What small digital scale do you suggest?

[Rocketvapor]

I have a scale that reads FAST and at even gram loads like 1.000, 2.000, 3.000 ALWAYS reads correct.
In fact, as long as the test weight is within about 4 counts it always reads x.000 grams.
0.997 grams through 1.003 grams always reads 1.000.
1.004 grams sometimes reads 1.000 or 1.004.
This also applies to Grain loads that are at or near the even gram.
It's not bad at in between loads doesn't seem to drift much.
A to D converter/software probably has missing data points but is strange that the problem exists @ 1, 2, 3 - - 20 grams. Like the scale has a 1 gram repeating LOOP.
ALWAYS test sensitivity to SMALL changes, maybe with a kernel or two of Varget.

 

[Lefty2506]

For years I used a Dillon digital scale that I bought in the late 90s. Along the way, I've tried some of the RCBS and Lyman systems that would automatically trickle out the correct charge, but I always found that they tended to drift and loose their zero a LOT. A couple of years ago my Dillon scale died on me. I was in a pinch, so I bought a cheap 50gram digital scale from Amazon for around $20. It runs on a pair of AA batteries or off of a USB adapter, but I found that it's a lot more consistent when using the AA batteries. I suspect that's because the power coming in on the USB isn't very clean. It turned out to be the fastest and most accurate scale I've ever used. It's accurate to around +/-0.02gr. I took a sharpie and wrote the weight of one of my pans on the pan, and check the weight every time I pick the pan up off the scale. I'm not going to say that its on par with the expensive lab grade scales, but its far better than +/- 0.1gr that most digital powder scales are advertised to have. For the price, I decided to buy a 100gram version as a spare. I found similar issues with it with regards to using a USB adapter to power it. I also found that the accuracy was only half as good as the 50gram version which makes sense as I'm sure they are using the same 4-20ma or 0-10v analog signal for both applications. So my recommendation would be to go spend $30 on Amazon for a 50gram drug dealer scale

 

[DiffEQ]

I use a Belding & Mull powder drop and US Solid scale USS-DBS46-3.
I paid about $140 for mine back around 2018. They're $80 now. Probably the only reloading tool that costs less now than pre-pandemic pricing.

 

[Rocketvapor]

Full bridge load cells used in many scales usually have an output of around 3mv per volt of excitation. To eliminate MOST temperature effects in the electronics a ratiometric measurement is used. A STABLE voltage source is used to control both bridge excitation and AtoD reference (not separate sources). Reference goes up, or down and the ratio is maintained.

Loads cells can be somewhat temperature compensated but the Zero state is the most effected zone. Many scales will have a good bit of preload with platens to get some load on the load cell. Temperature must be stable or the compensation is chasing a moving target Air temperature is not always the best indicator of actual load cell temperature.

Perform a zero and full scale will probably be long term stable.
Tilt and a change in geographic location will impact full scale calibration.

 

[Dimner]

I'm learning a ton in this thread. Both from the wonderful information provided and the testing I'm doing.

I loaded some fire forming loads with the PL-50 last night. Man that scale is jumpy! Very fast and sensitive. The load weight I was going for was 36.00. To reach that weight was very difficult for the scale, I'm guessing due to rounding issues. Using Shooters World Precision, a smallish extruded like varget, I would pour and then trickle the powder. I would either end up on 35.98 or 36.02. I'd remove the pan and replace it to verify the weight. Doing this a couple times I would either add or subtract one granule or powder and then pour into the case.

After every 7th round loaded I would tare the scale, make sure the 2g check weight was 30.86 grains, place the pan and tare the scale, then place the check weight in the pan. Doing this made sub groups of the 25 rounds.

After 25 rounds, I checked the results with the TRX-925. I found the results really interesting.

Each sub group was very consistent +/-0.02gr between the 7 rounds. However when comparing the weights of subgroup A to Subgroup B, the powder weight was nearly .08gr different.

It seems as though, re-taring the scale made my powder weights vastly different. I have no idea why. I tried another 25 rounds and tared the pan before each single powder throw. Results were all over the place. I know this is a lot of detail and boring to most, but maybe someone can help explain what is going on?

 

[Rocketvapor]

It is possible the ZERO is a band a couple counts wide.
Scale senses you are close to zero and displays, yup, you guessed it, ZERO.
Maybe don't zero at zero, but some small weight.
Tare pan, add the 30.86gn 2 gram weight, in the pan.
Should read 30.86gn.
Add 36gn of powder, hopefully reading 66.86gn.
Dump on the TRX.

It is also possible that pressing on the Tare button is felt inside or disturbing level.

The scale in the youtube video I posted had a +/- dead band on even gram loads.
Test some pennies or dimes between the little 4 button black scale and your TRX.

 

[Rocketvapor]

I have a scale that was advertised as 0.001 gram resolution.
Ha 0.002 in gram mode, 0.05 in grain mode, but does have 0.005 carat resolution which IS one milligram resolution.
0.02% per count @ 10 g---- 0.01% per count @10 g---- 0.032% per count @ 10 g

 

[Webster]

The OP was gonna use this for a traveling scale, so a beam scale wouldn't be near as handy if you ask me -- which ya didn't. It doesn't seem to be a critical precision situation, and after all, everybody isn't Rocketvapor.

Lyman has a few basic electronics that I've got that are consistent for my purposes, and I believe I weigh +/- 0.1 quicker than I can with my beam scales.

Question??-- It seems like I've heard that the small scales are more accurate operating on the battery than with the power cord. Any truth to that?.. jd

He says he is going to the cabin. Does this mean he is just going to pop off rounds in the woods? Save yourself $100 and reload at home before you go to the cabin. If you buy a cheap scale you will probably never use it again when you get home. You said you could reload at a friends house when at the cabin. I assume he has the equipment.

 

[Rocketvapor]

Using a Good scale at home, or a Cheap scale at home it's always good to really evaluate your weighing process to see if errors are the scale, the weighing environment, or the Weigh Master. Some error SHOW, some DON'T.
I guess what you don't see is OK
Moving to another location adds even more uncertainty.

 

[AlNyhus]

And much of the obsession with exact weights borders on the ridiculous.

Ducking the bricks headed my way (weigh??)...

Good shootin' -Al

 

[Rocketvapor]

One small Nerf Brick coming at ya
I don't know if it's so much of an obsession with exact weights but maybe an obsession KNOWING how your weighing process is performing.
Those that spend the money for a GOOD balance probably expect performance to specs. Buying a cheap scale is a crap shoot. Some might be good, some not.
I think proving performance has merit.
If you only need +/- 0.1 grain, then watch your target.

 

[Dimner]

Yeah for me l, my testing and what not has to do with gaining confidence in the cheap scale for my purposes. Turns out the scale does heat when I do not use the tare function.

However, I got burned.by playing around with it too much. I found that the PL-50 has a calibration mode. So I ran it through that mode. It calls for a 50gram weight. I drop the weight on the platter and it says PASS. I got grains and try my 2 gram weight and it now shows. 31.96 grains. Dammit! So I try the 50 gram weight on the gram setting... 49.759 grams. Wait, what?

So all my readings are now skewed and not accurate. I think I trashed this scale.


Also @Webster , I'll reload at the cabin because it's something I like to do. I'll bring my stuff to my friends because all he has are lee dippers for powder measurement. Not sure why you assume my friends place is anywhere near the cabin. They are separate locations an hour between.

 

[Dimner]

And much of the obsession with exact weights borders on the ridiculous.

Ducking the bricks headed my way (weigh??)...

Good shootin' -Al

Cuz it's raining outside and I'm procrastinating doing laundry

 

[Rocketvapor]

A milligram scale, even good ones will likely NEED calibration before first use.
Unless you have the same local gravity as Canada, Japan, or even China where it was first calibrated. If your scale does not have a bubble level how can you place it level? How do you monitor scale temperature?
Sounds like this scale may be a lost cause.

 

[Doom]

For grins and giggles here is a test I ran on the GEM 20 when I first received it.


The test was to see what kind of repeatability the scale would have in two cases. One was zeroing the scale before each weighing and the other without zeroing before each weigh. The test test weight was a Nosler 60 gn Ballistic Tip.

Statistically the mean weight variation is statistically significant by Welchs T-Test but the standard deviation is not statistically significant by the F-Test. It is important to note that the mean difference is less than the scale resolution of 0.02 gn.

It is not possible to know what drift may occur in the scale when the drift is less than the display resolution. This is was to determine if this scale is subject to drift and if re-zeroing prior to each weigh would reduce the possible error. When the individual weigh points are plotted it appears that the weight in not zeroed case did decrease with time while the Zeroed case remained consistent as shown below. This is anecdotal evidence based on one test that the scale may be drifting. As a result I tare the scale before each weigh.

 

[Rocketvapor]

The low range (instead of 50 or 100grams) and the small load cell (low mass) for a milligram scale makes it good for trickling. Try your bullet and add 1, 2, 3, kernels of Varget to simulate final trickle to charge.

 

[Doom]

The low range (instead of 50 or 100grams) and the small load cell (low mass) for a milligram scale makes it good for trickling. Try your bullet and add 1, 2, 3, kernels of Varget to simulate final trickle to charge.

Aw, do I have to use Varget?

Just kidding.

 

[Rocketvapor]

The "One Kernel of Varget" is the universal test of scale sensitivity.

 

[Doom]

It took from 2 to 4 kernels to register a change. That was with the Bullet and also with just the pan zeroed. That is about what I expected from a cheap load cell based scale especially dropping one kernel at a time with tweezers from no real height.

I bought this scale as check scale for use with my Chargemaster. I'm not in need of one kernel resolution. All my shooting is 600yds or less and no competition.

Since I was in the files here is a test I ran on 28 charges of 8208 XBR thrown by the CM1500 at 25.4 gn. I didn't show it on the image but the 95% confidence interval based on this sampling is 25.3815 to 25.4125.

 

[Lefty2506]

I'm learning a ton in this thread. Both from the wonderful information provided and the testing I'm doing.

I loaded some fire forming loads with the PL-50 last night. Man that scale is jumpy! Very fast and sensitive. The load weight I was going for was 36.00. To reach that weight was very difficult for the scale, I'm guessing due to rounding issues. Using Shooters World Precision, a smallish extruded like varget, I would pour and then trickle the powder. I would either end up on 35.98 or 36.02. I'd remove the pan and replace it to verify the weight. Doing this a couple times I would either add or subtract one granule or powder and then pour into the case.

After every 7th round loaded I would tare the scale, make sure the 2g check weight was 30.86 grains, place the pan and tare the scale, then place the check weight in the pan. Doing this made sub groups of the 25 rounds.

After 25 rounds, I checked the results with the TRX-925. I found the results really interesting.

Each sub group was very consistent +/-0.02gr between the 7 rounds. However when comparing the weights of subgroup A to Subgroup B, the powder weight was nearly .08gr different.

It seems as though, re-taring the scale made my powder weights vastly different. I have no idea why. I tried another 25 rounds and tared the pan before each single powder throw. Results were all over the place. I know this is a lot of detail and boring to most, but maybe someone can help explain what is going on?

For me, trickling powder has always caused all of my digital scales to drift frequently. Ironically, my cheap Amazon digital scale drifts less than my Dillion or Lyman scales did (and also less than the RCBS/PACT scales I used to use). I weigh my pan empty, right after doing a calibration. For the sake of this example, lets pretend that my pan weighs 128.12 grains. While loading, if I pick up my pan do dump the powder into a case, the scale should ready -128.12. If it reads something that's off by more than .02, I consider re-zeroing everything. If the value is off by more than .04, I automatically tare and reweigh the charge that I just dumped. I've also used a sharpie to write the weight of the calibration standard in grains on the standard itself so I can quickly set it in my powder pan and make sure nothing has changed.

I've also gotten away from trickling in the standard sense. I throw my charge from an RCPS Uniflow that's mounted to my bench in an RCBS stand (not the sheet metal piece the thrower comes with). I throw directly into my pan, and then place the pan on the scale and weigh it. I have a second pan with a narrow spout that works great for scooping up a few kernels out of the charge, or for trickling a few more kernels into the powder charge. After 20 rounds or so you'll have a pretty good idea of how much you need to add or remove to get the charge where you want it. If its within +/- .04 grains, that's PLENTY good enough. For decades, our scales only read to 0.1 grains meaning that the exact same charge could ready 25.3, 25.4, or 25.5 grains and still be considered accurate. I can load this way a LOT faster than I could using something like an RCBS chargemaster.