
And, I stick with my statement to someone recently, they only measure the movement in a spring block, and so, they are spring balances with a fancy readout.
(I suppose there are other ways it is done, but not for the dollars we're talking about.)
Not to take anything away from Pete's neat solution, my admiration is undiminished, but does it not merely just fix the auto zero problem?

Howard

[Bindi2]

I agree most heartedly with AlanF
I've paid a lot of money for only a couple of scales, and always go back to the Redding.
Clean up the Redding knife edge vees with 1200 grip paper on a snap blade, and it will move with one granule.
Pulled an electonic one apart recently, and there's just too much to vary. Analogue Voltage Amplifers for a start.
And, I stick with my statement to someone recently, they only measure the movement in a spring block, and so, they are spring balances with a fancy readout.
(I suppose there are other ways it is done, but not for the dollars we're talking about.)
Not to take anything away from Pete's neat solution, my admiration is undiminished, but does it not merely just fix the auto zero problem?

Howard

I don't always agree with AlanF but I do in this instance. He is on the button 100%. Howard it is good to see I am not the only one.

[AlanF]

I don't always agree with AlanF but I do in this instance.

Hmmm... maybe I need to rethink on this....

[pjifl]

Just for the record, the top end electronic scales do not measure the deflection of a spring.

They use a variable current in a magnetic field to lift the pan. Then measure the current with an AD converter and display it digitally.
This system has been proven the best but is more expensive. All lab scales work this way.

My conclusions are.

1/ Cheaper digitals are only good for quick rough checks. At times they seem good enough and suck you in but always let you down.

2/ Very expensive digitals are useful. They usually have a sticky zero but are fast and trustworthy. Especially if you get one with an extra decimal place than needed.

3/ Gravity balances are inherently reliable if made well and in many cases are my preferred option. Unfortunately most are not made well enough.
The knife edges and supports should be made out of other materials. But then they would be more easily damaged.
The ones we see are also designed for a different market which dictates they max out at more than 500 grain or something like that. This compromises their performance at the typical 40 - 70 grain range we need.

I have modified gravity balance scales and used them for years before eventually splurging on a quality Digital.

You mill away most of the beam and reduce its weight. Other swinging weight mods would be worthwhile but are harder without a complete redesign. The beam should be shorter.

Then you up the sensitivity by raising the centre of mass. (notice most old Lab scales have an adjustment for this above the central knife edges.)
The sensitivity can be increased by a bit of chewing gum on the top of the beam in an emergency. Simply add enough such that the beam is not top heavy.

After this, best to change the damping.

Eventually, like Grandad's axe........

After this, with good knife edges and surfaces for them to ride on, they are excellent but still slower than a good set of digitals.

Peter Smith.

[williada]

Sound advice Peter.

[Bindi2]

I don't always agree with AlanF but I do in this instance.

Hmmm... maybe I need to rethink on this....

[Howard]

Excellent info by Pete.
I was wondering about how the top end ones work, and every time I thought to look into it, I was away from the internet, and when I was on the net, I forgot!
That seems to happen a lot!!
The bugbear with powder scales is, as Pete says, 500 grains max! Who the hell is going to load 500 grains of powder, and set fire to it off their shoulder!
I've always wondered what to do to make standard scales better, and nowI have the urge to look into it, and modify a set.
A far as speed goes, I dunno. Even if it halves the time taken to actually weigh a load, it's still a small portion of the cleaning, sizing, priming, bullet seating cycle.
Also, just to create a bit of controversy. For each of us, does it really matter how accurate the scales are, i.e. an exact measurement accroding to the standard, as opposed to repeatability?
E.G. if my load, measured on my scales, is 40g, but in reality, measured on a million dollar set of scales, is 40.5, does that matter? If all my testing is done with MY scales, and as long as ALL my loads are done on MY scales, it's accurate repeatability I'm after.
And here is the problem with - admittedly cheap, spring digital scales - the repeatability suffers because of the zero jumping around.
Howard

[Pommy Chris]

Just for the record, the top end electronic scales do not measure the deflection of a spring.

They use a variable current in a magnetic field to lift the pan. Then measure the current with an AD converter and display it digitally.
This system has been proven the best but is more expensive. All lab scales work this way.

My conclusions are.

1/ Cheaper digitals are only good for quick rough checks. At times they seem good enough and suck you in but always let you down.

2/ Very expensive digitals are useful. They usually have a sticky zero but are fast and trustworthy. Especially if you get one with an extra decimal place than needed.

3/ Gravity balances are inherently reliable if made well and in many cases are my preferred option. Unfortunately most are not made well enough.
The knife edges and supports should be made out of other materials. But then they would be more easily damaged.
The ones we see are also designed for a different market which dictates they max out at more than 500 grain or something like that. This compromises their performance at the typical 40 - 70 grain range we need.

I have modified gravity balance scales and used them for years before eventually splurging on a quality Digital.

You mill away most of the beam and reduce its weight. Other swinging weight mods would be worthwhile but are harder without a complete redesign. The beam should be shorter.

Then you up the sensitivity by raising the centre of mass. (notice most old Lab scales have an adjustment for this above the central knife edges.)
The sensitivity can be increased by a bit of chewing gum on the top of the beam in an emergency. Simply add enough such that the beam is not top heavy.

After this, best to change the damping.

Eventually, like Grandad's axe........

After this, with good knife edges and surfaces for them to ride on, they are excellent but still slower than a good set of digitals.

Peter Smith.

Hi Peter,
My recent experience with some scales Josh recommended which only cost me 100 dollars is very good. It is true I have not has them very long, but apparently Josh has had his set for a bit now and they really seem very good and reliable. I am confident my loads are within 0.02 of a grain using this scale which is good enough IMO. My only quibble is it needs regular taring, but I think this is probably most digital scale. As I mentioned on my thread on the subject Peter made up three loads on his modded beam scales and I tested them on my digital scale and they were all exactly the same proving you can get extreme accuracy from a beam scale. He told me though he spends hours reloading to get this level of accuracy. I can get the same accuracy much much faster with the digital. Since my thread I have experimented dribbling onto the digital as it is not supposed to be possible on cheap digital scales. Well I had no problems doing it with mine, even adding one kernel of powder was not a problem. To save repeating myself too much here is a link to my thread. viewtopic.php?f=5&t=6366
Cheers
Chris